Citation
Characterizing snow avalanches recorded in lake sediments of central Colorado

Material Information

Title:
Characterizing snow avalanches recorded in lake sediments of central Colorado
Creator:
Hickman, Zara Kathleen ( author )
Place of Publication:
Denver, Colo.
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
1 electronic file (127 pages) : ;

Thesis/Dissertation Information

Degree:
Master's ( Master of science)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Geography and Environmental Sciences, CU Denver
Degree Disciplines:
Environmental sciences

Subjects

Subjects / Keywords:
Avalanches -- Colorado ( lcsh )
Lake sediments -- Colorado ( lcsh )
Avalanches ( fast )
Lake sediments ( fast )
Colorado ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
Snow avalanches are common to high elevations of the Colorado Rocky Mountains; yet, little has been done to reconstruct them in the past. This thesis examines sediment characteristics of geomorphic events recorded in high alpine lakes of central Colorado. Sediment grain-size analysis, geochemistry, and magnetic sediment properties were used to characterize snow avalanches from other allochthonous disturbances such as floods and density currents from two sites, Cottonwood and Mirror lakes in the Sawatch Mountain Range. While data from the combined proxies provided the best evidence of different disturbances, grain size analysis was the single best indicator. Snow avalanche deposits had poorly sorted sediments with a medium to high grain size and higher frequency of larger particles (coarse silts and very fine sands). Avalanche deposits were leptokurtic (peaked) with a 'fat-tail' due to higher occurrence of larger grain sizes. Flood event deposits were found to be better sorted with a high mean grain size (fine to very fine sands) and a more leptokurtic distribution lacking the 'fat-tail' characteristic of an avalanche event. Density currents were found to be the most sorted of the three disturbance events, with the highest mean grain size (sands). Finally, events are not recorded across the lake basin, which is based on the location of the disturbance event (i.e. avalanche path or stream mouth), the amount of energy when it energy the lake, and the dissipation of that energy. Localized geomorphic disturbances have left distinct signatures within lake sediments that can be used to reconstruct geomorphic disturbance history.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: Adobe Reader.
Statement of Responsibility:
by Zara Kathleen Hickman.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
on10069 ( NOTIS )
1006905760 ( OCLC )
on1006905760
Classification:
LD1193.L547 2017m H53 ( lcc )

Downloads

This item has the following downloads:


Full Text
CHARACTERIZING SNOW AVALANCHES RECORDED IN LAKE SEDIMENTS OF
CENTRAL COLORADO by
ZARA KATHLEEN HICKMAN B.S., State University of New York College at Oneonta, 2010
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment Of the requirements for the degree of Master of Science Environmental Sciences Program
2017


2
This thesis for the Master of Science degree by Zara Kathleen Hickman Has been approved for the Environmental Sciences Program by
Christy E. Briles, Chair
Andrew Gray Daniel Liptzin
July 29, 2017


3
Hickman, Zara Kathleen (M.S., Environmental Sciences)
Characterizing snow avalanches recorded in lake sediments of central Colorado Thesis directed by Assistant Professor Christy E Briles
ABSTRACT
Snow avalanches are common to high elevations of the Colorado Rocky Mountains; yet, little has been done to reconstruct them in the past. This thesis examines sediment characteristics of geomorphic events recorded in high alpine lakes of central Colorado. Sediment grain-size analysis, geochemistry, and magnetic sediment properties were used to characterize snow avalanches from other allochthonous disturbances such as floods and density currents from two sites, Cottonwood and Mirror lakes in the Sawatch Mountain Range. While data from the combined proxies provided the best evidence of different disturbances, grain size analysis was the single best indicator. Snow avalanche deposits had poorly sorted sediments with a medium to high grain size and higher frequency of larger particles (coarse silts and very fine sands). Avalanche deposits were leptokurtic (peaked) with a fat-tail due to higher occurrence of larger grain sizes. Flood event deposits were found to be better sorted with a high mean grain size (fine to very fine sands) and a more leptokurtic distribution lacking the fat-tail characteristic of an avalanche event. Density currents were found to be the most sorted of the three disturbance events, with the highest mean grain size (sands). Finally, events are not recorded across the lake basin, which is based on the location of the disturbance event (i.e. avalanche path or stream mouth), the amount of energy when it energy the lake, and the dissipation of that energy. Localized geomorphic disturbances have left distinct signatures within lake sediments that can be used to reconstruct geomorphic disturbance history.


4
DEDICATION
I would like to dedicate this thesis to my family, biological and accumulated. Their overwhelming support and encouragement was integral to my education and success. I want to also thank my parents for always encouraging me to follow my dreams wherever they may take me. Through them I discovered my love of mountains and the curiosity to explore them.


5
ACKNOWLEDGEMENTS
I would like to take the opportunity to thank my advisor, Christy Briles, for her time, patience, guidance, and passion. Her dedication to teaching and advancing paleoecology is an inspiration. I would also like to thank my committee, Andrew Gray and Daniel Liptzin. Lastly thank you to the faculty and administration of the Geography and Environmental Sciences department of the University of Colorado Denver for their support.


6
TABLE OF CONTENTS
CHAPTER
I. INTRODUCTION..............................................................1
II. BACKGROUND INFORMATION................................................... 5
Physical environment.................................................. 5
Geology................................................................5
Geomorphology......................................................... 5
Climate............................................................... 6
Vegetation............................................................ 7
Disturbance ...........................................................8
Snow avalanche causation and Colorado record..........................11
Proxy data and natural archive....................................... 14
Particle Size Analysis............................................... 15
iii. gis suiABiurY analysis...................................................is
Elevation............................................................ 18
Slope.................................................................18
Aspect............................................................... 19
Geology...............................................................19
Accessibility and Cost Path...........................................19
IV. SITE DESCRIPTION.........................................................22
V. METHODS AND DATA ANALYSIS................................................26
Field methods.........................................................26
Laboratory methods....................................................26
Core preparation, lithology...........................................26
Magnetic susceptibility & loss-on-ignition............................27
Geochemistry/XRF......................................................27
Particle size analysis................................................28
Data analysis.........................................................28
Geochemistry/XRF......................................................28
Particle Size Analysis................................................29
VI. RESULTS..................................................................31
Lithology.............................................................31
Magnetic susceptibility & loss-on-ignition............................34
Geochemistry/XRF......................................................36
Particle size analysis................................................37


7
VII. DISCUSSION............................................................41
Distinguishing between autochthonous from allochthonous disturbance.41
Characterizing different allochthonous events (e.g. snow avalanches, floods, debris
flows)..............................................................46
Localized differences within lake records...........................54
Building on existing literature from Norway and future research.56
VIII. CONCLUSIONS...........................................................60
REFERENCES................................................................. 63
APPENDIX..............................................................70
A. MAGNETIC SUSCEPTIBILITY FOR COTTONWOOD LAKE (CWL15).................. 70
B. MAGNETIC SUSCEPTIBILITY FOR MIRROR LAKE (ML15EBC).................... 75
C. MAGNETIC SUSCEPTIBILITY FOR MIRROR LAKE (ML 131A) ....................78
D. LOSS-ON-IGNITION FOR COTTONWOOD LAKE (CWL15)..........................93
E. LOSS-ON-IGNITION FOR MIRROR LAKE (ML15EBC)............................96
F. LOSS-ON-IGNITION FOR MIRROR LAKE (ML131A).............................99
G. GEOCHEMISTRY FOR COTTONWOOD LAKE (CWL15)............................ 104
H. GEOCHEMISTRY FOR MIRROR LAKE (ML 15EBC)............................. 107
I. PARTICLE SIZE ANALYSIS (CWL15).......................................110
J. PARTICLE SIZE ANALYSIS (ML15EBC).............................. 114
K. PARTICLE SIZE ANALYSIS (MI. 131 A)............................ 117


8
LIST OF TABLES
TABLE
1. Summary of avalanche manifestations...............................................12
2. Statistical formulae for particle size parameter calculations..................... 17
3. GIS suitability analysis data summary............................................21
4. GRADISTAT grain size statistical summary.........................................30
5. Disturbance event summary for Cottonwood and Mirror Lake.........................51


9
LIST OF FIGURES
FIGURE
1. Identification of avalanche path zones........................................13
2. Procedural flow chart for suitability analysis................................20
3. Site map for Cottonwood Lake and Mirror Lake..................................23
4. Core lithology, loss-on-ignition, magnetic susceptibility, and Rb/Sr-ratios...32
5. Core particle compositions....................................................39
6. Particle Size Analysis Summary................................................40
7. Geomorphic Disturbance Summary................................................47


1
CHAPTER I INTRODUCTION
Snow avalanche events are large and quickly displaced masses of snow transported down steep slopes (Luckmann, 1977), and common at high elevations in mountainous terrain. They pose a significant threat to winter sport enthusiasts, tourism, and transportation. Where snow avalanches encounter boundaries of society, they can be costly to infrastructure and tourism (e.g. ski resort management). In Colorado alone, seven people on average lose their lives due to snow avalanches each year (Colorado Avalanche Information Center, 2016). They also impact ecological systems directly by removing vegetation, such as toppling down forests in their path (Bebi el al., 2009; Baker, 1992). By extension, avalanche paths can define growth patterns of vegetation on a hillside (e.g. shrubs versus conifers), and ultimately determine what plant communities can exist in an area (Baker, 1992; Veblen, 1994).
It can be difficult to quantify snow avalanche activity and frequency (Pain el al., 1998). Besides recorded events around residential, recreational areas, and roads (Atkins, 2006), another way used to examine them is by identifying and dating tree rings that reflect scaring or differential growth patterns resulting from an avalanche event (Bebi et al., 2009; Pederson el al., 2006; Szychowska-Krapiec and Krapiec, 2001). Within central Colorado, the region of focus for this thesis, studies using dendrochronological tree-ring data have been conducted to expand historic frequency of snow avalanche events in more remote landscapes (Simonsin, 2012; Veblen, 1994). However, the tree scar method can be problematic if an area experiences multiple avalanches, if the event kills many or all the trees resulting in a low sample pool, and if the tree species involved does not respond or produce quality regrowth scars (Reardon et al., 2008; Veblen, 1994; Corona, 2012; Shroder, 1978; Bebi et al.,
2009).


2
Lake sediments are an under-used archive for reconstructing avalanche events. Lakes, unlike trees, often accumulate undisturbed sediment for thousands of years, and can continuously record events that integrate signals from an entire watershed. This sediment is deposited overtime by organisms living in the lake, or allochthonous sediment from flood or debris flow (Vasskog et al., 2011). Local geomorphic activity (floods, slope failure, snow and rock avalanches, etc.) and other environmental variables (wind-blown erosion) influence depositional transport of sediments to the watershed. For example: snow avalanches deposit allochthonous course-grained inorganic sediments and plant remains that are significantly different from year-to-year production of organic fine-grained autochthonous sediments. Lake sediments are commonly used to reconstruct historical ecological patterns and changes, using proxies such as pollen and charcoal, but often lakes with layers of disturbed allochthonous sediment with low organics, air pockets, and sometimes large macrofossils, are overlooked and discarded as the material disrupts the gradual sedimentation rate of autochthonous organic sediments (Briles et al, 2012; Del Priore, 2015). In this thesis, using a multi-proxy approach,
I demonstrate how disturbed allochthonous sediment presents an opportunity to explore material that may be evidence of past disturbance events. Until now, the only lake sediment records that have used sediment properties to reconstruct avalanche events are those from Norway (e.g. Nesje et al., 2007; Vasskog etal., 2011; Vasskog etal., 2012; Corona etal., 2013).
Grain size distribution is one of the most fundamental properties of lake sedimentary deposits (Folk and Ward, 1974). The distribution of grain size within a lake deposit is influenced by parent material, as well as transport and deposition processes (Beierle et al., 2002). Particle size distribution of lake sediments as a paleoenvironmental proxy has been used to help reconstruct disturbance and origin through particle transport, and to distinguish between transport processes (e.g. snow avalanches and flooding events) (Nesje et al., 2007; Vasskog et al., 2011; Jin 2005). Through multiple proxies (i.e. magnetic susceptibility, loss-on-ignition, and geochemistry) and the interpretation of grain-size


3
distributions using statistical tools, the type of mass-wasting events can be determined (e.g. snow avalanche versus flooding event) (Vasskog et al., 2011; Wilhelm el al., 2013).
Given that lake sediments have yet to be used outside of Norway to reconstruct snow avalanche history, the main objective of this thesis research is to determine the differences between autochthonous sediments, and allochthonous instantaneous disturbances, in the southern Rocky Mountains (Sawatch Mountains) of Colorado. Proxy data was used to distinguish methods of sediment transport and deposition (e.g. snow avalanches versus flood event), allowing for the characterization of snow avalanches from other mass wasting events. Proxies used in this study included: magnetic susceptibility to measure iron-bearing elastics; loss-on-ignition to measure lake productivity; x-ray fluorescence to determine sediment geochemistry; and particle-size analysis to describe sediment grain size distributions. A Geographic Information System (GIS) was used to determine suitable lakes based on slope, aspect, parent material, and location of visible avalanche shoots into the lake. The study addressed the following questions:
1. What characteristics distinguish autochthonous and allochthonous/disturbance events in a lake sediment record?
2. What sediment properties differentiate snow avalanches from other geomorphic disturbance (flood, debris flows, etc.) in lake sediments?
The thesis is developed around seven chapters. Chapter two describes central Colorados present-day physical environment, geomorphology, geology, climate, vegetation, and dominant disturbance regimes. It also discusses the nature and uses of proxy data for reconstructing past environments. Chapter three describes the study area and characteristics of Cottonwood Lake and Mirror Lake. Chapter four details field and laboratory methods and how the lake sediment, particle size data, and geochemistry and were analyzed. Chapter five presents results of the study, including lithologies of the cores, particle size distributions, magnetic susceptibility, loss-on-ignition, and geochemistry data. Chapter six discusses how the proxy data can be used to characterize avalanche


4
signals from other geomorphic disturbance events that result in an interruption of autochthonous sediments into the basin. Chapter seven summarizes the findings of the study.


5
CHAPTER II
BACKGROUND INFORMATION Physical environment
The Sawatch Mountains of central Colorado, are part of the southern Rocky Mountain chain. The Rocky Mountains stretch south from New Mexico north into Alaska, within western North America. The Colorado Rocky Mountains begin just west of the Front Range highland (1600 m) and extend across the state to the western most boundaries. They included over 300 peaks rising above 4,267 meters (m), and the second tallest mountain in the continental United States (Mount Elbert, 4,401 m elevation). The Colorado Rockies are characterized by high peaked ranges, with deep cut valleys or parks dispersed throughout the ranges (Pazzaglia and Kelley, 1998). The Continental Divide is the primary hydrological drainage barrier within the Rocky Mountains and separates distinct drainages between the Atlantic Ocean (east), Gulf of Mexico (southeast) and Pacific Ocean (west).
Geology
The ancestral Rocky Mountains first formed by processes of rapid uplifts from the North American and South American plates beginning approximately 300 million years ago (mya) during the Pennsylvanian Period throughout the western United States and Canada (Kluth and Coney, 1981). The Southern Rocky Mountains (north-central New Mexico north through Colorado and into southeastern Wyoming) experienced a secondary intense geologic uplift during the Laramide orogeny (Late Cretaceous, approximately 70-40 mya) resulting in the topography and relief of todays mountains. The uplift of the Sawatch Range created a separation between the Taylor Park and South Park regions and are an anticline (Tweto, 1975; and Brugger, 2006).
Geomorphology
Central Colorado experienced two Pleistocene glacial periods with the latest ending approximately 18,000 years ago. Due to the regions high elevation, it experienced both rock and ice


glaciers (Refsnider and Brugger, 2007). The Sawatch Range was characterized by glacial activity that significantly influenced the formation of valleys and cirque basins of the surrounding landscape
6
(Barsch, 1977). The transient deglaciation process created the regions significant topographic relief and many high elevation kettle lakes (Brugger, 2006). Prominent Pleistocene moraines exist in both Taylor Park and the upper Arkansas River valley (Refsnider and Brugger, 2007). Glacial extent of the Sawatch Range occurred from 16.3 +/- 1.6 to 22.2 +/- 2.8 ka based on exposure ages on boulders (Brugger 2006, Refsnider and Brugger 2007). These last glacial maximums (LGM) advances correspond with others calculated throughout the Southern Rocky Mountains (Fall, 1997; Brugger, 2010).
Climate
Colorado climates are influenced by regional geography, specifically by elevation and topography, resulting in microclimates. Colorados intercontinental North American location is not influenced by normal maritime precipitation regimes. Summers at lower elevations are typically hot and dry, while in the mountains they are often cool and wet due to mountain thunderstorms. Winters are cool and dry at lower elevations, while cold with high snowfall at higher elevations in the mountains. Overall, Colorado climate is characterized as cool and dry with low humidity (semi-arid). Regional temperature averages range between 6.9 C (Buena Vista, 2427 m; east of the Continental Divide) and 0.9 C (Crested Butte, 2715 m; west of the Continental Divide). Current average rain precipitation ranges from 27 cm per year (at Buena Vista) and 61 cm at Crested Butte (climatic data from Western Regional Climate Center [http://www.wrcc.dri.edu/summary/climsmco.html]). Much of the regions precipitation occurs within the winter months, December to April. Winter precipitation, primarily snow fall, factors into local snow pack equivalence that results in snow melt run-off in the spring and summer months. Average snowfall in the region ranges between 104cm (at Buena Vista) and 551 cm per year at Crested Butte (units in snow water equivalents, precipitation averages from US Climate Data [http://www.usclimatedata.com/climate/Colorado/united-


states/3175]). Westerly prevailing winds are prominent, but are strongly determined by local topography. Due to steep and sharp relief, climate ecotones exist with rising elevation (Mast et al.,
7
1998).
Air masses from the Pacific Ocean in winter months bring snow to the west of the Continental Divide, but this rarely impacts precipitation patterns east of the Continental Divide (Doesken et al., 2003). The western slope, including the study area, receives higher accumulations of snow in the winter, than the Front Range due to orographic lifting and the rain shadow that result on the leeward side of the Rockies (Fall, 1997a).
Vegetation
Local and regional topography plays a significant factor in flora distribution and biomass due to the elevation ecoclines (thermoclines) within the Southern Rocky Mountains (Stohlgren, 2000). Increased elevation thermoclines result in higher levels of precipitation and lower overall temperatures. Vegetation in the Rocky Mountains is divided into four zones that correspond to rising elevation conditions, these include: shrub-steppe and grassland, (2,300-2,900 m); montane (2,800-3,000 m); subalpine (2,900-3,400 m); the alpine (> 3,500 m) (Fall, 1992b). Montane forests are composed of drought resistant trees such as lodgepole pine (Pinus contoria, Pseudotsuga menziesii and P i m isponderosa). Subalpine forests are dominated by Engelmann Spruce (Picea engelmannii) and subalpine fir (Abies lasiocarpa) forest. The alpine zone presents a significant ecotone associated with tree line that limits tree growth to under 4 meters (Aplet et al., 1988).
At high elevations, the presence or absence of vegetation can impact disturbance activity.
Due to the more extreme climatic and topographic conditions of higher elevations, a higher probability for soil erosion exists (Meusburger et al., 2010). Higher concentrations of vegetation biomass was found to decrease rates of soil erosion caused by disturbance events in alpine environments (Martin et al., 2010). Vegetation intercepts rainfall and increases evapo-transportation, which then reduces runoff flow and channel creation, a positive feedback that leads to increased


disturbance event impacts, such as landslides and mud debris flows (Bochet et al., 2006). Areas above the tree line ecotone with lower percent vegetative cover coincide with increased rock slide
8
activity and soil erosion (Meusburger et al., 2010).
Disturbance
An environmental disturbance is a discrete event that alters a physical system and has profound and immediate impacts to the surrounding landscape with varying long-term (decadal to centennial scale) effects (White and Pickett, 1985; Baker, 1992). Although the overall health of the system may not experience a negative drastic change by each individual event, there are both discrete and cumulative impacts to regional geology and ecosystems. Community ecological health, such as species richness is found to increase in the presence of intermediate disturbance (Connell, 1978; Lubchenco, 1978). Due to the size and infrequent nature of large mass-wasting events few studies exist that encompass the entirety of ecological responses due to lack of experimental control (Michener and Haeuber, 1998). There are several types of disturbance events which occur in central Colorado. Disturbances to this region can range from small and semi-annual to large and infrequent having varying impacts due to the event type, scale, and frequency (Paine et al., 1998). Fire is considered the most frequent and dominant disturbance event in the central Colorado region during warm, dry summer months (Anderson, 2008a). Colorados fluctuating temperatures and precipitation can also trigger events such as snow avalanches, floods, and debris flows within the monsoon (June to September) and winter seasons (Vivoni et al., 2006). Such events can alter watershed flow dynamics, as well as cause the triggering of disturbance from surrounding topographic relief.
Fire events are common disturbances of central Colorado that can shift montane to subalpine forest vegetation structure on varying spatial scales, and influencing forests from decades to centuries (Whitlock et al., 2010). Severe fire events occur approximately 100 to 300+ years in the subalpine forest of central Colorado (Fall, 1997). Much of what is known about historic fire regime has been recovered through proxies, such as charcoal in lake sediments and tree-ring fire scars. Climate is a


dominant factor influencing fires, particularly during ENSO cycles, which increase the impacts of an already dry arid environment (Higuera et al., 2014; Schoennagel et al., 2005; and Sherriff et al.,
9
2001). The presence of potential fuel (biomass) located throughout the subalpine forest and canopy influence overall fire ignition and spread.
During the winter months when precipitation is dominated by snow fall, disturbance can occur in the form of snow-avalanches. The semi-arid environment of Colorado results in a dry snow pack that is not as cohesive or binding as a maritime regional snow pack (LaChapelle, 1969). Avalanche slides can be unpredictable, but typically slide on topographic slopes with angles from 25 to 50 and when: (1) stress to existing snow slabs increases such as with significant precipitation accumulation; (2) strength of an existing slab decreases, such as above freezing temperatures, causing the deterioration of existing internal snow layers; and (3) propagation of the initial failure, or a positive feedback loop once the initial slide occurs (Voight et al., 2011). Other factors correlated with snow avalanche activity are the terrains directional aspect, wind patterns, underlying geological composition, intensity of solar radiation, local vegetation, and the local average temperature fluctuation.
Flood disturbance events occur infrequently due to extreme localized summer precipitation events (Wilhelm et al., 2013). These disturbance events influence watershed dynamics such as tributary and channel structure and rate of sediment plumes (Schillereff, 2014). Watersheds experience annual hydraulic pulses due to precipitation and spring snow melt from higher elevations, but large flooding events are rare. Due to irregular frequencies, little is known about large scale flooding (Michener and Haeuber, 1998). There is an increased frequency of flood events that correlate with warming climates (e.g. Medieval Climate Anomaly 800-1300 A.D.) (Wilhelm et al., 2013). With flood frequency and intensity increasing because of warming climates (Knox, 2000) it is important to derive information from historic flood events and resulting ecological response. Lake sediments are an archive of flood disturbance when they have an inflowing stream. Flood events


10
produce a sediment plume within a lake composed of clastic material and organic matter that leave behind coarse-grained laminated layers (Lowe and Walker, 1997; Schillereff et al., 2014). Besides event frequency, studies on flood deposits in lake sediments have also been able to determine severity (Lowe and Walker, 1997).
Rock landslides are also localized infrequent disturbance events common to the high elevations of central Colorado. Rock slides and debris flows although bound by similar factors of snow avalanches, can vary in characteristics due to the composition and causation (e.g. slope, geological composition, and climate). Rockslides are composed of rock mixed with earth, snow, or ice (Voight et al., 2011). They occur more often in higher elevation areas with slopes ranging from 25-40 above tree line where vegetation significantly decreases and the landscape is composed of stronger materials such as exposed loose soil, cobbles, and boulders (Hungr, 1995). Debris flows have longer displacement paths and can be fluid and flow-like in nature due to saturation (Hungr, 1995). Mud flows occur after intense rain storms where runoff loosens debris on steep terrain.
Debris flows are a major cause of slope erosion and scour the landscape with deep channel-like tracks (Berti et al., 1999).
Density currents, or turbidity currents occur due to failing slope sediments above or within a body of water and/or turbid plumes of water introduced to the internal slope that move quickly downhill, releasing a mass suspension of sediment into the water column (Gould, 1951; Mortimer, 1971). Density currents, which are layers of sediment rich in water that enter less-dense, non-turbid waters and flow in layers beneath the surface (Lamb and Mohrig, 2009). Density currents are often caused by debris flows due to slope instability quickly introducing sediments into the lake (Zhang et al., 2014b).
Mountain (Dendroctonus ponderosae) and spruce beetle (Dendroctonus rufipennis) outbreaks are well documented and are becoming an increasingly more prominent issue for forest management. The pine beetle is native to the Rocky Mountains, lives and breeds primarily within ponderosa, lodge


11
pole, scotch and limber pine tree species. D. ponderosae is showing exponential increases in comparison to its historic population due to changing climatic factors (Carroll et al 2003; Chapman el al., 2012). Open forest resulting from extensive beetle kill decrease snow stability leading to increased avalanche potential under ideal weather conditions. After the trees root systems decay, the outbreaks can also lead to slope instability and increased sediment inputs to streams and waterbodies (Ryan et al., 2014).
Snow Avalanche Causation and the Colorado Historic Record
On February 2, 2016, an avalanche occurred west of Buena Vista on Cottonwood Pass that resulted in one fatality. A storm began on January 31 and ended February 1 with approximately 15 inches of snow fall. Temperatures were consistently low throughout the storm and fluctuated between -20 to -15 C until February 2. Wind speed ranged between 16-24 km/h that could have wind loaded steeper slopes. The unintentional release (u- An classification) occurred on a southeastern facing slope with a 36 slope by a recreational snow biker. As the recreationalist ascended to steeper terrain, the start zone of the avalanche was triggered from above burying the victim. Due to the cold temperatures and added stress from the new snow fall, snow crystalline patterns did not coalesce, resulting in a slide (Baker, 1986).
Avalanche events are caused by complex mechanisms that vary due to climactic conditions which make them difficult to predict. (Schweizer et al., 2003). However, it is known that avalanche formation is most strongly controlled by the interaction between the regional topography, local climactic conditions, and existing snowpack. Added stresses that can release snow slabs result from quick surficial loading in proximity to the start zone (Figure 1), progressive loading to the start zone due to precipitation, and random releases (Schweizer et al 2003). The triggers of slide activity are uncertain and are divided into two groups: dynamic and static (Luckman, 1977). Static factors include slope, topographic angle and aspect, and geologic composition. Dynamic variables are weather related, such as temperature fluctuation and wind loading (Butler, 1986). Due to difference


12
in meteorological changes throughout the season there are several classifications of avalanche manifestations: mass movement, slush flows, and creeps (Schweizer et al., 2003). Table 1 provides a description of the avalanche types and their triggers.
Table 1: Summary of primary avalanche manifestations, descriptions, and triggers
Avalanche Type Description Trigger Variables
Mass movement (dry-snow) A dry, intercontinental snow pack. Release begins underneath a strong cohesive slab (Schweizer, 1999). Snow stratification is rate dependent, meaning overall stress must exceed slab layer strength (Narita, 1983). Wind loading, precipitation, random release.
Slush flow (wet-snow) Occur in intercontinental mountain ranges in the spring when temperatures begin to rise consistently (Baggi and Schweizer 2008) and snowpack is saturated with runoff water. Primary slide manifestation in ranges characterized by maritime snow pack (ex. Sierra Nevada). Water cycle, topography, snowpack stratification, air temperature, rain, bedrock.
Creep Slow deformation of snowpack (Mathews and Mackay 1963) Topography, temperature fluctuation


13
Figure 1: Identification of avalanche path zones represented by Mirror Lake's existing paths on the northern slope of the lake. (Photograph credit: Zara Hickman)
Several key terms are used in describing avalanche movement and path. The starting and run out zones are important for defining the spatial nature of a slide event. The starting zone is where the slide first begins to slip. This is primarily due to the inconsistencies in snow crystal formations, but can be a reaction to wind loading or precipitation (Hungr, 1995). As the movement travels down the track, the final resting place of the mass disperses throughout the run-out zone. The run-out zone is where most of the debris and material collected by the event is displaced and deposited.
Snow avalanches have profound impacts to local ecosystems and landscape that are immediately felt (Pain et cil., 1998; Baker, 1992). Although the overall health of a system is not drastically changed by each disturbance event, there are impacts to regional geomorphology and vegetation. Avalanche paths can redefine growth patterns on a hillside, and ultimately determine


14
what vegetation can exist in an area. Such key events can create natural patches and variability as well as overall health in surrounding environments and ecosystems (Baker, 1992; Veblen, 1994).
Much of what is known about Colorados avalanche records exist in observational accounts (Atkins, 2006). However, tree-ring data has been used to establish historic frequency of events and path history (Alestalo, 1971). Although results of such studies have produced evidence for events, this method can only establish minimal frequency at best (Simonsin, 2012; Veblen, 1994). Dendrochronological studies from Glacier National Park (GLAC) focused around existing avalanche paths and determined a lack of event frequency resolution in times of known recent activity (Butler et al., 1979). Since remote areas lack infrastructure, avalanche accounts are lacking. In Colorado, treering dating methods have been conducted in Ophir, but a study like that of GLAC has yet to be conducted to determine resolution reliability (Carrera, 1979).
Proxy data from natural archives
Proxy data preserved in natural archives (such as lake sediments, ice cores, tree-rings, etc) allow for the reconstruction of past ecosystems, environmental change, including disturbances to the landscape on a range of temporal and spatial scales. Paleoenvironmental and paleoclimatic data can be gathered from, but not limited to, lake and ocean sediment cores, ice cores, tree-ring records, corals, and packrat maddens. Not all proxy data can provide the same level of resolution or accuracy. Tree-ring records and corals provide annual data records of temperature and precipitation; however, extent is limited to centennial to millennial scales in the case of tree-rings (Andersen et al., 2008). Lake and ocean sediments, and the material they preserve, such as pollen and charcoal, lack resolution, but can extend back millennia. Lake sediments also preserve data that can be used to reconstruct depositional history, regional climate variability and localized geomorphic disturbance (Rothwell et al., 2006). But uniting multiple threads of evidence into a comprehensive analytical framework, a multi-proxy approach can provide a more robust analysis with higher overall resolution of ecosystem processes and change than achieved through one alone.


15
Sediment stratigraphy and lithological properties provide information on the occurrence and impacts of individual disturbance events within proximity to the lake field site (Vasskog et al., 2011; Nesje, 2007). Magnetic susceptibility can be used to identify allochthonous pulses from potential erosional events (Gedye et al., 2000). Loss-on-ignition data examines lake productivity and allochthonous sediment deposition. Geochemistry and elemental ratios are used to determine depositional history and chemical erosion relationships with local geology (Croudace et al., 2006; Dasch, 1969; Jin et al, 2006). Particle size analysis is used to determine deviation from normal distributions and to characterize transport and depositional processes through several different parameters, such as mean grain size, sorting, skewness, and kurtosis (Folk and Ward, 1957; Blott and Pye, 2001).
Particle Size Analysis
Grain size characteristics are an elemental part of particle size analysis. Paleoecological studies have utilized particle size characteristics to understand geomorphological changes to the landscape and resulting watershed impacts (Folk and Ward, 1957; Weltje and von Eynatten, 2004). Particle size distributions (PSD) along with natural proxy data can provide insight to a watersheds sediment origin, sediment transportation process, and depositional characteristics (Folk and Ward, 1957; Bui et al., 1990; Friedman and Sanders, 1978). Environmental factors such as geology, climatic conditions, regional weather, vegetation, and geomorphology influence the chemical and physical weathering regimes that control the grain size of eroded sediments (Jin et al., 2001; Weltje and von Evanatten, 2004). Fluvial sediment sizes are further influenced by transport and depositional processes, such as water fluvial flows, aeolian, or due to topography within the channelized system (Watson et al., 2013).
The differences in specific particle size characteristics can be used to interpret transport methods as varying mechanisms (e.g. sorting, skewness, kurtosis) transport different particle sizes


16
reflecting different movement (Garrow, 1982; and McLaren and Bowles, 1985). For example, recent studies have utilized particle size analysis to understand changes to historic estuary dynamics (Watson et al., 2013). PSD has also been used to understand the role climate plays on flood events, resulting hyperpycnal density currents, and other geologic hazards (Zhang, 2014a; Zhang et al., 2014b). Particle size distributions were used to show that most turbidities were caused by extreme floods, rather than slope failure, suggesting climate change implications (Zhang, 2014a; Zhang et al., 2014b). Increasingly, particle size analyses have been used to understand increased erosional processes due to wildfire (Wondzell and King, 2003). The understanding of these processes, and impacts to deposits, is an important tool in understanding often infrequent geomorphic disturbance (snow avalanches, rock avalanches, floods, and density currents) and their landscape impacts (McLaren and Bowles, 1985).
Estimation of particle size distribution can be accomplished through a variety of methods including, but are not limited to, direct measurements, sieving, and laser granulometry (Blot and Pye, 2001). Regardless of method, these techniques are used to describe the variation of particle diameter (D) within a population of particles. Various statistical packages and statistics aid in the interpretation of particle size characteristics by visualizing distribution curves. Statistical results are presented in two prominent units: phi transformations, and metric units. Phi units, have been termed Tog-normal and transform values to from the metric scale (D in millimeters) using the following equation (0 = log2D) (Folk and Ward, 1957; Blott and Pye; 2001). Plotting multiple particle size characteristics can often reveal relationships and trends that otherwise would not have been seen. Common variables include mean particle size, standard deviations, sorting, skewness, and kurtosis (Krumbein and Pettijohn, 1938; Folk and Ward, 1957). Mean particle size (graphical mean.*,,) is the average diameter of grain size (Table 2); where m is equal to the mid-point of each class interval for both metric (mm) and phi unit (0m). Sediments are then broken down into texture classes by fraction (clays, silts, and sands) (Blot and Pye, 2001; Table 3). Standard deviations ((Jg), or sorting, determine


17
the variance of a defined population assuming normal conditions (Table 2) (Blot and Pye, 2001). Sorting then refers to the spread of the particle size distribution, reflecting how mixed the sediment is from well-sorted (low values) to poorly-sorted (high values). Skewness and kurtosis are used to describe distributions as then differ from normal distributions (e.g. Gaussian Distribution). Skewness (Skg ) is further used to determine shifts of distributions are fine skewed or coarse skewed (Blott and Pye, 2001). Whereas kurtosis (Kg) refers to grains relative to average and how peaked the distribution is (Table 2) (Blott and Pye, 2001). Although phi units have been preferred traditionally by sedimentologists, metric units are used more commonly by the geological community as resulting parameters are easier to interpret (Blott and Pye, 2001). This study utilized metric units for this study to most easily compare results to pervious literature investigating historic geomorphic disturbance (Nesje etal., 2007; Vasskog et al., 2011).
Table 2: Statistical formulae for the particle size parameter calculations (Krumbein and Pettijohn, 1938; Folk and Ward, 1957; Blot and Pye, 2001).
Phi Unit Graphical Mean Standard Deviation Skewness Kurtosis



18
CHAPTER IV
GIS SUITABILITY ANALYSIS OF LAKES PRONE TO AVALANCHE ACTIVITY
Site selection of Colorado lakes more prone to local snow avalanches was determined through a suitability analysis using a GIS with ESRI ArcGIS 10.2 software. Five criteria, elevation, slope, aspect, geology, and accessibility were used based on physical environment and regional geomorphology literature that highly influence avalanche triggers and events (Hungr, 1995). The criteria and workflow used within suitability analysis are discussed below (Figure 2).
Elevation
Snow avalanches and rock slides can both produce disturbance events. Although rock avalanches can transport ice and snow, these events are composed primarily of large rocks and debris. Snow avalanches are a mass movement of snow that can transport and mix of slope material and debris (Table 1). Rock avalanches, also common to central Colorado, occur most frequently in tundra zones where trees, larger shrubs, and ground vegetation do not exist to anchor the soil (Martin et al., 2010). As rock avalanches are highly dependent on topography, specific elevation thresholds were considered. Colorado lake locations and surrounding environment were reclassified from DEM raster data to identify only lakes that occur below tree line to control for disturbance that was not caused by soil erosion and rock avalanche activity (3352 m).
Slope
Snow avalanches occur most frequently on slopes ranging from 20to a 50 pitch (Vasskog et al., 2011). Topography of immediate local terrain surrounding the lake sites were analyzed to locate slopes prone to increased slide activity. Using a digital elevation model (DEM) of central Colorado,


19
pitch was assessed using slope and hillshade tools. The resulting slope rasters were reclassified to determine distances > 100 m from lakeshore.
Aspect
Temperature fluctuations can increase slide probability by impacting the snow pack stratigraphic layers and crystal cohesion. Southern facing slopes experience higher daily solar radiation and temperature range then other directional aspects (Schweizer et al., 1999). A DEM was used within the aspect analysis to determine slope direction of the localized terrain. Values determined from the aspect analysis (0-360) were reclassified to represent major directional headings. Southern facing slopes near lakes were identified.
Geology
Established snow avalanche paths form based on geomorphic structure and underlying bedrock composition (Butler, 1990). Gullies, couloirs, and path structures occur most often over sedimentary bedrock due to slab like physical properties (Luckman, 1977). Local geologic data was overlaid with DEM derived hill shades to determine sedimentary rock proximity to existing avalanche paths and lakes.
Accessibility and Least Cost Path
Lakes identified through the above criteria were analyzed for overall accessibility for travel and sampling gear. Colorado road and trail data was used to determine access ease for a 4x4 vehicle or trail length to carry gear in to field site (<1 km). Distances from the University of Colorado Denver was used to calculate travel times and gasoline costs.


20
Figure 2: Procedural flow chart of steps taken within the GIS suitability analysis performed to identify field sites lakes most likely to capture snow avalanche sediment deposits. Data layer information and metadata available below in Table 2.
This analysis produced five potential field sites throughout Colorado: Chicago Lakes, Cottonwood Lake, Mirror Lake, Maroon Lake, and Crystal Lake. Mirror Lake and Cottonwood Lake were selected due to existing paleoecological records from the central Colorado region and ease of accessibility from Denver, Colorado. Table 2 provides details and sources of data used within this analysis.


21
Table 3: Data summary for criteria used within Suitability analysis.
Layer Shapefile Projection Source
Elevation CO DEM UTM 13N National Map
Lakes COLakes UTM 13N Colorado Division of Water Resources
Geology COGeology UTM 13N US Geological Survey
Roads Cords UTM 13N Colorado Department of T ransportation
Trails Co_trls UTM 13N Colorado Department of T ransportation


22
CHAPTER IV SITE DESCRIPTION
Mirror (Lat. 384437.53 N, Long. 1062555.51 W, 3,347 m elevation) and Cottonwood (Lat. 384648 N, Long. 1061640, 2,911 m elevation) lakes are located within the subalpine and upper montane zones of the Sawatch Mountain Range in central Colorado (Figure 3). The Sawatch Mountains are a fault-bounded range composed predominantly of Precambrian crystalline rocks that were uplifted during the Laramide Orogeny (Refsnider and Brugger 2007; Tweto, 1987). The present physical landscape of the alpine and subalpine zones have been influenced significantly by Pleistocene glacial activity. The Sawatch Range is also characterized by the presence of rock glaciers, both active and inactive (Refsnider and Brugger, 2007).


23
A. Mirror Lake
B. Cottonwood Lake
108 W____________106 W_____________104 W____________102 W
41 N- WY
Colorado
Denver
39N- UT Mirror Lake Cottonwood Lake
1 Buena Vista
37N- 1 1
NM 0 25 50 75 100
U OU fO 1UO
108 W 106 W 104 W 102 W
Figure 3: Site Map. Location of Mirror and Cottonwood Lake in Colorado, a. Photograph taken of Mirror Lake in July 2015. b. Photograph taken of Cottonwood Lake in September 2015. c. Contour lines depicting elevation (m) and topographic features of the Mirror Lake regions with local existing avalanche tracks depicted in purple; i. represents ML131A core taken in 2013, and ii. Represents location of ML15EBC core. d. Contour lines depicting elevation (m) and topographic features of the Cottonwood Lake region with local existing avalanche tracks depicted in purples; iii. Represents the location of the CWL15 core taken in September of 2015; iv. Represents the location of an existing semi-annual debris flow path.


24
Mirror Lake is approximately 29 kilometers west of Buena Vista, within the Taylor Park watershed. The kettle lake was created from Pleistocene glaciers that sculpted the landscape leaving behind blocks of ice and a deep depression that fdled with water after the glacier retreated. The immediate eastern slopes, with angles ranging between 25 and 47 degrees, have been carved by avalanches that lead to the lakes edge. East Willow Creek flows into Mirror Lake from the south side. Local bedrock is composed of glacial drift, felsic gneiss, sandstone, and plutonic rock. Glacial drift is composed of left over glacial melt including gravel, sand, clay and erratics (Klassen and Thompson, 1993). The Taylor Park region is characterized by harsh winters with high snow fall accumulation and warm summers. Regional temperature (Crested Butte, 2715 m) averages 0.9 C with an average snowfall of 551cm per year. Local and regional topography has a profound influence on micro-climate causing distinct ecoclines and microclimates.
Cottonwood Lake is approximately 13 kilometers from Buena Vista, within the Arkansas River basin just east of Cottonwood Pass summit and located within the Collegiate Peaks Wilderness Area (San Isabel National Lorest). Steep slopes occur on both the northern and southern slopes of the lake that range between 25 and 47 degrees, and evidence exists of previous avalanches and debris flows. The lake experiences occasional mud flows on the south-east side of the shore. Cottonwood Lake is also a kettle lake within the South Cottonwood Canyon system and has an inflowing stream coming in from the western edge. Surrounding geological bedrock composition consists of granite as the dominant rock along with plutonic rock, biotite gneiss (schist) and glacial drift. Xenolithic evidence, such as rock and organic debris, are visible at the base of established avalanche paths and debris flow channels along the south side of the lake.
Regional climate varies with elevation east of Cottonwood pass. The Arkansas River Valley lies along the north-western edge of the banana belt, a microclimate that produces warmer average temperatures throughout the year (Buena Vista 6.9 C; Salida 7.4C climate information from US


25
Climate [http://www.usclimatedata.com/climate/Colorado/united-states/3175]). The average annual precipitation is 27 cm per year in Buena Vista and 61 cm per year in Crested Butte.


26
CHAPTER IV
METHODS AND DATA ANALYSIS Field Methods
Cores sampled for this study from Mirror Lake were taken in September 2013 (Del Piore, 2015) and in July of 2015. The 2013 core (ML 131A) was taken off the north-eastern shore towards the center of the lake. The ML15EBC was taken from the north-eastern central bank of Mirror Lake (18 m water depth) within proximity to an existing avalanche path location (Figure 3). In September 2015, the Cottonwood Lake core (CWL15) was sampled off the south-western shore towards the center of the lake (Figure 3). The cores were taken from an anchored floating platform with a 5-cm diameter Fivingston square-rod piston sampler (Wright et al., 1983). Approximately the top meter of sediment was captured from the lakes in 2015 in order to capture the most recent (-2000 years) geomorphic events. The core MF15EBC recovered 0.92 meters of sediment. The Cottonwood Fake core CWF15 was extracted from 7.4 m water depth and 0.98 m sediment was retrieved. The unconsolidated sediment at the top of the profile was not captured in the sample, but captured using a Klein short coring devise. The cores were measured for length and lithology described. The cores were placed in plastic wrap and PVC-pipe for protection and transported back to the laboratory and refrigerated.
Laboratory Methods
Core preparation and lithology
Each core was opened individually and lengths taken to compare against recorded field measurements. Cores were split lengthwise and lithology described in detail. Cores were subsampled at continuous one-centimeter increments and stored in 2-oz labeled Whirl-Pak bags. One-half of each core was left un-sampled and archived at the University of Colorado Denver Paleoecology, Palynology, and Climate Change Faboratory.


27
Magnetic susceptibility & loss-on-ignition
Sediment magnetic susceptibility was continuously measured on Cottonwood Lake (CWL15) and Mirror Lake (ML15EBC) cores to determine clastic inputs of iron-bearing sediments (Gedye et al., 2000). A Bartington MS2E point sensor was moved along each intact sediment core at room-temperature and at one-centimeter intervals for the length of each core. Measurements were recorded in units of centimeter-gram-second (cgs). Loss-on-ignition (LOI) was performed to assess changes in background lake sediment productivity and sediment input. LOI was measured at two centimeter intervals, and at one centimeter intervals in areas of interest. One-cm3 samples were dried for twenty-four hours at 90C to remove moisture from the samples. Percent organics were calculated based on weight-loss after heating the samples for two hours at 550C. Lastly, total percent carbonates were determined from subsequent weight-loss after heating for an additional two hours at 900 C (Dean, 1974). Magnetic susceptibility and loss-on-ignition analysis was performed on the ML 131A in 2013 using the same methods (Del Piore, 2015).
X-RAY fluorescence
X-Ray fluorescence (XRL) was measured on the CWL15 and ML15EBC cores to determine the geochemistry and elemental makeup of the lake sediment cores pulled from both lakes (Linkenbinder et al., 2014). Measurements of the LOI residual sediment were taken using a handheld Olympus Delta scanner on the residual dried sediment yielded from LOI analysis at 2-cm intervals and at higher resolutions in locations of interest. The sediment was homogenized by grinding the sediment using an agate mortar and pestle and packed into p-XRL sample holders and covered with a 4 micron ultralene window film for measurement. XRL measurements were run in both soil and geochemistry (yielding averages of three measurement) modes. Geochemistry data was consolidated and visualized using C2 software to determine trends in composition with regards to the locations of the disturbance event layers.


28
Particle Size Analysis
Particle size distributions (PSD) were determined on 1 cm3 of sediment at 2 cm intervals, and at finer 1-cm intervals for areas of disturbance layer interest. Samples were treated with hydrogen peroxide (30%) for 24 hours to oxidize organics and then deflocculated with heating treatments and distilled water at 70C as per Gray el al. 2010. Treated sediment was transferred to scintillation vials and combined with sodium hexametaphosphate for transportation. Samples were analyzed at Dr. Andrew Grays lab at the University of California Riverside Environmental Science Lab using a Beckman-Coulter (LS-13-320) Grain Size Analyzer. The resulting PSDs were used to describe the distribution of grain sizes throughout the CWL15 and ML15EBC cores. The ML131A cores areas of visual interest for particle size analysis focused around previously noted disturbances as per Del Piorie, 2015.
Data Analysis
X-Rav Fluorescence
Element concentrations derived through XRF are used to infer relationships between lake sediments and the surrounding geological environment. A ratio of rubidium (Rb) to strontium (Sr) was used to identify disturbed allochthonous material from gradual autochthonous sediments produced within the lake. A higher ratio suggests sediments that have been chemically weathered (Jin et al., 2001; Jin el al., 2006). Lower ratio values infer sediment of more detrital nature brought down by mass wasting events. Ratios were calculated as Rb/Sr. Autochthonous sediments were interpreted as consistently higher ratio values (>0.42.) Allochthonous sediments were distinguished by the presence of lower ratio values (<0.42). This threshold was determined through a combination of observations in data where decreases in ratios correspond with dark coarse grained layers within the cores and values utilized within current literature (Vasskog et al., 2011).


29
Particle Size Analysis
Statistical data processed through GRADISTAT, a program used to quickly analyze and produce grain size statistics as per Folk and Ward (1957), was used to determine grain size characteristics (in metric units) specific to autochthonous versus allochthonous sediments (Blott and Pye, 2001). Thresholds reported in current literature (Vasskog et al., 2011) were used to assess similar conditions to the Mirror and Cottonwood cores. Background autochthonous sediments were composed of higher levels of consolidated clays and silts. Autochthonous material had varied levels of sorting between cores, with mean grain sizes less than 50 pm. Allochthonous disturbed sediments were composed of coarser silts and sands and were determined as poorly sorted material with a mean grain size greater than 50 pm (Nesje et al., 2007; Vasskog et al., 2011).
Differences in geomorphic disturbances were determined within the allochthonous sediments by assessing the relationships between sorting and grain size. Avalanche events (Figure 6; Event 1) were determined as events with a mean grain size >50 pm with sorting values > 3 pm. Flooding events (Figure 6; Event 2) were determined as events with a mean grain size >50 pm with sorting values between 3 and 2.5 pm. Density currents (Figure 6; Event 3) were characterized to have a mean grain size >50 pm and sorting values <2.5 pm.
Averages of autochthonous sediment particle size distributions were taken from all three cores and plotted against specified geomorphic events 1-3 (snow avalanche, flood, density current) to determine distribution characteristics. Folk and Ward skewness and kurtosis descriptions rendered through GRADISTAT were used to describe differences within in each of the distributions. Table 3 describes particle size ranges as per Folk and Ward (1957).


30
Table 4: GRADISTAT particle size descriptions (Blott and Pye, 2001) adapted from Folk and Ward, 1957.
Sorting____________________Skewness________________________Kurtosis
Very Well Sorted <0.35 Very fine skewed +0.3 to+1.0 Very Platykurtic <0.67
Well Sorted 0.35-0.50 Fine Skewed +0.1 to+0.3 Platykurtic 0.67-0.90
Moderately well sorted 0.50-0.70 Symmetrical +0.1 to-0.1 Mesokurtic 0.90-1.11
Moderately sorted 0.70-1.00 Coarse skewed -0.1 to -0.3 Leptokurtic 1.11-1.51
Poorly Sorted 1.00-2.00 Very coarse skewed -0.3 to -1.0 Very Leptokurtic 1.50-3.00
Very Poorly Sorted 2.00-4.00 Extremely leptokurtic >3.00
Extremely poorly sorted >4.00


31
CHAPTER V RESULTS
The results of the analyses have been separated into the following sub-sections: core descriptions and analysis, magnetic susceptibility & loss-on-ignition, geochemistry/XRF, and particle size distributions. The core descriptions and analysis section describes the Cottonwood and Mirror lake sediment cores based on visual inspection and observations of major sedimentation changes and physical composition. Magnetic susceptibility of sediments and loss-on-ignition of the sediments were used to determine clastic inputs and lake productivity, respectively. The next section describes changes in geochemistry and fluctuations of elemental composition within the records. In the last section, particle size analysis and statistical characteristics are described to determine differences throughout the sediment cores.
Lithology
Core Descriptions and Analysis
Lake cores from both Cottonwood and Mirror lakes were composed primarily of fine detritus gyttja (FDG), sediment containing a combination of organic matter, inorganic precipitates, and varying mineral composition (Oregon State, Sediment in Lakes). Coarse-grained sediment, highly visible minerals (mica and quartz), macrofossils (wood and other botanical debris), and sediment greater than 1 mm were found interspersed within the FDG sediments.


32
a. Cottonwood Lake
l.iihology
l.osc-On-Ignition
Magnetic Susceptibility Geochemistry
b. Mirror Lake 15EBC
........................ i t i i i i i i i i i i i i i i i i i >
0.1 0.2 0. 04 0 5 0.6 8 II 14 17 20 21 0 8 18 24 52 40 0.0 0.2 0.4 0.6 0 0.24 0.48 0.72
Bulk Density (g/cmJ) % Organic Content % Mineral Residue (cgs) Rb/Sr-Ratio
c. Mirror Lake 131A
Figure 4: Study Core Lithologies, a. data from Cottonwood lithology, loss-on-ignition, magnetic susceptibility and geochemistry data. b. data from Mirror Lake (15EBC) lithology, loss-on-ignition, magnetic susceptibility and geochemistry data. c. data from Mirror Lake (131 A) lithology, loss-on-ignition and magnetic susceptibility.


33
Cottonwood Lake (CWL15)
The CWL 15 sediment core was composed of light brown FDG with two large sections of 2-3 cm thick inorganic sediment intermixed in coarse detritus gyttja (CDG) (Figure 4a). FDG at the bottom of the core, from 100 to 83 cm depth, transitioned to the first section of inorganic and CDG sediment layers (82 to 67 cm depth). There were visual traces of mica and macrofossils such as woody debris and pine needles within the CDG. The dark coarse layers transitioned back to a lighter FDG, from 67 to 48 cm depth, with one inorganic sediment layer at 58 to 56 cm depth. A second zone of dark inorganic layers and CDG with visible macrofossils and mica occurred from 48 to 21 cm depth. The top 20 cm of the core consisted of FDG.
Mirror Lake (ML15EBQ
The ML15EBC core was predominately composed of a light FDG sediments with frequent visible darker silt bands towards the beginning of the record between 90 to 68 cm depth. Layers of dark silt also occurred at 62 to 60.5 cm and 57.5 to 56 cm depth. A darker grey-brown layer of coarse silt occurred at 85 to 84 cm depth. Since no visibly-identifiable sediment layers existed in the top part of the core (53 to 0 cm depth) this analysis focused only on the lower half to the record (100 to 53 cm depth) (Figure 4b).
Mirror Lake ML131A
The ML 131A core was originally described in Del Piore 2015. The core was composed of dark gyttja characterized by coarser and darker gyttja layers with a high abundance of macrofossils, mica and silicates (Figure 4c). From 200.5 to 184 cm depth sediment was observed as FDG with a darker coarse layer from 187 to 185 cm depth. The ML 131A core transitioned back to FDG (184 to 104 cm depth) with darker layers occurring at 159 to 156 cm and 128 to 109 cm depth. Core composition then transitioned from FDG (104 to 6 cm depth) to dark coarser layers present at 88 to 86, 72 to 71, 69 to 67, 34 to 28 and 19 to 17 cm depth.


34
Magnetic susceptibility & loss-on-ignition
Cottonwood Lake (CWL15)
The CWL15 sediment magnetic susceptibility increased from an average of 1.2 cgs at the bottom of the core to an average of 2.1 cgs at the top of the core. At the bottom of the core, from 89 to 87 cm depth, magnetic susceptibility fluctuates from 1.34 to 0.802 cgs before increasing quickly to
2.5 cgs at 70 cm depth. The magnetic susceptibility remained low from 70 to 51 cm depth at an average of 0.65 cgs before increasing to 2.21 cgs at 47 cm depth. Between 47 to 20 cm depth, magnetic susceptibility fluctuated between 1.04 to 2.97 cgs before decreasing quickly to 0.79 cgs at 19 cm depth. From 19 cm depth to the top of the core, large fluctuations occurred between 0.302 and
3.05 cgs. Peaks in magnetic susceptibility occurred at depths of 70 cm at 2.5 cgs, 22 cm at 2.97 cgs, 12 cm at 2.58 cgs, and 7 cm at 3.05 cgs. These pulses suggest an increase in allochthonous ironbearing sediment to the lake.
Loss-on-ignition data in the CWL15 core indicates a decrease in organics from the bottom of the core (average 10.9%) to the top (average 8.8%); although even at the bottom of the core, total organic content never exceeded 25%. Undisturbed FDG sediments displayed lower levels of organic content, whereas in areas which correspond with darker coarse layers, organic content increased. Areas of FDG averaged 9.5% organics, while the coarse FDG, between 80 to 79 cm depth, ranged between 14 and 22%. Smaller increases also occurred at 72 (18.4%) and 59 cm depth (15.6%).
Mirror Lake (ML15EBC)
The ML15EBC sediment magnetic susceptibility remained consistently low throughout the core, with an average of 0.153 cgs and a slight decrease to 0.0759 cgs towards the top of the core starting at 60 cm. At the bottom of the core, the magnetic susceptibility fluctuated slightly and decreased between 100 to 91 cm depth from 0.112 to 0.0573 cgs. From 90 to 82 cm depth, there was a slight increase from 0.104 to 0.192 cgs before decreasing to 0.0653 cgs at 81 cm depth. Within the


35
middle section of the core, magnetic susceptibility increased slightly between 80 to 74 cm depth from 0.123 to 0.175 cgs, respectively, before increasing quickly to 0.594 cgs at 73 cm depth. Magnetic susceptibility at the top of the core remained consistently low with little fluctuation from 73 to 53 cm depth. The only noticeable pulse within the ML15EBC core occurred at 73 cm depth (0.594 cgs).
The loss-on-ignition data for Mirror Lake (ML15EBC) records a decrease in organics from the bottom (average 16.3%) of the core to the top (average 11.8%). The percentage of organics did not exceed 22% anywhere within the core. The uninterrupted FDG sediments displayed lower levels of organics (mean 15.4%). Areas of darker sediment layers had slightly higher levels of organic material (16.8%). At 84 cm depth, organics dropped to 7.6% indicating a pulse in inorganic material into the lake.
Mirror Lake (ML 131A)
The ML 131A core magnetic susceptibility and loss-on-ignition data was originally analyzed and completed in 2014 (Del Piore, 2015). Magnetic susceptibility increased gradually from the bottom of the core to the top from 1.09 to 2.4 cgs. From the 200 to 183 cm depth, magnetic susceptibility fluctuated slightly before rising from 1.09 to 2.434 cgs. From 184 to 143.5 cm depth, the cores magnetic susceptibility increased from 1.44 to 2.11 cgs The magnetic susceptibility decreased from 1.90 to 0.734 cgs before increasing at 119 cm depth to 1.19 cgs. Magnetic susceptibility decreased and remained relatively low until 110 cm depth at 1.18 cgs before increasing to 3.14 at 106 cm depth. Values became more variable from 105 to 60 cm before distinct sharp pulses began to occur at 59 cm depth at 3.9 cgs to 51.5 cm depth at 11.5 cgs. At the top of the core magnetic susceptibility decreased from 45 to 10 cm depth. Values varied slightly through the middle of the core 60 to 45 cm depth (1.47-5.10 cgs).
The ML 131A loss-on-ignition analysis showed a decrease in organic material from the bottom (average 15.1%) to the top of the core (average 9.4%) and did not exceed 25% anywhere


36
within the record. In areas of undisturbed FDG sediments, percentages of organic content remained relatively low (only 8.1%). The bottom of the core, (from 196 to 150 cm depth), displayed an average of 14.2% organics. Within the middle of the core FDG sediments (at 148 cm depth) showed 12.6% organics which decreased to 9.1% at 120 cm depth. Organic material remained low at the top (7.3 to 9.8%) from 20 to 10 cm depth. A layer of darker material at 19 cm depth recorded 7% organics.
Geochemistry/XRF
Cottonwood Lake (CWL15)
The CWL15 geochemistry primarily focused on the elemental composition of rubidium (Rb) and strontium (Sr) as they can be used to suggest differences in autochthonous and allochthonous materials. Rb/Sr-ratio values fluctuated throughout the record. At the bottom of the core, from 97 cm to 85 cm depth, ratio remained high fluctuating between from 0.41 and 0.43. Rb/Sr-ratio then decreased from 84 to 72 cm depth from 0.44 to 0.31. In the middle of the core (from 65 to 55 cm depth), composed primarily of FDG sediment, Rb/Sr-ratios increased to 0.5. From 47 to 23 cm depth, where darker coarse sediment layers occurred Rb/Sr-ratios decreased to 0.33 with greater variability. The top of the core (from 22 to 0 cm depth), composed primarily of FDG sediments displayed relatively higher levels of Rb/Sr-ratios (0.502).
Mirror Lake (ML15EBC)
The ML15EBC Rb/Sr-ratios remained relatively high throughout the core with some variability. The bottom of the ML15EBC core (99 to 86 cm depth) recorded Rb/Sr-ratio of 0.52 to 0.56. Rb/Sr ratios decreased after 83 cm depth to 0.29 before abruptly increasing again after 78 cm depth to an average of 0.48. The middle of the core (84 to 70 cm depth), composed of FDG, remained relatively high (0.51) from 78 to 68 cm depth before decreasing to 0.25 after 67 cm depth. The top of the core, from 65 to 53 cm depth, composed primarily of FDG, ratios remained high (0.52-


37
0.67) with little variability and experienced the highest ratios occurred at 58 (0.6) and 54 cm (0.67) depth.
Particle Size Distribution
Cottonwood Lake (CWL15)
The Cottonwood Lake core was primarily composed of sand (56.8%) and lower amounts of clay (4.2%) and silt (39.2) (Figure 5a). Increased percentages of sand occurred in inorganic layers (83 to 67, and 47 to 23 cm depth), while higher clay and silt percentages occurred in the FDG. Inorganic layers contained very poorly sorted particles (83 to 67 cm depth, and 48 to 23 cm depth) and became more sorted as the sediment transitioned back to FDG (67 to 57 cm depth) (Figure 6a). The distribution of average FDG sediments had a mesokurtic to platykurtic symmetrical grain size. Distributions taken from specific darker coarse layers displayed a more leptokurtic and a coarsely skewed distribution (see Table 3). Event 1 represents the grain-size distribution from 68 cm, event 2 at 70 cm, and event 3 at 78 cm depth. (Figure 6b).
Mirror Lake (ML15EBC)
The Mirror Lake 15EBC core was composed of silt (77%) and lower amounts of sand (13.3%) and clay (98%) (Figure 5b). Sand remained low (<10%) throughout the core, but at 84 cm depth, a sharp increase from 10 to 55% occurred. Mean grain size remained low and did not exceed 1 mm in the core. The ML15EBC core was relatively poorly-sorted but at 84 cm depth became very poorly sorted. The distribution of averaged FDG sediments displayed a mesokurtic symmetrical grain size curve (Figure 6d). The area of interest at depth 84 cm represented by event 3 (Figure 6d) became more leptokurtic with a coarsely skewed grain size distribution (Table 3) suggesting a shift from
autochthonous sediments to allochthonous.


38
Mirror Lake (ML 131A)
The ML131A core was composed of sand (average 55.6%) with slightly less silt (average 42%) and little clay (average 11%) (Figure 5c). Areas of increased sand (81.4%) and the larger grain size (166 pm) occurred in the inorganic sections, whereas silt was highest in the FDG (82.8%). Mean grain size increased throughout the record towards the top of the core where sediments displayed the largest mean sizes (230.6 pm). Grain sizes greater than 1 mm occurred five times throughout the record (27.5, 55, 65, 67, 106 cm depths). Layers (0 to 29.5 cm depth) of coarser materials were composed of poorly sorted material with larger mean grain sizes, contrasting with the more well sorted material with smaller grain size of FDG sediments. The distribution of average FDG sediments showed mesokurtic and slightly leptokurtic curves that were slightly skewed towards coarser grain sizes. Disturbances layers displayed more leptokurtic with slightly coarse skewed distributions (Figure 6f). Event 1 represents the grain-size distribution at 55 cm depth, event 2 at 27.5 cm depth, and event 3 at 17.5 cm depth. It is important to note that the ML131A core was processed for PSD as a secondary analysis of ML15EBC.


39
Figure 5: Percentage compositions of three lake cores as indicated in legend above. a.Cottonwood Lake (CWL15). b. Mirror Lake (ML15EBC). c. Mirror Lake (ML131A).


40
Cottonwood Lake (CVVL15)
Mirror Lake (ML131A)
Figure 6: Particle size analysis for the CWL15, ML15EBC, and ML131A cores, a. CWL15 Folk and Ward method bivariate plot of sorting (pm) versus mean grain size (pm), b. Plot of CWL15 average grain size (pm) distributions of background sedimentation versus grain-size distributions of differing types of events; 1 (sample 46 at depth 68 cm), 2 (sample 48 at depth 70 cm), and 3 (sample 56 at depth 78 cm), c. ML15EBC Folk and Ward method bivariate plot of sorting (pm) versus mean grain size (pm) d. Plot of ML15EBC average grain size (pm) distributions of background sedimentation versus a differing event (sample 91 at depth 84 cm), e. ML131A Folk and Ward method bivariate plot of sorting (pm) versus mean grain size (pm), f. Plot of CWL15 average grain size (pm) distributions of background sedimentation versus grain-size of differing events; event 1 (sample 106 at depth 55 cm), event 2(sample 100 at depth 27.5), and event 3 (sample 99 at depth 17.5 cm).


41
CHAPTER VI DISCUSSION
The following chapter discusses sediment deposition characteristics of snow avalanche events in high elevation lakes of central Colorado, and how they can be distinguished from autochthonous sediments and other disturbance events that deposit allochthonous sediment (e.g. flood and debris flow). Cottonwood and Mirror Lake sediment records are compared to demonstrate how these events can be separated within two separate watersheds in the Sawatch Mountains of central Colorado. In addition, two cores from Mirror Lake allow for a within-lake comparison.
Distinguishing between autochthonous from allochthonous disturbance
Allochthonous sediments deposited through instantaneous disturbance events (e.g. snow avalanche or floods, or debris flows) are often visibly identifiable among more abundant autochthonous fine detritus gyttja (FDG) and other internally redistributed allochthonous sediments through time. Allochthonous layers, deposited after a disturbance event, are often distinguished as coarser grained sediment (e.g. silts to small pebbles) with higher abundance of macrofossils brought into the lake (Nesje et al., 2007). The CWL15 and ML131A cores both contain distinct sections of disturbed sediment composed of coarser silt and sandy material with macrofossils and traces of mica not found in the FDG sediments (Figure 3a). The ML15EBC core, in contrast, is composed of almost all FDG with infrequent brown grey band and one dark brown sediment layer at 84 cm depth suggesting a core composed primarily of undisturbed autochthonous material (Figure 4b.). Therefore, the differences in sediment characteristics in the Mirror Lake cores, suggests that the disturbance events are localized and are not recorded throughout the lake.
Allochthonous sediments deposited by mass wasting events such as avalanches are known to display distinct geochemical signatures differing from FDG sediments subject to chemical weathering associated with the watershed (Vasskog et al., 2011; Croudace et al., 2006). The geochemical


42
analysis of the sediments, using x-ray fluorescence (XRF), provides a means of determining sediment provenance and parent material. For example, Rb/Sr-ratios have been used as a proxy for evaluating chemical weathering and past climates within individual watersheds (Jin et al., 2001). Rubidium (Rb) is linked to detrital clays more resistant to weathering processes typically found in autochthonous sediments, whereas Strontium (Sr), can be used to reflect activity during chemical and physical breakdown (Jin et al., 2001; Jin et al., 2006). This ratio increases significantly in the presence of weathered rock, reflecting chemical erosion, versus fresh or newer rock (Dasch, 1969). Rb/Sr-ratios are well reflected within igneous (basalt and granite) rock. The Rb/Sr-ratio can therefore be used to distinguish weathered sediments and material introduced by mass wasting events within the Mirror and Cottonwood lake records as local geology contains high concentrations of basalt and other igneous rock. Autochthonous sediments display higher values of RB/Sr-ratios due to the slower organic and clay accumulation. Mass-wasting events (i.e. snow avalanches) display lower Rb/Sr-ratios indicating an increased amount of detrital material and inorganics that are not subjected to gradual modes of weathering.
The CWL15 sediment had an average Rb/Sr-ratio of 0.43, but displayed the lowest ratios (0.34) and greater variability in areas of coarse grained sediment. At the bottom of the core between the depths of 83 to 67 cm depth, the greatest variability occurred in the dark coarse-grained sediments. Locations of lower Rb/Sr ratios, corresponding with coarse layers suggest sedimenttransported into lakerapidly and that has not sustained extensive chemical weathering. Contrastingly, highest ratiosof CWL15 of 0.57 occurred within areas of uninterrupted lighter FDG found at 59 cm depth (Figure 4a). Similarly, ratios of ~0.5 occurred at 16 cm depth within FDG sediments. ML15EBC, composed primarily of FDG had an average Rb/Sr-ratio of 0.45 as well as more consistently higher values than CWL15, with some fluctuation (Figure 4b). Although minor variability occurred within ML15EBC, the higher ratios suggest a consistent and relatively undisturbed sediments that have undergone chemical weathering. The depths at which the higher and lower values occurred within CWL15 and


43
ML 131A suggest differences between autochthonous sediments naturally broken down through time from allochthonous sediments abruptly deposited into the lake via geomorphic disturbance.
Analysis of grain size distributions (e.g. core percent composition, mean grain size, sorting, kurtosis, and skewness) are useful indicators of sediment transport and deposition mechanisms (Folk and Ward, 1957; Beierle etal., 2002). Visibly identified allochthonous sediments in core CWL15 record large grain sizes with higher percentages of sand (70.2%) (Figure 4a). Similar layers in ML131A also coincided with the highest percentages of sand (53.6%) (Figure 4c). A layer from 117 to 99 cm depth records an increased percent sand from 37% to 61% before decreasing to 33% in areas of FDG (Figure 5c). The FDG sediments record high instances of finer silts (73%) and consolidated clays (8.6%) in CWL15 between of 65 to 49 cm depth (Figure 4a). The ML15EBC core had the highest levels of medium silt (80.2 %) throughout the core, but at depth 84 cm, a sharp increase from 13% to 53% in coarse grain sand occurred, before sharply declining back to 7.7% indicating the location of a geomorphic disturbance event (Figure 4b) and suggesting core composition of autochthonous sediment with a layer of allochthonous deposited material.
Particle size distributions can also distinguish between autochthonous and allochthonous disturbances. Disturbance events quickly introduce substantial amounts of detrital material of varying sizes into the lake. Occurrences of mean grain size larger than 1 mm are not seen within lake cores in areas of FDG, and largest grain diameters were only recorded within disturbed sediments (Figure 5). Previous literature has determined that particles with D > 1 mm to be a strong indicator of snow avalanche events, but in the absence of particles below this threshold other indicators within particle size characteristics exist to be used to identify snow avalanche signals. (Nesje el al., 2007; Vasskog el al., 2011). For example, CWL15, mean grain size, exceeding 1 mm occurred 12 times in areas of allochthonous sediments (83 to 67 cm and 47 to 23 cm depth; Table 4), whereas the autochthonous sediments at the top of the CWL15 core (6 to 2 cm depth) were well sorted and did not exceed an average grain size of 41 pm (Figure 6a). The ML15EBC cores only disturbance layer (84 cm depth)


44
had a mean grain size of 49 pm (Figure 6c) and the FDG unit did not exceed a mean grain size of 29.25 (im, suggesting a core composed primarily of FDG sediments. The ML131A core recorded mean grain size exceeding 1mm four times (Table 4) with areas of visible disturbance contrasting to undisturbed FDG with a mean grain size not exceeding 48 pm (Figure 6e).
Due to rapid transport and deposition, allochthonous disturbance sediment is poorly sorted and are distinguishable from well-sorted autochthonous sediment. Highest sorting values (indicative of poorer sorting) coinciding with higher mean grain size are strong indicators of allochthonous disturbance events (Vasskog et al. 2011). Although sorting of all three cores was poor, FDG sections were not as poorly sorted as allochthonous disturbance events with medium to large grain sizes. FDG sediments within CWL15 were poorly sorted, but with only a low to medium mean grain size (13 to 47 pm). RSE at depths of 83 to 67 cm and 47 to 23 cm depths recorded both large grain sizes and poor sorting compared to autochthonous FDG sediments (Figure 6a).
The difference between particle size characteristics of autochthonous and allochthonous disturbance sediment can be visualized using bivariate plots of particle size distribution descriptors (Beierle et al., 2002). Background sediments of all three cores have symmetrical grain size distributions, with a slight skew towards coarse silts (Figure 6b,d,e, Table 3). Disturbance events from the cores are coarsely skewed (Table 3) and distinct from the autochthonous sediment curves of CWL15 and ML15EBC. The disturbance events have higher percentages of larger grains that creates a bimodal fat-tail of sand-sized particles, evident within the CWL15 and ML131A cores. The slight bimodal distribution suggests a more heterogeneous particle matrix that is introduced to the system through a quick and fluid pulse event (Spencer, 1963). Autochthonous FDG distributions of both CWL15 and ML15EBC displayed lower meso- or platykurtic (less peaked) distributions (Table 3). The meso and platykurtic (less peaked) distributions of autochthonous sediments contrast with disturbance event distributions of all three cores that are far more leptokurtic (more peaked) and skewed to the left by the presence of coarse silts and sands. The ML 131A autochthonous sediment


45
distribution is slightly more leptokurtic than the other two cores, which may be due to the fewer number of samples analyzed in proximity to visibly coarser event layers.
In summary, the visual, geochemical, and particle size analyses allow for the characterization between autochthonous-derived FDG sediments and those deposited through rapid deposition during a disturbance event. The visible presence of more detrital material and higher occurrence of mica infers rapidly mixed sediments introduced into the record in comparison to sediment which has been gradually decomposed and weathered through time. The visual differences of background sediments and disturbance layers further correspond with the geochemistry in both CWL15 and ML15EBC.
The presence of higher Rb/Sr-ratio (>0.42) suggests that background sediments have experienced a higher degree of chemical weathering, while the lower ratios (<0.42) found in disturbance layers indicates the entrainment of less chemically weathered sediments through strong physical weathering (Vasskog et al., 2011; Boggs, 2001). Allochthonous disturbance sediments contain very poorly sorted material (coarser silts and sands) with a higher mean grain-size (52-172 pm) which suggests an increased depositional energy entering the lake (Nesje et al., 2007). Furthermore, sediment that is chemically weathered through time were defined by Gaussian distributions, contrasting distinctly with than those derived from a disturbance, which were leptokurtically peaked and coarsely skewed negatively to the right due to the presence of larger grains (Spencer, 1963) (Figure 6 b,d,e.). The differences in distributions suggests allochthonous disturbance material deposited by mass-wasting events contain larger, newer, and more detrital material that has not been subjected to the degree of chemical weathering found in autochthonous FDG sediments.
The results suggest that several proxies should be used to characterize geomorphic disturbance within a lake basin. Visual inspection of core lithology is useful for quickly identifying areas of disturbance, but not all events are visible. For example, at the top of the CWL15 core, what appears to be uninterrupted autochthonous sediments (1cm depth) (Figure 4a), are disturbance events identified by their poorly sorted sediments with a higher mean grain size of 84 pm. This is also


46
evident within ML 131A where a darker FDG (5 cm depth) was found to contain poorly sorted materials with a mean grain size of 49 pm. Conversely, the ML1EBC had visible dark gray banding, which was initially interpreted to be disturbed sediments, but lacked particles with medium to high mean grain size. Similar problems arise when interpreting Rb/Sr-ratios. Current literature has noted that using elemental analyses provides the strongest separation of autochthonous and allochthonous sediments (Vasskog, et al., 2011; Jin et al., 2006); however, ML15EBC composed of FDG sediments with very low mean grain size still had fluctuations and the presence of relatively low Rb/Sr-ratios. Given all the measures conducted in this study to characterize geomorphic disturbance events, particle size analysis provides the most powerful results; however, other proxies help to verify disturbance from non-disturbance events.
The effectiveness of other proxy indicators examined in this study are influenced by local and regional characteristics such as climate and topography. For example, measurable spikes in magnetic susceptibility have been well documented within the literature to indicate pulses of iron-bearing clastic sediments entering the lake (Wondzell and King, 2003; Blake et al., 2006). This proxy is limited to watersheds with iron-bearing sediments or where fire events were frequent and intense enough to influence the magnetics of sediments. Both Mirror and Cottonwood lakes, however, record peaks during events identified through particle-size analysis and geochemistry. The geology of the watersheds did not contain iron-bearing bedrock. Loss-on-ignition, which is used to measure lake productivity, also did not decrease during a geomorphic disturbance. This is likely due to large macrofossils being brought into the lake along with the inorganic sediments.
Characterizing different allochthonous events (e.g. snow avalanches, floods, debris flows)
The grain-size properties of allochthonous disturbance layers can be used to distinguish differences in geomorphic mass-wasting events common in high elevation lakes such as: snow avalanches, floods, and density currents (Vasskog et al., 2011). Statistical differences in particle size distributions can be used to infer distinct methods of transport and depositional characteristics of


47
particles into the watershed (Tanner, 1991). Using statistics calculated in GRADISTAT software, grain-size properties were used to describe and interpret different RSE (Folk and Ward, 1957; Blott and Pye, 2001). By plotting derived grain size parameters together, signatures of three disturbance events can be identified within both Cottonwood and Mirror lake sediment cores (Figure 6). Within this sub-section, differences in disturbance event signatures are discussed.
Darker color
No Macrofossils
Well-sorted
High Grain Size
Low Rb/Sr-Values
Very Leptokurtic
Density Current
Lighter color
No Macrofossils
Well-sorted
Low Grain Size
High Rb/Sr-Values
Platykurtic
Grain Size
Avalanche Event
Background
Sediment
Darker color
Increased Macrofossils
Better-sorted
High Grain Size
Low Rb/Sr-Values
Very Leptokurtic
Flood Event
Gram Size
Darker color
Macrofossils c
Poor-sorted c
Med-High Grain Size M
Low Rb/Sr-Values
Leptokurtic
Grain Size
Figure 7: Disturbances the watershed. Summary of disturbance characteristics and transport into the lake.


48
Snow Avalanche Events
Snow avalanche events (Event 1) within Cottonwood and Mirror lakes are associated with poorly, sorted medium-to-high mean grain size, with higher concentrations of larger particles. The presence of grain sizes greater than 1 mm is a strong indicator of avalanche activity; however, this depends on the type of the slide (Vasskog et al., 2011). Both CWL15 and ML131A showed avalanche events to be the most poorly sorted of the three disturbances with the largest range of grain sizes (Figure 6a).
In both cases, instances of coarse sand exceeding 1 mm were evident; however, layers were described as coarse silts and fine sands (Table 3). Avalanche events (Figure 6b,d,e; Event 1) were identified by the presence of a fat-tail representing the higher concentrations of coarse sands. This contrasts with the other types of disturbances that while more highly leptokurtic (Figure 6b,d,e; Events 2 and 3; Table 3), do not display the same concentration or abundance of coarse sands (Figure 6b).
The poorly sorted sediment deposited from a snow avalanche is inferred from the intense mixing of debris transported down an avalanche path and deposited in a fan-shape feature at the end of the runout zone, or lake. Further variation in sorting and sediment mixing exists due to the differing types of snow avalanche activity (Hungr, 1985). For example, slush flows are associated with inconsistent early spring weather conditions, such as higher temperatures, and fluctuations are thought to transport more material due to decreasing stability of snow pack and decaying crystal layering (Schweiser et al., 2003). On the south-western edge of Cottonwood Fake, near where the sediment core was taken, the avalanche path suggests slush flow slides, based on topography and changes in vegetation, including large woody-detrital debris deposited at the lakes edge. When a snow avalanche slide encounters the waters edge, the energy of the slide causes material to fan out over the water before settling out and allowing debris to sort through the water column. The quick pulse-like nature of the snow avalanche event is evident by the poor sorting and mixed deposition of grain sizes. Such transport variability is reflected in all three cores, which are more leptokurtic and coarsely skewed than the FDG sediments. It is worth noting that as Mirror Fake is at a higher


49
elevation and experiences colder average temperatures, so avalanche material is more likely to encounter lake ice, leaving a fan of sediment on the ice that enters the record upon melt. This allows the deposition of larger particles further off shore, which is likely why the ML15EBC core lacks higher resolution in its event signals than in ML 131A.
It has been suggested that rapid precipitation events, common to the central Rocky Mountains, could cause a mud flow into the lake record, as indicated for the southern Sierra Nevada Mountains (Anderson, 1990). Would such an event leave similar signatures within the lake sediment as snow avalanches? The signals would be different because winter precipitation events are uncommon to central Colorado Rocky Mountains, with most winter precipitation occurring in the form of snowfall, contrasting with the wetter winters (snow and winter rains) that characterize the southern Sierra Nevada Mountains (Anderson, 1990). This should not be confused with sediments deposited by spring and summer thunderstorms. Such deposits would be more likely to cause hypemycal flood pulses or slope failures depositing sediments with different grain size characteristics closer to those of a flood or density current.
Flood Events
Flood events (Event 2) can be visually differentiated from other disturbance events within lake sediment records through higher minerogenic composition and presence of macrofossils. Identified flood events within the ML131A core contained high concentrations of macrofossils such as pine needles and other woody debris. Flooding events are associated with large mean grain size (very fine sand to fine sand), but are recognized by their well-sorted deposits. Flood event deposits are more leptokurtic and coarsely skewed than avalanches, but lack the characteristic bimodal fat-tail reflecting coarse grain size. This suggests a difference in transport and deposition that is characteristic of a higher influx of water and sediment to a lake that lacks the energy and velocity of a snow avalanche. For example, avalanche deposits are more poorly-sorted then flood deposits due to the higher transport energy of slide activity.


50
Density Current Events
Density currents (Event 3) are interpreted as the least frequent disturbance events identified within Cottonwood and Mirror lakes. Density currents, or turbidity currents appear to be discemable from the other geomorphic disturbances by large mean grain sizes (very fine to fine sand) that are the most well sorted of the three disturbance events (Figure 6) (Vasskog et al., 2011). Observed density currents display the most peaked leptokurtic distributions that are the most coarsely skewed to the right (Figure 6b,d,e). These characteristics reflect the responsive transport initiated by sediment failure and the fast fluid movement down a slope into and through the lake (Meiburg and Kneller, 2009). Density currents appeared to be the only disturbance event distinguishable within all three lake sediment cores. These deposits are due to slope failure and downslope transport carrying coarser eroded materials that contrast with the finer autochthonous sediments that have degraded over time (Johansen et al., 2001; Wondzell and King, 2003). Within the CWF15 and MF15EBC cores, density current events appear to have had the largest mean grain size (110-179.2 pm) (Table 4). The observed larger particles within the suspended sediment filter out into the record leaving behind layers, or turbidities that appear to be well sorted with a higher mean grain size. Flood-caused turbidities that can be recorded in watersheds such as Cottonwood and Mirror lakes can be further used to determine hydrological changes to the area (Zhang, 2014).
It is important to acknowledge that current literature accepts the identification of density currents and hyperpycnites through a variety of methods including those of this study, but with the addition of others including stratigraphic and horizon characteristics, spatial distributions, and multiple core scales (Famb and Mohrig, 2009; Zhang, 2014a; Zhang et al., 2014b). Further investigation into the signals found within the Cottonwood and Mirror cores is needed to increase confidence of the interpretations of this study.


51
Table 5: Summary of event characteristics within the CWL15, ML15EBC, and ML131A core including type, mean grain size, sorting value, and Folk and Ward descriptions.
Event Core Depth (cm) Event Type Magnetic Susceptibility % Organics Rb/Sr Ratio
1 CWL15 1 A 0.84 8.84
2 CWL15 8-10 A 0.68 9.5 0
3 CWL15 18-20 A 0.78 8.34 0.45
4 CWL15 23-28 A 2.5 10.1 0.29
5 CWL15 32-37 A 1.83 11.03 0.36
6 CWL15 39-45 F 0.92 10.3 0.19
7 CWL15 47 A 0.71 10.7
8 CWL15 68 A 0.67 12.46 0.44
9 CWL15 69 DC 0.62 8.6 0.51
10 CWL15 70 F 0.78 11.8 0.46
11 CWL15 71 DC 2.5 11.8 0.45
12 CWL15 72-74 F 1.83 13.98 0.37
13 CWL15 75 A 0.92 11.13 0.42
14 CWL15 76 F 0.71 11.75 0.42
15 CWL15 77 A 0.67 11.2 0.37
16 CWL15 78 DC 0.62 14.1 0.27
17 CWL15 79 A 0.55 22.77 0.3
18 CWL15 80-81 F 0.8 14.28 0.4
19 CWL15 83-91 A 0.78 11.5 0.43
20 CWL15 93 DC 11.7 0.32
21 CWL15 95-97 A 8.9 0.45
22 ML15EBC 84-85 DC 0.12 7.6 0.45
23 ML131A 5 A 2.52 7.65
24 ML131A 17.5 DC 1.18 16.96
25 ML131A 27.5 F 2.2 8.65


52
26 ML131A 29.5 DC 2.2
21 ML131A 32 F 5.03
28 ML131A 55 A 3.3 3.89
29 ML131A 68 F 1.5 10.31
30 ML131A 70.5 A 0.63
31 ML131A 106 A 3.14 8.95
Event abbreviations: A- Snow avalanche; F- Flood; DC- Density current
Event Event Type Mean GS (pm) Sorting (pm) Grain-Size Description Sorting Description Kurtosis Description
1 A 54.5 3.75 Very Coarse Silt Poorly Sorted Leptokurtic
2 A 54.7 3.88 Very Coarse Silt- Very Fine Sand Poorly Sorted Meso- Leptokurtic
3 A 63.9 3.7 Very Coarse Silt- Very Fine Sand Poorly Sorted Leptokurtic
4 A 58.4 3.91 Very Coarse Silt- Very Fine Sand Poorly- Very Poorly Sorted Meso- Leptokurtic
5 A 64.3 3.78 Very Coarse Silt- Very Fine Sand Poorly- Very Poorly Sorted Meso- Leptokurtic
6 F 101.7 2.75 Very Fine Sand Poorly Sorted Leptokurtic- Very Leptokurtic
7 A 60.2 3.66 Very Coarse Silt Poorly Sorted Leptokurtic
8 A 101.5 4.28 Very Fine Sand Very Poorly Sorted Very Leptokurtic
9 DC 110.1 2.48 Very Fine Poorly Very


Sand Sorted Leptokurtic
10 F 98.91 2.92 Very Fine Sand Poorly Sorted Very Leptokurtic
11 DC 129.7 2.48 Fine Sand Poorly Sorted Leptokurtic
12 F 113 2.77 Very Fine Sand Poorly Sorted Leptokurtic
13 A 84.7 3.53 Very Fine Sand Poorly Sorted Leptokurtic
14 F 138.8 2.5 Fine Sand Poorly Sorted Very Leptokurtic
15 A 89.02 3.12 Very Fine Sand Poorly Sorted Leptokurtic
16 DC 172.4 2.34 Fine Sand Poorly Sorted Leptokurtic
17 A 91.4 3.17 Very Fine Sand Poorly Sorted Leptokurtic
18 F 120.9 2.86 Very Fine Sand Poorly Sorted Very Leptokurtic
19 A 96.26 3.71 Very Fine Sand Poorly Sorted Leptokurtic
20 DC 148 2.59 Fine Sand Poorly Sorted Very Leptokurtic
21 A 72.05 3.6 Very Fine Sand Poorly- Very Poorly Sorted Meso- Leptokurtic
22 DC 49.25 3 Very Coarse Silt Poorly Sorted Leptokurtic
23 A 153.4 3.16 Fine Sand Poorly Sorted Mesokurtic
24 DC 230.6 2.49 Fine Sand Poorly Sorted Mesokurtic
25 F 225.1 2.62 Fine Sand Poorly Sorted Platykurtic
26 DC 114.9 2.36 Very Fine Poorly Leptokurtic


54
Sand Sorted
27 F 55.84 2.9 Very Coarse Silt Poorly Sorted Mesokurtic
28 A 58.3 3.71 Very Coarse Silt Poorly Sorted Leptokurtic
29 F 73.82 2.86 Very Fine Sand Poorly Sorted Very Leptokurtic
30 A 77.81 4.8 Very Fine Sand Very Poorly Sorted Mesokurtic
31 A 78.34 3.24 Very Fine Sand Poorly Sorted Leptokurtic
Event Abbreviations: A- Snow avalanche; F- Flood; DC- Density current
Localized differences within lake records
The disturbance events recorded at Cottonwood and Mirror Lake are highly localized and to reconstruct the mass-wasting history of the basin requires a network of cores. The differences between ML15EBC and ML 131A provide the best example of event detection. Both cores were taken from the same location off the north-eastern shore, with ML15EBC closer to shore to align with existing avalanche paths, and the ML 131A taken further off shore near the center of the lake (Figure 3). This change in location resulted in cores with visibly different sediment records. The ML131A core has visible detrital debris throughout the core that corresponds with layers of disturbance. The sediment analyses on the ML 131A core identified three types of disturbances; however, only one type in the ML15EBC.
The overall lack of event detection in the ML15EBC core may be the result of the closer proximity to the existing avalanche path and subsequent run out zone on the north-eastern shore.
Snow avalanches reach their highest velocity and slide energy at the bottom of the avalanche track and run-out zone, before fanning out into the lake (or over the flatter landscape) as they lose kinetic energy (Hungr, 1995). Therefore, the ML15EBC core location may have been in the zone of


55
avalanche translation rather than deposition, particularly during conditions of a frozen lake surface. Under this scenario, the energy of slide activity would be too great to leave record in ML15EBC sediments, while the location of the ML 131A core was far enough from shore to catch debris settled out into the water or on the ice. However, several types of events other than avalanches can register closer to the shore. Sediment failures, often resulting in mud debris flows and turbidities deposit local distributions that are associated with identifiable steep topography (Mulder el al., 2001). For example, in the ML15EBC core the only event was the density current at 84 cm depth, that suggests slope instability from the energy of the slide and increased mixing of deposits. In addition, on the southern bank of Cottonwood Lake (Figure 3) a well-defined semi-annual debris flow channel exists. In September of 2016, a sediment core was taken within proximity to the flow channel. When extruded and inspected back at the laboratory, the core lost its shape immediately and broke apart, suggesting a higher level of oxygen trapped within the sediments from the turbid disturbance of sediment. Larger rocks and pebbles, far exceeding the 1 and 2 mm threshold, were present throughout the core. The core physically contrasts from the CWL15 core, and those closer to avalanche tracks. The differing lithologies suggests the core taken near the debris flow was significantly disturbed and would not serve as a good core to reconstruct debris flow event frequency; however, the CWL15 core picked up traces of density currents that could have been a result of debris flow activity from the channel. Debris flows occur most often in the late spring mud season during significant rains in conjunction with snow melt. This creates fast moving surficial erosion with high velocity down steep topography. The channels derived from debris flows have deeper channels than avalanche tracks, are lined with large rocks and lack flag trees and new shrub growth associated with frequent disturbance.


56
Figure 8: Example of sediment core extracted from base of mud debris flow channel off the south-eastern bank of Cottonwood Lake. The sediment is very unconsolidated and doesn't maintain shape.
In summary, core sampling locations are an important consideration for future research and multiple cores need to be taken to reconstruct past mass-wasting event disturbance. If the ML15EBC core was the only core taken from Mirror Lake little to no activity would have been identified producing a low-resolution record. For future studies cores should be taken around the basin, from central locations, and especially near mass-wasting landscape features or incoming steams. Cores taken from the center of the lake tend to capture most events with the least amount of mixing due to transport.
Building on existing literature from Norway and future research
The lake sediment records examined within this thesis corroborate with findings in current literature (Nesje et al., 2007; Vasskog el al. 2011). Snow avalanche activity can be identified from autochthonous background sediments most effectively using particle size analysis and geochemistry. For example, Norwegian field sites have different environmental characteristics (e.g. local geology,


57
topography, or climate) than those of central Colorados Rocky Mountains. This has the potential to influence disturbance signals and thresholds within the core. For example, steeper topography could result in greater transport energy, and would carry and deposit large particle sizes.
The studies in Norway were conducted on a larger lake (Oldevatnet; area 8.03 km2); however, the signals were picked up on much smaller kettle lakes characteristic of subalpine environments of Colorado. This study suggests that geomorphic disturbances from mass-wasting events are highly localized, and can potentially be missed from cores extracted from large lakes such as Oldevatnet. The extraction of multiple cores from smaller lakes can be used to show the relationship between spatial location and proxy resolution and confidence. For example, Mirror Lake cores are distinctly different due to the proximity of avalanche paths on the north-eastern shore and resulting snow avalanche transport and debris deposit causing the ML15EBC to have low resolution, and the ML131A to more accurately pick up event signals. This is reaffirmed by the CWL15 and the September 2016 core taken at the base of the debris flow channel. Large lakes and cores taken large distances apart increase the potential for record error and decrease overall resolution for disturbance to the watershed.
Using a GIS, the study identified landscapes and lakes that had the greatest potential for capturing mass-wasting/avalanche histories. Current literature utilized field sites situated in large glacial valleys where the probability of snow avalanche activity is high. North America has several possibilities for avalanche/mass-wasting reconstruction and the use of a GIS significantly helps narrow down potential sites, but also the most accessible and cost-effective sites.
Future research would involve developing an age model for the sites so that a paleo-disturbance reconstruction can be developed. Radiocarbon dating of macrofossils found in the cores has been problematic due to the number of older macrofossils washed into the lake sediments during disturbance events. Moving forward it would be useful to date material (i.e. pollen or charcoal),


58
avoiding large macrofossils, within the FDG/autochthonous sediments. Alternatively, optically stimulated luminescence (OSL) dating could be used (Chen and Pagonis, 2011)
Once a chronology can be established, specifically for Cottonwood Lake which has no age model, the record can be compared to regional paleoclimatic records for central Colorado (Anderson 2011; Cook et al., 2004; Mann et al., 2008). Current paleoecological studies from central Colorado suggest experienced a shift from a summer dominated precipitation regime to a winter precipitation regime -2000 yr BP with less effective summer moisture (Del Piore, 2015). The increase in winter precipitation suggests increase in historic snow avalanche event frequency due to higher snowpack (Martin et al., 2001; Lazar et al., 2008; Vasskog et al., 2011). Frequency of snow avalanche events, and other geomorphic disturbances can be compared to historic climatic records to determine if trends exist regarding changes in climate. Paleoflood deposits are particularly sensitive to climate change (i.e. higher frequency of precipitation events and snow melt) (Knox, 2000). Hyperpycnal, and related density currents are linked to climate by flood frequency and impact can record changes to climate (Mulder et al., 2001). Linking flood disturbance layers to regional paleoclimate data is useful can also be useful to further interpret apparent density currents and turbidities as indicators of past hydrologic changes and activity within the two watersheds with potential climate implications (Zhang, 2014; Molnar, 2004).
Other directions for future research would be to examine the impacts of subalpine and alpine wildfire on disturbance event, such as debris flow frequency. Wildfire influences levels of local erosion by removing vegetation from slopes susceptible to failure (Wondzell and King, 2003). The absence of vegetation allows already erodible soils to be exposed to increased surface erosion. Wildfire debris flow relationships have been well documented in southern California and the western slope of Colorado, but little is known about fire debris flow relationships within central Colorado (Wells, 1987; Cannon and DeGraff, 2009; Cannon et al., 2001). Fire increases infiltration soil-slip to the landscape, creating ideal conditions for debris flows during precipitation events (Cannon et al.,


59
2001). Established debris flow channels, such as those of Cottonwood and Mirror lakes, are susceptible to overland flows and further accentuated by the influence of wildfire (Wondzell and King, 2003). The combination of sediment chronology and multiple proxies (magnetic susceptibility, loss-on-ignition, geochemistry, and particle size distributions) would be ideal to further determine this relationship in central Colorado.


60
CHAPTER VII CONCLUSION
In conclusion, the sediment cores from Cottonwood and Mirror lake offer new, high resolution data on avalanche and other mass-wasting disturbance common to high elevations of lakes of central Colorado. This data distinguishes between autochthonous and allochthonous sediments deposited by various disturbance including snow avalanche, flooding, and deposits left from density currents. The research was structured around the following three questions:
1. What characteristics distinguish autochthonous and allochthonous/disturbance sediments in a lake sediment record?
Allochthonous sediments deposited within sediment cores can often be visibly identifiable among autochthonous sediment as darker colored bands composed of coarser material (e.g. silts to small pebbles) with higher abundance of macrofossils. Although certain events are not visibly distinguishable and must be identified using other proxy data analyses (e.g. particle size distributions). Autochthonous sediment is visibly identifiable as lighter FDG composed of consolidated clays and medium silts. Allochthonous sediments deposits display geochemical trends differing from autochthonous background sedimentation. Allochthonous sediments display lower Rb/Sr-ratio values (<0.42) with higher value variability. Autochthonous sediments yield higher ratio values (>0.42) with less variability.
Allochthonous sediments show larger mean grain sizes with higher percentages of sand and coarser silts as disturbance events quickly introduce substantial amounts of detrital material into the lake in varying degrees of decomposition. Grain size >lmm can be used as a strong indicator of mass-wasting disturbance, but requires further characteristics to determine the type of geomorphic disturbance. It should also be noted that some allochthonous sediments contain grains ranging from 50-200 pm. The rapid depositional energy of disturbance results in allochthonous sediments that are


61
coarse grained and poorly sorted compared to the relatively well sorted autochthonous sediment with smaller grain sizes. Differences between autochthonous and allochthonous sediments are visualized using bivariate plots. Autochthonous sediments display a symmetrical grain size distribution with only a slight skew from coarser silts (Table 3). RSE layers are coarsely skewed (Table 4) distributions with curves have higher concentrations of larger grains that create a bimodal fat-tail of sand sized particles. Autochthonous distributions display lower meso- or platykurtic less-peaked curves (Table 3) due to smaller standard deviations like that of a Gaussian distribution curve. Disturbance event distributions are far more leptokurtic, skewed by the presence of coarser silts and sands.
The results suggest that the use of several proxies to distinguish geomorphic disturbance events. Visual inspection of core lithology can be useful but not all events are visible. Similarly, interpreting Rb/Sr-ratios can be used as an indicator of autochthonous and allochthonous material through chemical weathering; however, areas of allochthonous material show high variability in ratio values and can include higher ratios within areas of visibly identified disturbance. Disturbance events are characterized by their high mean grain size and poorly sorted sediment due to high energy and rapid transport methods. Local environmental variability can impact proxy confidence. Magnetic susceptibility is used to identify disturbance, specifically fire events, within lake sediments; however, if geologic bedrock does not contain iron-bearing material, disturbance can go undetected. Loss-on-ignition, used to infer lake productivity, did not decrease during a geomorphic disturbance event likely due to large macrofossils being brought into the lake along with the inorganic sediments.
2. What sediment properties define snow avalanches from other mass wasting events (flood, debris flows, fire, etc) in lake sediments?
Snow avalanche events are recognizable within lake sediments as poorly sorted (higher sorting values) material with a medium to high mean grain size. The wide range of mean grain size is likely


62
due to several types of snow avalanche slides and characteristics (Hungr, 1985, Table 1). Avalanche deposits have relatively low Rb/Sr-ratios that contrast with surrounding FDG sediments. Avalanche disturbances are the most poorly sorted of the three disturbances, which included floods and density currents (Figure 6a, Table 4). Distributions of snow avalanche events are identifiable by the presence of a fat-tail of higher concentrations of coarser silts and sands, and their grain-size distributions are more leptokurtic and coarsely skewed than background autochthonous sediments, differing from both floods and density currents (Table 1, Table 4).
Flood events can be visually inferred by higher minerogenic composition (mica and quartz) and increased macrofossil concentration. They are also associated with larger mean grain size (very fine sand to fine sand) (Table 3), but differ from avalanche deposits as they are better sorted (lower sorting values). Flood event distributions are more leptokurtic and coarsely skewed (Table 3, Table 4) than avalanche distributions, but lack the fat-tail of larger grain sizes (Figure 7). This suggests different transport and deposition of water pulses into a lake versus the energy and velocity of a snow avalanche depositing material at the bottom of the run-out zone.
Density currents appear to differ from avalanche events as material that have the largest mean grain size, (very fine to fine sand) (Table 3, Table 4) and are the best sorted of the three geomorphic disturbance events (Figure 7). Density currents, or turbidities display more leptokurtic distributions than avalanche events which are most coarsely skewed (Table 3, Table 4).


63
REFERENCES
Alestalo, J. (1971) Dendrochronological interpretation of geomorphic processes. Fennia 105, p 1-140
Anderson, R., Smith, S., Lynch, A., and Geils, B. (2010). The pollen record of a 20th century spruce beetle (Dendroctonus rufipennis) outbrea in a Colorado subalpine forest, USA. Forest Ecology and Management 260, p 448-455.
Anderson, R., Allen, C., Toney, J., Jass, R., and Bair, A. (2008) Holocene vegetation and fire regimes in subalpine and mixed conifer forests, southern Rocky Mountains, USA. International Journal of Wildland Fire 17, p 96-114
Aplet, G., Laven, R., and Smith, F. (1988) Patterns of community dynamics in Colorado Engelmann spruce-subalpine fir forests. Ecological. 69:2, p 312-319
Baggi, S., Schweizer, J. (2009) Characteristics of wet-snow avalanche activity: 20 years of
observations from a high alpine valley (Dischma, Switzerland). Natural Hazards. 50, p 97-108
Baker, W. (1992) The landscape ecology of large disturbances in the design and management of nature reserves. Landscape Ecology. 7:3
Barsch, D. (1977) Nature and importance of mass-wasting by rock glaciers in alpine permafrost environments. Earth Surficial Processes. 2, p 231-245
Beierle, B., Lamoureux, S., Cockbum, J., Spooner, I. (2002) A new method for visualizing particle size distributions. Paleolimnol. 27, p 279-283
Blake, W., Walbrink, P., Doerr, S., Shakesby, R., and Humphreys, G. (2006) Magnetic enhancement in wildfire-affected soil and its potential for sediment-source ascription. Earth Surface Processes and Landforms. 31, p 249-264
Boggs, S. Principles of stratigraphy and sedimentology (2002) Prentice Hall Publishing
Bocet, E., Poesen, J., and Rubio, J. (2006) Runof and soil loss under individual plants o a semi-arid Mediterranean scrubland: Inluence o plan morpholo and rainfall intensity. Earth Surface Processes and Landforms. 31, p 536-549
Brugger, K. (2006) Late Pleistocene climate inferred from the reconstruction of the Taylor River glacier complex, southern Sawatch Range, Colorado. Geomorphology. 75, p 318-329


64
Brun, E., Martin, E., Simon, V., Gendre, C., Coleou, C. (1989) An Energy and Mass Model of Snow Cover Suitable for Operational Avalanche Forecasting. Journal of Glaciology. 35:121, p 333-342
Bui, E., Mazullo J., Wilding L. (1990) Using quartz grain size and shape analysis to distinguish between Aeolian and fluvial deposits in the Dallol Boss of Niger (West Africa). Earth Surface Processes andLandforms. 14, p 157-166
Butler, D., Walsh, S. (1990) Lithologic, Structural, and Topographic Influences on Snow-Avalanche Path Location, Eastern Glacier National Park, Montana. Annals of the Association of American Geographers. 80:3, p 362-378
Cannon, S., Kirkham, R., and Parise, M. Wildfire-related debris-flow initiation processes, Storm King Mountain, Colorado. Geomorphology. 39, p 171-188
Cannon, S., and DeGraff, J. The increasing wildfire and Post-fire Debris-Flow threat in western USA, and implications of consequences of climate change. Landslides- Disaster Risk Reduction.
P 177-190
Carrara, P. (1979) The determination of snow avalanche frequency through tree-ring analysis and historical records at Ophir, Colorado. The Geological Society of America. 90:8, p 773-780
Chen, R., and Pagonis, V. (2011) Optically Stimulated Luminescense: A Simulation Approach. John Wiley & Sons, Ltd, Chichester, UK.
Connell, J. (1978) Diversity in Tropical Rain Forests and Coral Reefs. Science. 199 (4335) p 1302-
1310
Cook, E., Woodhouse, C., Eakin, M., Meko, D., and Stahle, D. (2004). Long-term aridity changes in the western United States. Science. 306, p 1015-1018.
Corona, C., Saez, J., Stoffel, M., Rovera, G. (2012) Seven centuries of avalanche activity at Echalp (Queyras massif, southern French Alps) as inferred from tree rings. The Holocene. 23:2, p
292-304
Croudace, I., Rindby, A., Rothwell, G. (2006) ITRAX: description and evalutation of a new multifunction X-Ray core scanner. The Geological Society 267, p 51-63
Dasch, E. (1969) Strontium isotopes in weathering profiles, deep-sea sediments, and sedimentary rocks. Geochemica Cosmachimica Acta. 33 p 1521-1552


65
Dean, W. (1974) Determination of carbonate and organic matter in calcareous sediments by loss on ignition comparison to other methods. Journal of Sedimentary Petrology. 44, p 242-248
Del Piore (2015) 4,000 Years of Environmental Change in Central Colorado: A Paleoecological Perspective. University of Colorado Denver, p 1-117.
Doesken, N., Pielke, R., and Bliss, O. (2003). Climate of Colorado. Climatography of the United States 60 http://climate .atmos.colostate.edu/climateofcolorado .php
Fall, P. (1997a) Fire history and composition of the subalpine forest of western Colorado during the Holocene. Journal of Biogeography. 24 p 309-325
Finkenbinder, M., Abbott, M., Edwards, M., Fangdon, C., Steinman, B., Finney, B. (2014) A 31,000-year record of paleoenvironmental and lake-level change from Harding lake, Alaska, USA. Quaternary Science Reviews. 87, p 98-113
Friedman, G. and Sanders, J. (1978) Principles of Sedimentology. Wiley, New York
Garrow, H. (1982) A theoretical model of source and transport effects on grain-size distributions.
Eleventh International Congress on Sedimentology. p 80
Gedye, S., Jones, R., Tinner, W., Ammann, B. and Oldfield, F. (2000) The use of mineral magnetism in the reconstruction of fire history: a case study from Fago di Origlio, Swiss Alps.
Paleogeography, Paleoclimatology, Palaecology. 164, p 101-110
Gray, A., Pasternack, G., and Watson, E. (2010) Hydrogen peroxide treatement effects on the particle size distribution of alluvial and marsh sediments. The Holocene 20:2
Gregory, K. and McIntosh, W. Paleoclimate and paleoelevation of the Oligocnene Pitch-Pinnacle flora, Sawatch Range, Colorado. Geological Society of America Bulletin. 108:5 p 545-561
Higuera, P., Briles, C., and Whitlock, C. (2014) Fire-regime complacency and sensitivity to
centennial-through millennial-scale climate change in Rocky Mountain subalpine forests, Colorado, USA. Journal of Ecology. 102, p 1429-1441
Hungr, O. (1995) A Model of the runout analysis of rapid flow slides, debris flows, and avalanches
Canadian Geotechnical Journal. 32, p 610-623
Jin, Z., Cao, J., Wu, J. and Wang, S. (2006) A Rb/Sr record of catchment weathering response to Holocene climate change in Inner Mongolia. Earth Surface Processes and Landforms.


66
Johansen, M., Hakonson, T., Breshears, D. (2001) Post-fire runoff and erosion from rainfall
simulation: contrasting forests with shrublands and grasslands. Hydrological Processes 15, p 2953-2965
Krumbein, W. (1938) Size frequency distribution of sediments and normal phi curve. Journal of
Sedimentary Petrology 8, p 84-90
Kluth, C. and Coney, P. (1981) Plate tectonics of Ancestral Rocky Mountains. Geology
Knox, J. (2000) Sensitivity of modem and Holocene floods to climate change. Quaternary Science Reviews. 19, p 439-457
LaChapelle, E. IV. 1. Principles of Avalanche Forecasting. Associate Committee on Geotechnical Research
Lamb, M. and Mohrig, D. (2009) Do hyperpycnal-flow deposits record river-flood dynamics? Geological Society of America. 37:12 p 1067-10170
Lowe, J., and Walker, M. (1997) Reconstructing Quaternary Environments. Longman, London. 2nd Ed. 446
Lubchenco, J. Plant Species Diversity in a Marine Intertidal Community: Importance of Herbivore
Food Preference and Algal Competitive Abilities. The American Naturalist. 112:983, p 23-29
Luckman, B. (1977) The Geomorphic Activity of Snow Avalanches. Geografiska Annaler. 59:1, p p 31-48
Mann, M., Zhang, A., Hughes, M., Bradley, R., Miller, S., and Rutherford, S. (2008). Proxy-based reconstructions of hemispheric and global surface temperature variations over the past two millennia. Proceedings of the National Academy of Sciences. 105, p 13252-13257.
Mast, J., Veblen, T., and Linhart, Y. (1998) Disturbance and climatic influences on age structure of Ponderosa Pine at the Pine/Grassland Ecotone, Colorado Front Range. Journal of Biogeography. 25:4, p 743-755
Martin C., Pohl, M., Alewell, C., Komer, C., Rixen, C. (2010) Inerrill erosion a isturbed alpine sies: Effects of plant functional diversity and vegetation cover. Basic and Applied Ecology 11, p
619-626
McLaren, P., and Bowles, D. (1985) The effects of sediment transport on grain-size distributions.
Journal of Sedimentary Petrology 55:4 p 457-470


67
Meiburg, E. and Kneller, B. (2009) Turbidity Currents and Their Deposits. Annual Review Fluid Mechanics. 42, p 135-156
Meusburger, K., Konz, N., Schaub, M., Alewell, C. (2010) Soil erosion modelled with USLE and PERSERA using QuickBird derived vegetation parameters in an alpine catchment.
International Journal of Applied Earth Observation and Geoinformation 12, p 208-215
Molnar, P. (2004) Late Cenozoic Increase in Accumulation Rates of Terrestrial Sediment: How Might Climate Change Have Affected Erosion Rates? Annual Review Earth Planet Science 32. P. 67-89
Mulder, T., Migeon, S., Savoye, B., Faugeres, J. (2001) Inversley graded turbidite sequences in the deep Meditteranean: a record of deposits from flood-generated turbidity currents? Geo-Marine Letters. 21:2 p 86-93
Paine, R., Tegner, M., and Johnson, E. (1998) Compounded Perturbations Yield Ecological Surprises.
Ecosystems. P 535-545
Pazzaglia, F. and Kelley, S. (1998) Large-scale geomorphology and fission-track thermochronology in topographic and exhumation reconstructions of the Southern Rocky Mountains. Rocky Mountain Geology. 33:2, p 229-257
Refsnider, K., and Brugger, K. (2007) Rock Glaciers in Central Colorado, U.S.A., as Indicators of Holocene Climate Change. Arctic, Antarctic, and Alpine Research. 39:1, p 127-136
Ryan, S., Bishop, E., and Daniels, M. (2014) Influence of large wood on channel morphology and sediment storage in headwater mountain streams, Fraser Experimental Forest, Colorado. Geomorphology 217:15 p 73-88
Schillereff, D., Chiverrell, R., Macdonald, N., Hooke, J. (2014) Flood stratigraphies in lake sediments: A review. Earth Science Reviews. P 1-63
Schoennagel, T., Veblen, T., Romme, W., Sibold, J., and Cook, E. (2005) ENSO and PDO variability affect drought-induced fire occurrence in Rocky Mountain subalpine forests. http://spot.colorado.edu/~schoenna/images/Schoennagel2005EcoApps.pdf
Schweizer, J. (1999) Review of dry snow slab avalanche release. Cold Regions Science and Technology. 30:1-3, p 43-57


68
Sherriff, R., Veblen, T., and Sibold, J. (2001) Fire History in high elevation subalpine forests in the Colorado Front Ranee. Ecoscience 8, p 369-380
Stohlgren, T., Owen, A., and Lee, M. (2000) Monitoring shifts in plant diversity in response to climate change: a method for landscapes. Biodiversity and Conservation 9, p 65-86
Szeic, J., Zeeb, B., Bennett, K., and Smol, J. (1998) High-resolution paleoecological analysis of recent disturbance in a southern Chilean Nothofagus forest. 20, p 235-252
Tweto, O. (1975) Laramide (Late Cretaceous-Early Tertiary) Orogeny in the Southern Rocky Mountains. Geological Society of America Memoirs. 144, pi-44
Tweto, O. (1987) Rock units of the Precambrian basement in Colorado. U.S. Geological Survey, professional paper. 1321
Vasskog, K., Nesje, A., Nage, E. (2011) A Holocene record of snow avalanche and flood activity reconstructed from a lacustrine sedientary sequence in Oldevatnet, western Norway. The Holocene
Vasskog, K., Paasche, O., Nesje, A., Boyle, J., Birks, H. (2012) A new approach for reconstructing glacier variability based on lake sediments recording input from more than one glacier. The Holocene. P 192-204
Vivoni, E.; Bowman, R., Wyckoff, R., Jakubowski, R., Richards, K. (2006) Analysis of a monsoon flood event in an ephemeral tributary and its downstream hydrologic effects. Water Resources Research. 42:3
Watson, E., Pasternack, G., Gray, A., Goni, M. and Woolfolk, A. (2013) Particle size characterization of historic sediment deposition from a closed estuarine lagoon, Central California. Estuarine, Coastal and Shelf Science. 126, p 23-33
Wells, W. The effects of fire on the generation of debris flows in southern California. Geological Society of America. 7, p 105-114
Weltje, G., von Eynatten, H (2004) Quanitative provenance analysis of sediments: review and outlook. Sedimentary Geology 171, p 1-11
White, P., and Pickett, S. (1985) Natural disturances and patch dynamics: an introduction White Academic Press, p 3-13


Whitlock, C., Higuera, P., McWethy, D., and Briles, C. (2010) Paleoecological perspectives on fire ecology: revisiting the fire-regimes concept. The Open Ecology Journal. 3, p 6-23
69
Wilhelm, B., Amaud, F., Sabatier, P. (2013) Paleoflood activity and climate change over the last 1400 years recorded by lake sediments in the north-west European Alps. Journal of Quaternary Science, p 189-199
Wondzell, S., and King, J. (2003) Postfire erosional processes in the Pacific Northwest and Rocky Mountain regions. Forest Ecology and Management 178, p 75-87
Zhang, X. (2014a) Lacustrine Turbidites from Tropical African Lakes as Indicators of Hydrologic and Climatic Changes. Syracuse University Dissertation- ALL p 182
Zhang, X., Scholz, C., Hecky, R., Wood, D., Zal, H., and Ebinger, C. (2014) Climatic control of the late Quaternary turbidite sedimentology of Lake Kivu, East Africa: Implications for deep mixing and geologic hazards. Geology 42:2p811-814


70
APPENDIX A
MAGNETIC SUSCEPTIBILITY FOR COTTONWOOD LAKE (CWL15)


Depth (cm) MS (cgs)
1 2.04E-01
2 2.11E+00
3 2.17E+00
4 1.04E+00
5 1.80E+00
6 2.48E+00
7 4.42E-01
8 3.05E+00
9 2.56E+00
10 3.02E-01
11 2.56E+00
12 2.58E+00
13 2.67E+00
14 1.47E+00
15 1.38E+00
16 1.17E+00
17 1.36E+00
18 1.14E+00
19 7.89E-01
20 1.08E+00
21 1.98E+00
22 2.97E+00
23 1.42E+00
24 2.23E+00
25 1.67E+00
26 1.91E+00
27 1.76E+00


28 1.31E+00
29 2.20E+00
30 1.57E+00
31 2.33E+00
32 1.93E+00
33 1.12E+00
34 1.34E+00
35 1.04E+00
36 1.33E+00
37 1.59E+00
38 1.27E+00
39 1.45+00
40 1.07E+00
41 1.35E+00
42 1.57E+00
43 2.21E+00
44 1.61E+00
45 1.60E+00
46 1.20E+00
47 3.77E-01
48 1.20E+00
49 8.37E-01
50 1.05E+00
51 7.75E-01
52 5.40E-01
53 5.82E-01
54 6.12E-01
55 5.06E-01
56 5.44E-01


57 5.23E-01
58 8.18E-01
59 4.88E-01
60 4.08E-01
61 5.42E-01
62 3.95E-01
63 6.67E-01
64 1.08E+00
65 8.35E-01
66 6.17E-01
67 7.44E-01
68 8.37E-01
69 6.80E-01
70 7.78E-01
71 2.50E+00
72 2.50E+00
73 1.53E+00
74 1.47E+00
75 9.15E-01
76 7.02E-01
77 6.65E-01
78 6.20E-01
79 5.40E-01
80 4.81E-01
81 1.11E+00
82 8.57E-01
83 4.02E-01
84 7.52E-01
85 9.62E-01


86 1.01E+00
87 7.94E-01
88 9.68E-01
89 1.33E+00


75
APPENDIX B
MAGNETIC SUSCEPTIBILITY FOR MIRROR LAKE (MLL15EBC)


Depth MS
(cm) (cgs)
53 1.17E-01
54 1.30E-01
55 1.37E-01
56 1.77E-01
57 1.45E-01
58 1.53E-01
59 1.35E-01
60 7.59E-02
61 1.20E-01
62 1.39E-01
63 1.71E-01
64 1.69E-01
65 2.38E-01
66 1.45E-01
67 2.05E-01
68 1.71E-01
69 1.82E-01
70 1.11E-01
71 2.56E-01
72 2.68E-01
73 5.94E-01
74 1.75E01
75 1.95E-01
76 1.42E-01
77 1.74E-01
78 2.13E-01
79 2.22E-01


77
80 1.23E-01
81 6.53E-02
82 1.83E-01
83 1.81E-01
84 9.96E-02
85 1.51E-01
86 1.66E-01
87 1.97E-01
88 1.32E-01
89 1.01E-01
90 4.63E-02
91 5.73E-02
92 1.69E-01
93 1.47E-01
94 2.06E-01
95 1.07E-01
96 9.53E-02
97 1.20E-01


78
APPENDIX C
MAGNETIC SUSCEPTIBILITY FOR
MIRROR LAKE (ML131A)


Depth (cm) MS (cgs)
6.5 1.82E+00
7 2.86E+00
7.5 1.19E+00
8 1.06E+00
8.5 3.18E+00
9 3.41E+00
9.5 1.48E+00
10 6.14E-01
10.5 1.63E+00
11 2.52E+00
11.5 2.02E+00
12 2.10E+00
12.5 2.62E+00
13 2.48E+00
13.5 2.36E+00
14 2.18E+00
14.5 2.51E+00
15 1.54E+00
15.5 1.83E+00
16 3.44E+00
16.5 4.13E+00
17 4.57E+00
17.5 2.63E+00
18 1.97E+00
18.5 1.82E+00
19 3.00E+00
19.5 2.59E+00


20 2.05E+00
20.5 2.95E+00
21 1.08E+00
21.5 1.18E+00
22 1.66E+00
22.5 2.15E+00
23 1.74E+00
23.5 1.14E+00
24 1.27E+00
24.5 3.45E+00
25 4.59E+00
25.5 3.83E+00
26 1.92E+00
26.5 1.75E+00
27 3.09E+00
27.5 1.25E+00
28 1.08E+00
28.5 1.99E+00
29 2.51E+00
29.5 2.10E+00
30 2.05E+00
30.5 2.03E+00
31 8.06E-01
31.5 1.09E+00
32 1.24E+00
32.5 2.02E+00
33 2.31E+00
33.5 1.66E+00
34 3.13E+00


34.5 1.33E+00
35 2.19E+00
35.5 2.90E+00
36 3.72E+00
36.5 2.24E+00
37 1.66E+00
37.5 2.49E+00
38 2.87E+00
38.5 4.26E+00
39 4.08E+00
39.5 3.62E+00
40 2.88E+00
40.5 4.60E+00
41 4.95E+00
41.5 3.21E+00
42 2.31E+00
42.5 2.58E+00
43 2.56E+00
43.5 1.40E+00
44 9.28E-01
44.5 2.09E+00
45 1.82E+00
45.5 3.63E+00
46 1.47E+00
46.5 2.04E+00
47 3.39E+00
47.5 3.43E+00
48 4.95E+00
48.5 5.10E+00


49 3.28E+00
49.5 2.18E+00
50 2.52E+00
50.5 3.38E+00
51 3.73E+00
51.5 1.08E+00
52 8.42E+00
52.5 3.83E+00
53 4.54E+00
53.5 3.36E+00
54 1.79E+00
54.5 3.02E+00
55 3.28E+00
55.5 3.84E+00
56 4.30E+00
56.5 3.27E+00
57 3.34E+00
57.5 3.07E+00
58 2.87E+00
58.5 3.14E+00
59 3.89E+00
60 2.04E+00
60.5 9.31E-01
61 3.38E+00
61.5 5.00E+00
62 3.20E+00
62.5 1.73E+00
63 1.63E+00
63.5 2.32E+00


64 9.11E-01
64.5 2.68E+00
65 4.30E+00
65.5 7.00E-01
66 6.93E-01
66.5 1.23E+00
67 1.76E+00
67.5 1.31E+00
68 1.40E+00
68.5 2.40E+00
69 2.03E+00
69.5 2.10E+00
70 1.48E+00
70.5 1.37E+00
71 1.31E+00
71.5 1.51E+00
72 1.15E+00
72.5 1.82E+00
73 1.24E+00
73.5 1.19E+00
74 2.18E+00
74.5 2.72E+00
75 4.10E+00
75.5 4.09E+00
76 2.79E+00
76.5 2.26E+00
77 1.42E+00
77.5 1.68E+00
78 1.47E+00


78.5 1.38E+00
79 2.37E+00
79.5 1.37E+00
80 1.55E+00
80.5 1.30E+00
81 9.22E-01
81.5 5.70E-01
82 5.44E-01
82.5 5.75E-01
83 7.65E-01
83.5 2.13E+00
84 2.22E+00
84.5 2.86E+00
85 2.58E+00
85.5 2.26E+00
86 1.26E+00
86.5 1.61E+00
87 2.49E+00
87.5 2.46E+00
88 2.38E+00
88.5 2.71E+00
89 1.99E+00
89.5 2.10E+00
90 1.66E+00
90.5 2.75E+00
91 1.91E+00
91.5 1.15E+00
92 2.21E+00
92.5 2.03E+00


93 3.63E+00
93.5 3.32E+00
94 2.46E+00
94.5 2.21E+00
95 2.74E+00
95.5 2.50E+00
96 2.10E+00
96.5 2.27E+00
97 1.40E+00
97.5 2.28E+00
98 1.95E+00
98.5 2.06E+00
99 1.01E+00
99.5 1.05E+00
100 2.09E+00
100.5 3.14E+00
101 1.98E+00
101.5 1.81E+00
102 1.61E+00
102.5 1.80E+00
103 1.56E+00
103.5 2.05E+00
104 1.85E+00
104.5 2.88E+00
105 3.53E+00
105.5 3.56E+00
106 3.34E+00
106.5 1.71E+00
107 2.08E+00


107.5 2.55E+00
108 2.63E+00
108.5 1.96E+00
109 9.43E-01
109.5 1.31E+00
110 1.01E+00
110.5 1.79E+00
111 1.49E+00
111.5 2.51E+00
112 1.92E+00
112.5 1.67E+00
113 2.33E+00
113.5 1.91E+00
114 1.91E+00
114.5 1.86E+00
115 1.88E+00
115.5 1.98E+00
116 2.95E+00
116.5 2.25E+00
117 7.77E-01
117.5 1.03E+00
118 2.11E+00
118.5 1.78E+00
119 1.81E+00
119.5 1.72E+00
120 1.21E+00
120.5 6.71E-01
121 1.99E+00
121.5 2.08E+00


122 2.94E+00
122.5 2.50E+00
123 2.64E+00
123.5 2.12E+00
124 2.41E+00
124.5 1.18E+00
125 1.17E+00
125.5 1.81E+00
126 1.95E+00
126.5 1.54E+00
127 1.87E+00
127.5 1.25E+00
128 1.62E+00
128.5 1.42E+00
129 8.81E-01
129.5 1.21E+00
130 1.38E+00
130.5 7.81E-01
131 1.15E+00
131.5 1.22E+00
132 7.34E-01
132.5 1.12E+00
133 1.19E+00
133.5 1.62E+00
134 1.74E+00
134.5 1.59E+00
135 1.27E+00
135.5 1.31E+00
136 9.80E-01


136.5 1.01E+00
137 1.06E+00
137.5 1.03E+00
138 1.63E+00
138.5 1.80E+00
139 2.27E+00
139.5 2.93E+00
140 1.99E+00
140.5 1.24E+00
141 1.93E+00
141.5 1.06E+00
142 1.90E+00
142.5 1.16E+00
143 2.14E+00
143.5 2.11E+00
144 2.02E+00
144.5 2.90E+00
145 2.11E+00
145.5 2.02E+00
146 2.90E+00
146.5 2.47E+00
147 2.30E+00
147.5 3.14E+00
148 2.42E+00
148.5 1.67E+00
149 3.97E+00
149.5 1.80E+00
150 9.43E-01
150.5 1.31E+00


151 1.58E+00
151.5 3.04E+00
152 2.75E+00
152.5 2.95E+00
153 2.63E+00
153.5 3.09E+00
154 2.76E+00
154.5 2.94E+00
155 3.13E+00
155.5 2.90E+00
156 3.14E+00
156.5 4.02E+00
157 3.81E+00
157.5 2.33E+00
158 2.20E+00
158.5 2.38E+00
159 2.14E+00
159.5 8.52E-01
160 7.37E-01
160.5 8.04E-01
161 1.18E+00
161.5 2.39E+00
162 1.99E+00
162.5 2.12E+00
163 2.10E+00
163.5 2.85E+00
164 2.29E+00
164.5 1.56E+00
165 1.84E+00


165.5 2.79E+00
166 2.23E+00
166.5 1.62E+00
167 1.30E+00
167.5 2.23E+00
168 2.00E+00
168.5 2.43E+00
169 1.74E+00
169.5 1.39E+00
170 1.92E+00
170.5 2.44E+00
171 1.77E+00
171.5 2.09E+00
172 2.21E+00
172.5 2.23E+00
173 2.39E+00
173.5 2.27E+00
174 2.56E+00
174.5 1.86E+00
175 1.78E+00
175.5 2.69E+00
176 2.53E+00
176.5 1.91E+00
177 1.70E+00
177.5 2.14E+00
178 1.72E+00
178.5 8.99E-01
179 9.13E-01
179.5 6.52E-01


180 8.04E-01
180.5 2.14E+00
181 2.09E+00
181.5 1.94E+00
182 7.71E-01
182.5 1.35E+00
183 2.36E+00
183.5 1.09E+00
184 1.14E+00
184.5 1.69E+00
185 1.82E+00
185.5 2.52E+00
186 1.88E+00
186.5 1.58E+00
187 2.23E+00
187.5 1.95E+00
188 2.52E+00
188.5 2.69E+00
189 2.55E+00
189.5 2.98E+00
190 3.17E+00
190.5 3.92E+00
191 3.13E+00
191.5 3.00E+00
192 3.51E+00
192.5 5.02E+00
193 3.38E+00
193.5 1.59E+00
194 2.14E+00


Full Text

PAGE 1

CHARACTERIZING SNOW AVALANCHES RECORDED IN LAKE SEDIMENTS OF CENTRAL COLORADO by ZARA KATHLEEN HICKMAN B.S., State U niversity of New York College at Oneonta, 2010 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment Of the requirements for the degree of Master of Science Environmental Sciences Program 2017

PAGE 2

2 This thesis for the Master of Science degree by Zara Kathleen Hickman Has been approved for the Environmental Sciences Program by Christy E. Briles, Chair Andrew Gray Daniel Liptzin July 29, 2017

PAGE 3

3 Hickman, Zara Kathleen (M.S., Environmental Sciences) Characterizing snow avalanches recorded in lake sediments of central Colorado Thesis directed by Assistant Professor Christy E Briles ABSTRACT Snow avalanches are common to high elevations of the Colorado Rocky Mountains; yet, little has been done to reconstruct them in the past. This thesis examines sediment characteristics of g eomorphic events recorded in high alpine lakes of central Colorado. Sediment grain size analysis, geochemistry, and magnetic sediment properties were used to characterize snow avalanches from other allochthonous disturbances such as floods and density cur rents from two sites, Cottonwood and Mirror lakes in the Sawatch Mountain Range. While data from the combined proxies provided the best evidence of different disturbances, grain size analysis was the single best indicator. Snow a valanche deposits had poorly sorted sediments with a medium to high grain size and higher frequency of larger particles (coarse silts and very fine sands). Avalanche deposits were leptokurtic nt deposits were found to be better sorted with a high mean grain size (fine to very fine sands) and a more leptokurtic be the most sorted of the thre e disturbance events, with the highest mean grain size (sands). Finally, events are not recorded across the lake basin, which is based on the location of the disturbance event (i.e. avalanche path or stream mouth), the amount of energy when it energy the lake, and the dissipation of that energy. Localized geomorphic disturbances have left distinct signatures within lake sediments that can be used to reconstruct geomorphic disturbance history.

PAGE 4

4 DEDICATION I would like to dedicate this thesis to my family, biological and accumulated. Their overwhelming suppor t and encouragement was integral to my education and success. I want to also thank my parents for always encouraging me to follow my dreams wherever t hey may take me. Through them I discovered my love of mountains and the curiosity to explore them.

PAGE 5

5 ACKNOWLEDGEMENTS I would like to take the opportunity to thank my advisor, Christy Briles, for her time, patience, g uidance, and passion Her dedication to teaching and advancing paleoecology is an inspiration. I would also like to thank my committee, Andrew Gray and Daniel Liptz i n Lastly thank you to the faculty and administration of the Geography and Environmental Sciences department of the University of Colorado Denver for their support.

PAGE 6

6 TABLE OF CONTENTS C HAPTER I. II. BACKGROUND Physical environment Geomorphology. Climate ... Vegetation Disturbance Proxy data and natural archive 15 III. Accessibility an IV. V. 26 Laboratory methods Core prepa Magnetic susceptibility & loss on ............. 27 27 Particle size analysis Data analysis VI. Lithology Magnetic susceptibility & loss on Particle size analysis

PAGE 7

7 VII. Distinguishing between autochthonous from allochtho Characterizing different allochthonous events (e.g. snow avalanches, floods, debris Building on existing literature from Norw VIII. 70 A. MAGNETIC SUSCEPTIBILITY FOR COTTONWOOD LAKE (CWL15) 70 B. MAGNETIC 5 C. MAGNETIC SUSCEPTIBILITY FOR MIRROR LAKE (ML131A) 8 D. LOSS ON 3 E. LOSS ON 6 F. LOSS ON IGNITION 9 G. 4 H. 7 I. 10 J. 4 K. 7

PAGE 8

8 LIST OF TABLES TABLE 1. Summary of avalanche manifestations 12 2. 3. GIS suitability analysis data summary 21 4. GRADISTAT grain size statistical summary 30 5. Disturbance event summary for Cottonwood and Mirror Lake 51

PAGE 9

9 LIST OF FIGURES FIGURE 1. Identification of avalanche path zones 13 2. Procedural flow chart for suitability analysis 20 3. Site map for Cottonwood Lake and Mirror Lake 23 4. Core lithology, loss on ignition, magnetic susceptibility, and Rb/Sr ratios 32 5. Core particle compositions 39 6. Particle Size Analysis Summary 40 7. Geomorphic Disturbance Summary 47

PAGE 10

1 C HAPTER I INTRODUCTION Snow avalanche events are large and quickly displaced masses of snow transported down steep slope s (Luckmann 1977) and common at high elevations in mountainous terrain. They pose a significant threat to winter sport enthusiasts, tourism, and transportation. Where snow avalanches encounter boundaries of society they can be costly to infrastructure and tourism ( e.g. ski resort management). In Colorado alone, seven people on average lose their lives due to snow avalanches each year (Colorado Avalanche Information Center 2016). They also impact ecological systems directly by removing vegetation, such as topp l ing down forests in their path (Bebi et al ., 2009; Baker 1992). By extension avalanche paths can define growth patterns of vegetation on a hillside (e.g. shrubs versus conifers), and ultimately determine what plant co mmunities can exist in an area (Baker 1992; Veblen 1994). It can be difficult to quantify snow avalanche activity and frequency (Pain et al ., 1998). Besides recorded events around residential, recreational areas, and roads (Atkins 2006), another way used to examine them is by identifying and dating tree rings that reflect scaring or differential growth patterns resulting from an avalanche event (Bebi et al ., 2009 ; Pederson et al ., 2006; Szychowska Krapiec and Krapiec 2001). Within central Colorado, the region of focus for this thesis, st udies using dendrochronologic al tree ring data have been conducted to expand historic frequency of snow avalanche events in more remote landscapes (Simonsin 2012; Veblen 1994). However, the tree scar method can be problematic if an area experiences mult iple avalanches, if the event kills many or all the trees resulting in a low sample pool, and if the tree species involved does not respond or produce quality regrowth scars (Reardon et al ., 2008; Veblen 1994; Corona 2012; Shroder 1978; Bebi et al ., 2009).

PAGE 11

2 Lake sediments are an under used archive for reconstructing avalanche events. Lakes, unlike trees, often accumulate undisturbed sediment for thousands of years, and can continuously record events that integrate signals from an entire watershed. This sediment is deposited overtime by organisms living in the lake, or allochthonous sediment from flood or debris flow (Vasskog et al ., 2011). Local geom orphic activity (floods, slope failure, snow and rock avalanches, etc.) and other environmental variables (wind blown erosion) influence depositional transport of sediments to the watershed. For example: snow a valanches deposit allochthonous course grained inorganic sediments and plant remai ns that are significantly different from year to year produ ction of organic fine grained autochthonous sediments Lake sediments are commonly used to reconstruct historical ecological patterns and ch anges, using proxies such as pollen and charcoal, but often lakes with layers of disturbed allochthonous sedi ment with low organics, ai r pockets, and sometimes large m acrofossils, are overlooked and discarded as the material disrupts the gradual sedimentation rate of autochthonous organic sediments ( Briles et al 2012; Del Priore, 2015 ) In this thesis, using a multi proxy approach, I demonstrate how d isturbed allochthonous sediment presents an opportunity to explore material that may be evidence of past disturbance event s. Until now, the only lake sediment records that have used sediment properties to reconstruct avalanche events are those from Norway (e.g. Nesje et al ., 2007; Vasskog et al ., 2011; Vasskog et al ., 2012 ; Corona et al., 2013). Grain size distribution is one of the most funda mental properties of lake sedimentary deposits ( Folk and Ward, 1974 ). The distribution of grain size within a lake deposit is influenced by parent material, as well as transport and deposition processes (Beierle et al ., 2002 ) Particle size distribution of lake sediments as a paleoenvironmental proxy has been used to help reconstruct disturbance and origin through particle transport and to distinguish between transport processes (e.g. snow avalanches and flooding events) ( Nesje et al ., 2007; Vasskog et al ., 2011 ; Jin 2005 ). Through multiple proxies (i.e. magnetic susceptibility, loss on ignition, and geochemistry) and the interpretation of grain size

PAGE 12

3 distributions using stat i sti cal tools, the type of mass wasting events can be determined (e.g. snow avalanche versus flooding event) (Vasskog et al ., 2011; Wilhelm et al ., 2013). Given that lake sediments have yet to be used outside of Norway to reconstruct snow avalanche history, the main objectiv e of this thesis research is to determine the differences between autochthonous sediments and allochthonous instantaneous disturbances in the southern Rocky Mountains (Sawatch Mount ain s) of Colorado. Proxy data was used to distinguish methods of sedimen t transport and deposition (e.g. snow avalanches versus flood event), allowing for the characterization of snow avalanches from other mass wasting events. Proxies used in this study include d : magnetic susceptibility t o measure iron bearing clastics; loss on ignition t o measure lake productivity; x ray fluorescence to determine sediment geochemistr y; and particle size analysis to describe sediment grain size distributions A Geographic Information System ( GIS ) was used to determine suitable lakes bas ed on slope, aspect, parent material, and location of visible avalanche shoots int o the lake. The study addressed the following questions: 1. What characteristics distinguish autochthonous and all ochthonous/disturbance events in a lake sediment record? 2. What sediment properties differentiate snow avalanches from other geomorphic disturbance (flood, debris flows, etc. ) in lake sediments? The thesis is developed around seven present day physical environment, ge omorphology, geology, climate, vegetation, and dominant disturbance regimes. It also discuss es the nature and uses of proxy data for reconstructing past environments. Chapter three describes the study area and characteristics of Cottonwood Lake and Mirror Lake. Chapter four details field and laboratory methods and how the lake sediment, particle size data, and geochemistry and were analyzed. Chapter five presents results of the study, including lithologies of the cores, particle size distributions, magnetic susceptibility, loss on ignition, and geochemistry data. Chapter six discusses how the proxy data can be used to characterize avalanche

PAGE 13

4 signals from other geomorphic disturbance events that result in an interruption of autochthonous sediments into the basin. Chapter seven summarizes the finding s of the study.

PAGE 14

5 C HAPTER II BACKGROUND INFORMATION Physical environment The Sawatch Mountains of central Colorado, are part of the southern Rocky Mountain chai n. The Rocky Mountains stretch south from New Mexico north into Alaska, within western North America. The Colorado Rocky Mountains begin just west of the Front Range highland (16 00 m) and extend across the state to the western most boundaries. They included over 300 peaks rising above 4,267 meters (m), and the second tallest mountain in the continental United States ( Mount El bert, 4,401 m elevation ). The Colorado Rockies are cha racterized by high peaked ranges, with deep cut valleys or parks d ispersed throughout the ranges (Pazzaglia and Kelley, 1998 ). The Continental Divide is the primary hydrological drainage barrier within the Rocky Mountains and separates distinct drainages between the Atlantic Ocean (east), Gulf of Mexico (southeast) and Pacific Ocean (west). Geology The ancestral Rocky Mountains first formed by processes of rapid uplifts from the North American and South American plates beginning approximately 300 million years ago (mya) during the Pennsylvanian Period throughout the western United States and Canada (Kluth and Coney, 1981). The Southern Rocky Mountains (north central New Mexico north through Colorado and into southeastern Wyoming) experienced a secondary i ntense geologic uplift during the Laramide orogeny (Late Cretaceous, approximately 70 mountains. The uplift of the Sawatch Range created a separation between the Taylor Park and South Park regions and are an anticline (Tweto, 1975; and Brugger, 2006). Geomorphology Central Colorado experienced two Pleistocene glacial pe riods with the latest ending approximately 18,000 years ago. Due to the regions high elevation, it experienced both rock and ice

PAGE 15

6 glaciers (Refsnider and Brugger 2007). The Sawatch Range was characterized by glacial activity that significantly influenced the formation of valleys and cirque basins of the surrounding landscape (Barsch 1977 ). The transient significant topographic relief and many high elevation kettle lakes (Brugger 2006). Prominent Pleistocene moraines exist in both Taylor Park and the upper Arkansas River valley (Refsnider and Brugger 2007). Gla cial extent of the Sawatch Range occurred from 16.3 +/ 1.6 to 22.2 +/ 2.8 ka based on exposure ages on boulders (Brugger 2006, Refsnider and Brugger 2007). These last glacial maximums ( LGM ) advances correspond with others calculated throughout the South ern Rocky Mountains ( Fall, 1997; Brugger, 2010 ) Climate Colorado climates are influenced by regional geography, specifically by elevation and topography, resulting in microclimates. influenced by normal m aritime precipitation regimes. Summers at lower elevations are typically hot and dry while in the mountains they are often cool and wet due to mountain thunderstorms. Winters are cool and dry at lower elevations while cold with high snowfall at higher elevations in the mountains. Overall, Colorado climate is characterized as cool and dry with low humidity (semi arid) Regional t emper ature averages range between 6.9 C (Buena Vista 2427 m ; east of the Continental Divide ) and 0.9 C (Crested Butte 2715 m ; west of the Continental Divide ). Current average rain precipitation ranges from 2 7 cm per year (at Buena Vista) and 61 cm at Crested Butte (climatic data from Weste rn Regional Climate Center [http://www.wrcc.dri.edu/summary/climsmco.html]). Much of precipitation primarily snow fall, factors in to local snow pack equivalence that r esults in snow melt run off in the spring and summer months. Average snowfall in the region ranges between 104cm (at Buena Vista) and 551 cm per year at Crested Butte ( units in snow water equivalents, precipitation averages from US Climate Data [http://ww w.usclimatedata.com/climate/Colorado/united

PAGE 16

7 states/3175]). Westerly prevailing winds are prominent, but are strongly determined by local topography. Due to steep and sharp relief, climate ecotones exist with rising elevation ( Mast et al ., 1998 ). Air masses from the Pacific Ocean in winter months bring snow to the west of the Continental Divide, but this rarely impacts precipitation patterns east of the Continental Divide ( Doesken et al ., 2003 ). The western slope, including the study area, receives hi gher accumulations of snow in the winter, than the Front Range due to orographic lifting and the rain shadow that result on the leeward side of the Rockies (Fall 1997a). Vegetation Local and regional topography plays a significant factor in flora distrib ution and biomass due to the elevation ecoclines (thermoclines) within the Southern Rocky Mountains ( Stohlgren 2000 ). I ncreased el evation thermoclines result in higher levels of precipitation and lower overall temperatures. Vegetation in the Rocky Mounta ins is divided into four zones that correspond to rising elevation conditions, these include: shrub steppe and grassland, (2,300 2,900 m); montane (2,800 3,000 m); subalpine (2,900 3,400 m); the alpine (> 3,500 m ) (Fall 1992b). Montane forests are compos ed of drought resistant trees such as lodgepole pine ( Pinus contoria Pseudotsuga menziesii and Pinus ponderosa ) Subalpine forests are dominated by Engelmann Spruce ( Picea engelmannii ) and subalpine fir ( Abies lasiocarpa ) forest. The alpine zone presents a significant ecotone associated with tree line that limits tree growth to under 4 meters (Aplet et al ., 1988). At high elevations, the presence or absence of vegetation can impact disturbance activity Due to the more extreme climatic and top ographic conditions of higher elevations a higher probability for soil erosion exists (Meusburger et al ., 2010). Higher concentrations of v egetation biomass was found to decrease rates of soil erosio n caused by disturbance events in alpine environments ( Martin e t al ., 2010). Vegetation intercepts rainfall and increases evapo transportation which then reduces runoff flow and channel creation a positive feedback that leads to increased

PAGE 17

8 disturbance event impacts such as landslides and mud debris flows ( Bochet et al ., 2006 ) Areas above the tree line ecotone with lo wer percent vegetative cover coincide with increased rock sli de activity and soil erosion (Meusburger et al ., 20 10 ). Disturbance An e nvironmental disturbance is a discrete event that alters a physical system and has profound and immediate impacts to the surrounding landscap e with varying long term (decadal to centennial scale) effects (White and Pickett 1985; Baker 1992). Although the overall health of the system may not experience a negative drastic change by each individual event, there are both discrete and cumulative impacts to regional geology and ecosystems Community ecological health, such as species richness is found to increase in the presence of intermediate disturbance (Connell 1978; Lubchenco 1978 ) Due to the size and infrequent nature of large mass wasting events few studies exist that encompass the entirety of ecological responses due to lack of experimental control (Michener and Haeuber 1998). There are several types of disturbance events which occur in central Colorado. Disturbances to this region can range from small and semi annual to larg e and infrequent having varying impacts due to the event type, scale, and frequency (Paine et al ., 1998). Fire is considered the most frequent and dominant disturbance event in the centra l Colorado region during warm, dry summer months (Anderson, 2008a ). (June to September) and winter seasons (Vivoni et al ., 2006) Such events can alter watershed flow dynam ics, as well as cause the triggering of disturbance from surrounding topographic relief. Fire events are common disturbances of central Colorado that can shift montane to subalpine forest vegetation structure on varying spatial scales and influencing for est s from decades to centuries (Whitlock et al., 2010). Severe fire events occur approximately 100 to 300+ years in the subalpine forest of central Colorado (Fall, 1997). Much of what is known about historic fire regime has been recovered through proxies such as charcoal in lake sediments and tree ring fire scars Climate is a

PAGE 18

9 dominant factor influencing fires, particularly during ENSO cycles, which increase the impacts of an already dry arid environment (Higuera et al ., 2014; Schoennagel et al., 2005; and Sherriff et al ., 2001). The presence of potential fuel (biomass) located throughout the subalpine forest and canopy influence overall fire ignition and spread. During the winter months when precipitation is dominated by snow fall, disturbance can occur in the form of snow avalanches. The semi arid environment of Colorado result s in a dry snow pack that is not as cohesive or binding as a maritime regional snow pack ( LaChapelle 1969 ). Avalanche slides can be unpredictable, but typically slide on topographic slopes with angles from 2 5 to 50 and when : (1) stress to existing snow slabs increases such as with significant prec ipitation accumulation; (2) strength of an existing slab decreases such as above freezing temperatures causing the deterioration of existing internal snow layers; and (3) propagation of the initial failure, or a positive feedback loop once the i nitial slide occurs (Voight et al ., 2011 ). Other factors correlated s directional aspect, wind patterns, underly ing geological composition, intensity of solar radiation, local vegetation, and the local average temperature fluctuation. Flood disturbance events occur infrequently due to extreme localized summer precipitation events (Wilhelm et al ., 2013). These disturbance events influence watershed dynamics such as tributary and channel structure and rate of sediment plumes (Schillereff 2014). Watersheds experience annual hydraulic pulses du e to precip itation and spring snow melt from higher elevations but large flooding events are rare. Due to irregular frequencies, little is known about large scale flooding (Michener and Haeuber 1998). The re is an increased frequency of flood events tha t correlate with warming climates (e.g. Medieval Climate Anomaly 800 1300 A.D.) (Wilhelm et al ., 2013). With f lood frequency and intensity increasing because of warming climates (Knox 2000) it is important to derive information from historic flood events and resulting ecological response Lake sediments are an archive of flood disturbance when they have an inflowing stream Flood events

PAGE 19

10 produce a sediment plume within a lake composed of clastic material and organic matter that leave behind coarse grained laminated layers (L owe and Walker, 1997; Schillereff et al ., 2014). Besides event frequency, studies on flood deposits in lake sediments have also been able to determine severity ( Lowe and Walker, 1997). Rock landslides are also localized infreq uent disturbance events common to the high elevations of central Colorado. Rock slides and debris flows although bound by similar factors of snow avalanches, can vary in characteristics due to the composition and causation (e.g. slope, geological composit ion, and climate). Rockslides are composed of rock mixed w ith earth, snow, or ice (Voight et al ., 2011 ) They occur more often in higher elevation areas with slopes ranging from 25 40 above tree line where vegetation significantly decreases and the landscape is composed of exposed loose soil, cobbl es, and boulders (Hungr, 1995 ). Debris flows have longer displacement paths and can be fluid and flow like in nature due to sa turation (Hungr 1995). Mud flows occur after intense rain storms where runoff loosens debris on steep terrain. Debris flows are a major cause of slope erosion and scour the landscape with deep channel like tracks (Berti et al ., 1999). Density currents, or turbidity curre nts occur due to failing slope sediments above or within a body of water and/or turbid plumes of water introduced to the internal slope that move quickly downhill releasing a mass suspension of sediment into the water column (G ould, 1951; Mortimer, 1971). Density currents which are layers of sediment rich in water that enter less dense non turbid water s and flow in layers beneath the surface (Lamb and Mohrig, 2009). Density currents are often caused by debris flows due to slope instability quickly introducing sediments into the lake (Zhang et al ., 2014b) Mountain ( Dendroctonus ponderosae ) and spruce beetle ( Dendroctonus rufipennis ) outbreaks are well documented and are becoming an increasingly more prominent issue for forest management. The pine beetle is native to the Rocky Mountains, lives and breeds primarily within ponderosa, lodge

PAGE 20

11 pole, scotch and limber pine tree species. D. ponderosae is showing exponential increases in comparison to its historic population due to changing climatic factors (Carroll et al 2003; Chapman et al ., 2012). Open forest resulting from extensive beetle kill decrease snow stability leading to increased avalanche potential under ideal weather conditions. After the trees root systems decay, the outbreaks can also lead to slope instability and increased sediment inputs to streams and waterbodies (Ryan et al., 2014). Snow Avalanche Causation and the Colorado His toric Record On February 2, 2016, an avalanche occurred west of Buena Vista on Cottonwood Pass that resulted in one fatality A storm began on January 31 and ended February 1 with approximately 15 inches of snow fall. Temperatures were consistently low throughout the storm and fluctuated between 20 to 15 C un ti l February 2. Wind speed ranged between 16 24 km/h that could have wind loaded steeper slopes. The unintentional release ( u An classification) occurred on a south easte rn facing slope with a 36 slope by a recreational snow biker. As the recreationalist ascended to steeper terrain, the start zone of the avalanche was triggered from above burying the victim. Due to the cold temperatures and added stress from the new snow fall, snow crystal line p atterns did not coalesce, resulting in a slide (Baker 1986). Avalanche events are caused by complex mechanisms that vary due to climactic conditions which make them difficult to predict. (Schweizer et al., 2003). However, it is known that a valanche formation is most strongly controlled by the interaction between the regional topography, local climactic conditions, and existing snowpack. Add ed stresses that can release snow slabs result from quick surficial loading in proximity to the start zone (Figure 1 ) progressive loading to the start zone due to precipitation and random releases (Schweizer et al 2003). The triggers of slide activity are uncertain and are di vided into two groups : dynamic and static (Luckman 1977). Static factors include slope, topographic angle and aspect, and geologic composition. Dynamic variables are weather related, such as temperature fluctuation and wind loading (Butler 1986). Due to difference

PAGE 21

12 in meteorological changes throughout the season there are several classifications of avalanche manifestations: mass movement, slush flows, and creeps (Schweizer et al. 2003). Table 1 provides a description of the avalanche types and their triggers. Table 1 : Summary of primary avalanche m anifestati ons, descriptions, and triggers Avalanche Type Description Trigger Variables Mass movement (dry snow) A dry, intercontinental snow pack. Release begins underneath a strong cohesive slab (Schweizer 1999). Snow stratification is rate dependent, meaning overall stress must exceed slab layer strength (Narita 1983). Wind loading, precipitation, random release. Slush flow (wet snow) Occur in intercontinental mountain ranges in the spring when temperatures begin to rise consistently (Bagg i and Schweizer 2008) and snowpack is saturated with runoff water. Primary slide manifestation in ranges characterized by mariti me snow pack (ex. Sierra Nevada ). Water cycle, topography, snowpack stratification, air temperature, rain, bedrock. Creep Slow deformation of snowpack (Mathews and Mackay 1963) Topography, temperature fluctuation

PAGE 22

13 Figur e 1 : Identification of avalanche path zones represented by northern slope of the lake. (Photograph credit: Zara Hickman) Several key terms are used in de scribing avalanche movement and path. The starting and run out zones are important for defining the spatial nature of a slide event. The starting zone is where the slide first begins to slip. This is primarily due to the inconsistencies in snow crystal formations, but can be a reaction to wind loading or precipitation (Hungr 1995). As the movement travels down the track, the final resting place of the mass disperses throughout the run out zone. The run out zone is w here most of the debris and material collected by the event is displaced and deposited. Snow avalanches have profound impacts to local ecosystems and landscape that are immediately felt (Pain et al ., 1998; Baker 1992). Although the overall health of a system is not drastically changed by each disturbance event, there are impacts to regional geomorphology and vegetation. Avalanche paths can redefine growth patterns on a hillside, and ultimately determine

PAGE 23

14 what ve getation can exist in an area. Such key events can create natural patches and variability as well as overall health in surrounding environments and ecosystems (Baker 1992; Veblen, 1994). lanche records exist in observational accounts (Atkins 2006). However, t ree ring data has been used to establish historic frequency of events and path history ( Ale stalo, 1971 ). Although results of such studies have produced evidence f or events this method can only establ ish minimal frequency at best (Simonsin 2012; Veblen 1994). Dendrochronological studies from Glacier National Park (GLAC) focused around existing avalanche paths and determined a lack of event frequency resolution in times of known recent activity (Butl er et al ., 1979). Since remote areas lack infrastructure, avalanche accounts are lacking. In Colorado, tree ring dating methods have been conducted in Ophir, but a study like that of GLAC has yet to be conducted to determine resolution reliability (Carre ra, 1979). Proxy data from natural archives Proxy data preserved in natural archives (such as lake sediments, ice cores, tree rings, et c) allow for the re construction of past ecosystems, environmental change including disturbances to the landscape on a range of temporal and spatial scales. Paleoenvironmental and paleoclimatic data can be gathered from, but not limited to, lake and ocean sediment cores, ice cores, tree ring records, c orals, and packrat maddens Not all proxy data can provide the same level of resolution or accuracy. Tree ring records and corals provide annual data records of temperature and precipitation; however, extent is limited to centennial to millennial scales in the case of tree rings (Andersen et al ., 2008 ). L ake an d ocean sediments, and the materia l they preserve, such as pollen and charcoal, lack resolution but can extend back millennia. Lake sediments also preserve data that can be used to reconstruct depositional history, regional climate variability and local i zed geomorphic disturbance (Rothwell et al ., 2006). But uniting multiple threads of evidence into a comprehensive analytical framework, a multi proxy approach can provide a more robust analysis with higher overall resolution of ecosystem processes and cha nge than achieved through one alone

PAGE 24

15 Sediment stratigraphy and lithological properties provide information on the occurrence and impacts of individual disturbance events within proximity to the lake field site (Vasskog et al., 2011; Nesje, 2007). Magnetic susceptibility can be used to identify allochthonous pulses from potential erosional events (Gedye et al. 2000). Loss on ignition data examines lake productivity and allochthonous sediment deposition. Geochemistry and elemental ratios are used to determine depositional history and chemical erosion relationships with local geology (Croudace et al ., 2006 ; Dasch, 1969; Jin et al 2006 ). Particle size analysis is used to determine deviation from normal distributions and to characterize transport and depositional processes through several different parameters, such as mean grain size, sorting, skewness, and kurtosis (Folk and Ward 1957 ; Blott and Pye, 2001 ). Particle Size Analysis Grain size characteristics are an elemental part of pa rticle size analysis Paleoecological studies have utilized particle size characteristics to understand geomorphological changes to the landscape and resulting watershed impacts (Folk and Ward, 1957; Weltje and von Eynatten, 2004) P article size distribution s (PSD) along with natura l proxy data can provide insight to sediment origin, sediment transportation process and depositional characteristics (Folk and Ward, 1957; Bui et al ., 1990 ; Friedman and Sanders, 1978 ). Environmental factors s uch as geology climatic conditions, regional weather, vegetation, and geomorphology influence the chemical and physical weathering regimes that control the grain size of eroded sediments ( Jin et al ., 2001; Weltje and von Evanatten, 2004). Fluvia l sediment sizes are further influence d by transport and depositional processes such as water fluvial flows a e olian, or due to topography within the channelized system (Watson et al ., 2013). The differences in specific particle size characteristics can be used to interpret transport methods as varying mechanisms (e.g. sorting, skewness, kurtosis ) transport different particle sizes

PAGE 25

16 reflecting different movement (Garrow, 1982; and McLaren and Bowles, 1985). For example, r ecent studies have utilize d particle size analysis to understand changes to historic estuary dynamics (Watson et al ., 2013). PSD has also been used to understand the role climate plays on flood events, resulting hyperpycnal density currents and other geologic hazards (Zhang, 2014a; Zhang et al ., 2014b). Particle size distributions were used to show that most turbidities were caused by extreme floods, rather than slope failure suggesting climate change implications ( Zhang, 2014a ; Zhang et al ., 2014b). Increasingly, particle size analyses have been used to understand increased erosional processes due to wildfire (Wondzell and King, 2003). The understanding of these processes and impact s to deposits is an important tool in unde rstanding often infrequent geomorphic disturbance (snow avalanches, rock avalanches, floods, and density currents ) and their landscape impacts ( McLaren and Bowles, 1985 ) Estimation of particle size distribution can be accomplished through a variety of methods including, but are not limited to, direct measurements, sieving, and laser granulometry (Blot and Pye, 2001). Regardless of method, th ese techniques are used to describe the variation of partic le diameter ( D ) within a population of particles. V arious statistical packages and statistics ai d in the interpretation of particle size characteristics by visualizing distribution curves. Statistical results are presented in two prominent units: phi transformations, and metric units. Phi units, have been termed from the metric scale ( D in millimeters ) using the following equation ( ) (Folk and Ward, 1957; Blott and Pye; 2001). Plotting multiple partic le size characteristics can often reveal relationships and trends that otherwise would not have been seen. Common variables include mean particle size, standard d eviations sorting, skewness and kurtosis ( Krumbein and Pettijohn, 1938; Folk and Ward, 1957). Mean particle size (graphical mean, ) is the average diameter of grain size (Table 2); where m is equal to the mid point of each class interval for both metric ( m m ) and phi unit ( m ). Sediments are then broken down into texture classes by fraction ( clays, silts, and sands ) (Blot and Pye, 2001; Table 3). Standard deviations ( or sorting, determine

PAGE 26

17 the variance of a define d population assuming normal conditions (Table 2) (Blot and Pye, 2001). Sorting then refers to the spread of the particle size distribution, reflecting how mixed the sediment is from well sorted (low values) to poorly sorted (high values). Skewness and kurtosis are used to describe distributions as then differ f rom normal distributions (e.g. Gaussian Distribution). Skewness ) is further used to determine shifts of distributions are fine skewed or coarse skewed (Blott and Pye, 2001). Whereas k urtosis ( ) refers to grains relative to average and ho w distribution is (Table 2) (Blott and Pye, 2001). Although phi units have been prefer red traditionally by sedimentolo gists, metric units are used more commonly by the geological community as resulting parameters are easier to interpret (Blott and Pye, 2001). This study utilized metric units for this study to most easily compare r esults to pervious literature investigating historic geomorphic disturbance (Nesje et al. 2007; Vasskog et al. 2011). Table 2 : Statistical formulae for the particle size parameter calculations (Krumbein and Pettijohn, 1938; Folk and Ward, 1957; Blot and Pye, 2001). Phi Unit Graphical Mean Standard Deviation Skewness Kurtosis

PAGE 27

18 CHAPTER IV GIS SUITABILITY ANALYSIS OF LAKES PRONE TO AVALANCHE ACTIVITY S ite selection of Colorado lakes more prone to local snow avalanches was determined through a suitability analysis using a GIS with ESRI ArcGIS 10.2 software. Five cr iteria elevation, slope, aspect, geology, and accessibility were used based on physical environment and regional geomorphology literature that highly influence avalanche triggers and events (Hungr 1995). The criteria and workflow used within suitability analysis are discussed below (Figure 2) Elevation Snow avalanches and rock slides can both produce disturbance events Although rock avalanches can transport ice and snow, these events are composed primarily of large rocks and debris. Snow avalanches are a mass movement of snow that can transport and mix of slope material and debris ( Table 1). Rock avalanches, also common to central Colorado, occur mo st frequently in tundra zones where trees larger shrubs and ground vegetation d o not exist to anchor the soil ( Martin et al ., 2010 ). As rock avalanches are highly dependent on topography, specific elevation thresholds were considered. Colorado l ake locations and surrounding environment were reclassified from DEM raster data to identify only lakes that occur below tree line to control for disturbance that was not caused by soil erosion and rock avalanche activity ( 3352 m). Slope Snow avalanches o ccur most frequently on slopes ranging from 20to a 50 pitch ( Vasskog et al ., 2011 ) Topography of immediate local terrain surrounding the lake site s were analyzed to locate slopes prone to increased slide activity. Using a digital elevation model (DEM) of central Colorado,

PAGE 28

19 pitch was assess ed using slope and hillshade tools. The resulting slope raste rs were recla ssified to determine distances Aspect Temperature fluctuations can increase slide probability by impact ing the snow pack stratigraphic layers and cry stal cohesion. Southern facing slopes experience higher daily solar radiation and temperature range then other directional aspect s (Schweizer et al ., 1999). A DEM was used within the aspect analysis to determ ine slope direction of the localized terrain. Values determined from the aspect ana lysis (0 360) were reclassified to represent major directional he adings. Southern facing slopes near lakes were identified. Geology Established s now avalanche paths form based on geomorphic structure and underlying bedrock composition (Butler, 1990) Gullies, couloirs, and path structures occur most often over sedimentary bedrock due to slab like physical properties (Luckman, 1977). Local geologic data was over laid with DEM derived hill shades to determine sedimentary rock proximity to existin g avalan che paths and lakes Accessibility and Least Cost Path Lakes identified through the above criteria were analyzed for overall accessibility for travel and sampling gear Colorado road and trail dat a was used to determine access ease for a 4x4 vehicle or trail length to carry gear in t o field site (<1 km). D istances from the University of Colorado Denver was used to calculate travel times and gasoline costs.

PAGE 29

20 Figure 2 : Procedural flow chart of steps taken within the GIS suitability analysis performed to identify field sites lakes most likely to capture snow avalanche sediment deposits. Data layer information and metadata available below in Table 2. This analysis produced five potential field sites throughout Colorado : Chicago Lakes, Cottonwood Lake, Mi rror Lake, Maroon Lake, and Crystal Lake Mirror Lake and Cottonwood Lake were selected due to existing paleoecological records from the central Colorado region and ease of accessibility from Denver, Colorado. Table 2 provides details and sources of data used within this analysis.

PAGE 30

21 Table 3 : Data summary for criteria used within Suitability analysis. Layer Shapefile Projection Source E levation CO DEM UTM 13N National Map Lakes CO_L akes UTM 13N Colorado Division of Water Resources Geology CO_G eology UTM 13N US Geological Survey Roads Co_r ds UTM 13N Colorado Department of Transportation Trails Co_trls UTM 13N Colorado Department of Transportation

PAGE 31

22 C HAPTER IV SITE DESCRIPTION Mirror (Lat. 38 37.53 55.51 W, 3,347 m ele vation) and Cottonwood 48 40 2,91 1 m elevation) lakes are located within the subalpine and upper montane zones of the Sawatch Mountain Range in central Colorado (Figure 3) The Sawatch Mountains are a fault bo unded range composed predominantly of Precambrian crystalline rocks that were uplifted during the Laramide Orog eny (Refsnider and Brugger 2007; Tweto 1987). The present physical landscape of th e alpine and subalpine zones have been influen ced significantly by Pleistocene glacial activity. The Sawatch Range is also characterized by the presence of rock glaciers, both active and inactive (Refsnider and Brugger 2007).

PAGE 32

23 Figure 3 : Site Map. Location of Mirror and Cottonwood Lake in Colorado. a Photograph taken of Mirror Lake in July 2015. b Photograph taken of Cottonwood Lake in September 2015 c Contour lines depicting elevation (m) and topographic features of the Mirror Lake region s with local existing avalanche tracks depicted in purple ; i. represents ML131A core taken in 2013, and ii. Represents location of ML15EBC core. d. Contour lines depicting elevation (m) and topographic features of the Cottonwood Lake region with local existing avalanche tracks depicted in purples; iii. Represents the location of the CWL15 core taken in September of 2015; iv. Represents the location of a n existing semi annual debris flow path

PAGE 33

24 Mirror Lake is approximately 29 kilom eters west of Buena Vista, within the Taylor Park watershed. The kettle lake was created from Pleisto cene glaciers that sculpted the landscape leaving behind bl ocks of ice and a deep depression that filled with water after the glacier retreated The immediate eastern slopes, with angles ranging between 25 and 47 degrees, have been carved by East Willow Creek flows into Mirror Lake from the south side. Local bedrock is composed of glacial drift, felsic gneiss, sand stone, and plutonic rock Glacial drift is composed of left over glacial melt including gravel, sand, clay and erratic s ( Klassen and Thompson, 1993 ). The Taylor Park region is characterized by harsh winters with high snow fall accumulation and warm summers. Regional temperature (Crested Butte, 2715 m) averages 0.9 C with an average snowfall of 551cm per year. Local and r egional topography has a profound influence on micro climate causing dis tinct ecoclines and microclimates Cottonwood Lake is approximately 13 kilometers from Buena Vista, within the Arkansas River basin just east of Cottonwood Pass summit and located within the Collegiate Peaks Wilderness Area (San Isabel National Forest). S teep s lopes occur on both the northern and southern slopes of the lake that range between 25 and 47 degrees and evidence exists of previous avalanches and debris flows. The lake experiences occasional mud flows on the south east sid e of the shore. Cottonwood Lake is a lso a kettle lake within the South Cottonwood Canyon system and has an inflowing stream coming in from the western edge. Surrounding geological bedrock composition consists of granite as the dominant rock along with pl utonic rock, biotite gneiss (schist) and glacial drift Xenolithic evidence such as rock and organic debris are visible at the base of established avalanche paths and debris flow channels along the south side of the lake. Regional climate varies with elevation east of Cottonwood pass. The Arkansas River Valley lies along the north western edge of the banana belt, a microclimate that produces warmer average temperatures throughout the year (Buena Vista 6.9 C; Salida 7.4C climate informati on from US

PAGE 34

25 Climate [http://www.usclimatedata.com/climate/Colorado/united states/3175]). The average annual precipitation is 27 cm per year in Buena Vista and 61 cm per year in Crested Butte.

PAGE 35

26 CHAPTER IV METHODS AND DATA ANALYSIS Field Methods Cores sampled for this study from Mirror Lake were taken in September 2013 (Del Piore 2015) and in July of 2015. The 2013 core (ML131A) was taken off the north eastern shore towards the center of the lake. The ML15EBC was taken from the north ea stern central bank of Mirror Lake (18 m water depth) within proximity to an existing avalanche path location ( Figure 3 ) In September 2015, the Cottonwood Lake core (CWL15) was sampled off the south western shore towards the center of the lake (Figure 3) The cores were taken from an anchored floating platform with a 5 cm diameter Livingston square rod piston sampler (Wright et al 1983). Approximately the top meter of sediment was captured from the lakes in 2015 in order to capture the most recent (~2000 years) geomorphic events. The c ore ML15EBC re covered 0.92 meters of sediment The Cottonwood Lake core CWL15 was extra cted from 7.4 m water depth and 0.98 m sediment was retrieved. The unconsolidated sediment at the top of the profile was not captured in the sample, but captured using a Klein short coring devise. The cores were measured for length and lithology described. The cores were pla ced in plastic wrap and PVC pip e for protection and transported back to the laboratory and refr igerated. Laboratory Methods Core preparation and lithology Each core was opened individually and lengths taken to compare against recorded field measurements. Cores were split lengthwise and lithology described in detail. Cores we re subsampled at cont inuous one centimeter increments and stored in 2 oz labeled Whirl Pak bags. One half of each core was left un sampled and archived at the University of Colorado Denver Paleoecology, Palynology, and Climate Change Lab oratory

PAGE 36

27 Magnetic susceptibility & loss on ignition Sediment magnetic susceptibility was continuously m easured on Cottonwood Lake (CWL15) and Mirror Lake (ML15 EBC) cores to determine clastic inputs of iron bearing sediments (Gedye et al ., 2000). A Bartington MS2E point sensor was moved al ong each intact sediment c ore at room temperature and at one centimeter intervals for the length of each core Measurements were recorded in units of centimeter gram second (cgs). Loss on ignition (LOI) was performed to assess changes in background lake sediment productivity and sediment input. LOI was measured at two centimeter intervals, and at one centimeter inter vals in areas of interest. One cm 3 samples were dried for twenty four hours at 90 C to remove moisture from the samples Percent organics were calculated based on weight loss after heating t he samples for two hours at 550C Lastly, total percent carbonates were determined from subsequent weight loss after heating for an additional two hours at 900 C (Dean 1974). Magnetic susceptibility and loss on ignition analysis was performed on the ML131A in 2013 using the same methods (Del Piore, 2015). X RAY Fluorescence X Ray Fluorescence (XRF) was measured on the CWL15 and ML15EBC cores to determine the geochemistry and elemental makeup of the lake sediment cores pulled from both lakes (Finkenbinder et al ., 2014) Measurements of the LOI residual sediment were taken using a handheld Olympus Delta scanner on the residual dried sediment yielded from LOI analysis at 2 cm intervals a nd at higher resolutions in locations of interest The sediment was homogenized by grinding the sediment using a n agate mortar and pestle and packed into p XRF sample holders and covered with a 4 micron ultral ene window film for measurement XRF measurements were run in both soil and geochemistry (yielding averag es of three measurement) modes. Geochemistry data was consolidated and visualized using C2 software to determine trends in composition with regards to the locations of the disturbance event layers.

PAGE 37

28 Particle Size Analysis Particle size distributions (PSD) were determined on 1 cm 3 of se diment at 2 cm intervals, and at finer 1 cm intervals for areas of disturbance layer interest. Samples were treated with hydrogen peroxide (30%) for 24 hours to oxidize organics and then deflocculated with heating treatments and distilled water at 70 C as per Gray et al 2010 Treated sediment was transferred to scintillation vials and combined with sodium hexametaphosphate for transportation. Samples were analyzed at Dr. the University of California Riverside Environmental Science L ab using a Beckman Coulter (LS 13 320) Grain Size Analyzer. The resulting PSDs were used to describ e the distribution of grain sizes throughout the CWL15 and ML15EBC cores. The ML131A visual interest for particle size analysis focused around previously noted disturbance s as per Del Piorie, 2015. Data Analysis X Ray Fluorescence Element concentrations derived through XRF are used to infer relationships between lake sediments and the surrounding geolo gical environment. A ratio of r ubidium (Rb) to strontium (Sr) was used to identify disturbed allochthonous material from gradual autochthonous sediments produced within the lake A higher ratio suggests sediments that have been chemically weathered ( Jin et al ., 2001; Jin et al ., 2006 ) Lower ratio values infer sediment of more detrital nature brough t down by mass wasting events. Ratios were calculated as Rb/Sr. Autochthonous sediments were interpreted as consistently higher ratio values (>0.42.) Al lochthonous sediments were distinguished by the presence of lower ratio values (<0.42). This threshold was determined through a combination of observations in data where decreases in ratios correspond with dark coarse grained layers within the cores and values utilized within current literature (Vasskog et al ., 2011).

PAGE 38

29 Particle Size Analysis Statistical data processed through GRADISTA T, a program used to quickly an alyze and produce grain size statistics as per Folk and Ward ( 1957 ) was used to determine grain size characteristics (in metric units) specific to autochthonous versus allochthonous sediments (Blott and Pye, 2001) Thresholds reported in current literature (Vasskog et al ., 2011) were used to assess similar conditions to the Mirror and Cottonwood cores. Background autochthonous sediments were composed of higher levels of consolidated clays and silts. A utochthonous material had varied levels of sorting between cores, with mean grain sizes less than 50 m. Allochthonous disturbed sediments were composed of coarser silts and sands and were determined as poorly sorted material with a mean grain size greater than 50 m (Nesje et al ., 200 7; Vasskog et al ., 2011) Differences in geomorphic disturbances were determined within the allochthonous sediments by assessing the relationships between sorting and grain size. Avalanche events ( Figure 6; Event 1) were determined as events with a mea n grain size >50 m with sorting values > 3 m. Flooding events ( Figure 6; Event 2) were determined as events with a mean grain size >50 m with sorting values between 3 and 2.5 m. Density currents ( Figure 6; Event 3) were characterized to have a mean g rain size >50 m and sorting values <2.5 m. Aver ages of autochthonous sediment particle size distributions were taken from all three cores and plotted against s pecified geomorphic events 1 3 (snow avalanche, flood, density current) to determi ne distribution characteristics. Folk and Ward skewness and kurtosis descriptions rendered through GRADISTAT were used to describe differences within in each of the distribution s Table 3 describes particle size ranges as per Folk and Ward ( 1957 )

PAGE 39

30 Table 4 : GRADISTAT particle size descriptions (Blott and Pye, 2001) adapted from Folk and Ward, 1957. Sorting Skewness Kurtosis d Very Well Sorted <0.35 Very fine skewed +0.3 to +1.0 Very Platykurtic <0.67 Well Sorted 0.35 0.50 Fine Skewed +0.1 to +0.3 Platykurtic 0.67 0.90 Moderately well sorted 0.50 0.70 Symmetrical +0.1 to 0.1 Mesokur tic 0.90 1.11 Moderately sorted 0.70 1.00 Coarse skewed 0.1 to 0.3 Leptokurtic 1.11 1.51 Poorly Sorted 1.00 2.00 Very coarse skewed 0.3 to 1.0 Very Leptokurtic 1.50 3.00 Very Poorly Sorted 2.00 4.00 Extremely leptokurtic >3.00 Extremely poorly sorted >4.00

PAGE 40

31 CHAPTER V RESULTS The results of the analyses have been separated in to the following sub sections : core descriptions and analysis, m agnetic susc eptibility & loss on ignition, geochemistry/XRF, and p article size distributions. The core descriptions and analysis section describes the Cottonwood and Mir ror l ake sediment cores based on visual inspection and observations of major sedimentation changes and physic al composition. M agnetic susceptibility of sediments and loss on ignition of the sediments were used to determine clastic inputs and lake productivity, respectively. The next section describes cha nges in geochemistry and fluctuations of elemental composition within the records. In the last section, particle size analysis and statistical characteristics are described to determine differences throughout the sediment core s Lithology Co re Descriptions and Analysis Lake cores f rom both Cottonwood and Mirror l ake s were composed primarily of fine detritus gyttja (FDG) sediment containing a combination of organic matter, inorganic precipitates, and varying mineral composition Coarse grained sediment, highly visible minerals (mica and quartz), macrofossils (wood and other botanical debris), and sediment greater than 1 mm were found interspersed with in the FDG sediments.

PAGE 41

32 Figure 4 : Study Core Lithologies. a. data from Cottonwood lithology, loss on ignition, magnetic susceptibility and geochemistry data. b. data from Mirror Lake (15EBC) lithology, loss on ignition, magnetic susceptibility and ge ochemistry data. c. data from Mirror Lake (131A) lithology, loss on ignition and magnetic susceptibility.

PAGE 42

33 Cottonwood Lake ( CWL15) The CWL15 sedim ent core was composed of light brown FDG with two large sections of 2 3 cm thick inorganic sediment intermixed in coarse d etritus gyttja (CDG) (Figure 4 a ). FDG at the bottom of the core, from 100 to 83 cm depth, transitioned to the first section of inorganic and CDG sediment layers (82 to 67 cm depth). There were visual traces of mica and macrofossils such as woody debris and pine need les within the CDG. The dark coarse layers transitioned back to a lighter FDG, from 67 to 48 cm depth, with one inorganic sediment layer at 58 to 56 cm depth A second zone of dark inorganic layers and CDG with visible macrofossils and m ica occurred from 48 to 21 cm depth. The top 20 cm of the core consisted of FDG. Mirror Lake (ML15EBC) The ML15EBC core was pre dominately composed of a light FDG sediments with frequent visible darker silt bands towards the beginning of the record between 90 to 68 cm depth. Layers of dark silt also occurred at 62 to 60.5 cm and 57.5 to 56 cm depth. A darker grey brown layer of coarse silt occurred at 85 to 84 cm depth. Since no visibly identifiable sediment layers existed in the top part of the core (53 to 0 cm depth) this ana lysis focused only on the lower half to the record (100 to 53 cm depth) (Figure 4 b). Mirror Lake ML131A The ML131A core was orig inally described in Del Pior e 2015 The core was composed of dark gy ttja charac terized by coarser and darker gy ttja layers with a high abundance of macrofossi ls, mica and silicates (Figure 4 c). From 200.5 to 184 cm depth sediment was observed as FDG with a darker coarse layer from 187 to 185 cm depth. The ML131A core transit ioned back to FDG (184 to 104 cm depth) with darker layers occurring at 159 to 156 cm and 128 to 109 cm depth. Core composition then transitioned from FDG (104 to 6 cm depth) to dark coarse r layers present at 88 to 86 72 to 71 69 to 67 34 to 28 and 19 to 17 cm depth.

PAGE 43

34 Magnetic susceptibility & loss on ignition Cottonwood Lake (CWL15) The CWL15 sediment magnetic susceptibility increased from an average of 1.2 cgs a t the bottom of the core to an average of 2.1 cgs at the top of the core. At the bottom of the core from 89 to 87 cm depth, magnetic susceptibility fluctuates from 1.34 to 0.802 cgs before increasing quickly to 2.5 cgs at 70 cm depth. The magnetic s uscep tibility remained low from 70 to 51 cm depth at an average of 0.65 cgs before increasing to 2.21 cgs at 47 cm depth. Between 47 to 20 cm depth magnetic susceptibility fluctuated between 1.04 to 2.97 c gs before decreasing quickly to 0.79 cgs at 19 cm depth From 19 cm depth to the top of the core, large fluctuations occurred between 0.302 and 3.05 cgs. Peaks in magnetic susceptibility occurred at depths of 70 cm at 2.5 cgs, 22 cm at 2.97 cgs, 12 cm at 2.58 cgs, and 7 cm at 3.05 cgs. These pulses suggest an increase in allochthonous iron bearing sediment to the lake. Loss on ignition data in the CWL15 core indica tes a decrease in organics from the bottom of the core ( average 10.9 % ) to the top ( average 8.8%) ; although even at the bottom of the core total organic content never exceeded 25%. Undisturbed FDG sediments displayed lower levels of organic content, where as in areas whic h correspond with darker coarse layers, organic content increased. Areas of FDG averaged 9.5% organic s while the coarse FDG, between 80 to 79 cm depth ranged between 14 and 22%. Smaller increases a lso occ urred at 72 (18.4%) and 59 cm d epth ( 15.6%). Mirror Lake (ML15EBC) The ML15 EBC sediment magnetic susceptibility remained consistently low throughout the core, with an average of 0.153 cgs and a slight decrease to 0.0759 cgs towards the top of the core starting at 60 cm At the bottom of the core, the magnetic susceptibility fl uctuated slightly and decreased between 100 to 91 cm depth from 0.112 to 0.0573 cgs. From 90 to 82 cm depth, there was a slight increase from 0.104 to 0.192 cgs before decreasin g to 0.0653 cgs at 81 cm depth. Wit hin the

PAGE 44

35 middle section of the core, magnetic susc eptibility increased slightly between 80 to 74 cm depth from 0.123 to 0.175 cgs respectively, before increasing quickly to 0.594 cgs at 73 cm depth. Magnetic susceptibility at the top of the core remained consistently low wi th little fluctuation from 73 to 53 cm depth. The only noticeable pulse within the ML15EBC core occurred at 73 cm depth ( 0.594 cgs ) The loss on ignition data for Mirror Lake (ML15EBC) records a decrease in organics from the bottom ( average 16.3%) of the core to the top ( average 11.8% ) The percentage of organics did not exceed 22% anywhere within the core. The uninterrupted FDG sediments disp layed lower levels of organics (mean 15.4%). Areas of darker sediment layers had slightly higher level s of organic material ( 16.8% ) At 84 cm depth, organics dropped to 7.6% indicating a pulse in inorganic material into the lake Mirror Lake (ML131A) The ML131A core magnetic susceptibility and loss on ignition data was originally an alyzed and completed in 2014 (Del Piore 2015). Magnetic susceptibility increased gradually from the bottom of the core to the top from 1.09 to 2.4 cgs. From the 200 to 183 cm depth magnetic susceptibility fluctuated sligh tly before rising from 1.09 to 2.434 cgs. From 184 to 143.5 cm depth, ibility increased from 1.44 to 2.11 cgs The magnetic susceptibil ity decreased from 1.90 to 0.734 cgs before increasing at 119 cm depth to 1.19 cgs. Magnetic susceptibility decreased and remained relatively low until 110 cm depth at 1.18 cgs before increasing to 3.14 at 106 cm depth. Values became more variable from 105 to 60 cm before distinct sharp pulses began to occur at 59 cm depth at 3.9 cgs to 51.5 cm depth at 11.5 cgs. At the top of the core magnetic susceptibility decreased from 45 to 10 cm depth. Values varied slightly through the middle of the core 60 to 45 cm depth ( 1.47 5.10 cgs). The ML131A loss on ignition analysis showed a decrease in organic material from the b ottom ( average 15.1% ) to the top of the core ( average 9.4% ) and did not exceed 25% anywhere

PAGE 45

36 within the record. In areas of undisturbed FDG sediments, percentages of organic c ontent remained relatively low ( only 8.1 %) T he bottom of the core, ( from 196 to 150 cm depth ) displayed an average of 14.2% organics. Within the middle of the core FDG sediments ( at 148 cm depth ) showed 12.6% organics which decreased to 9.1% at 120 cm depth. Organic material remained low at the top (7.3 to 9.8%) from 20 to 10 c m depth. A layer of darker material at 19 cm depth recorded 7% organics Geochemistry/XRF Cottonwood Lake (CWL15) The CWL15 geochemistry primarily focused o n the elemental composition of rubidium (Rb) and str ontium (Sr ) as they can be used to suggest differences in autochthonous and allochthonous materials Rb/Sr ratio values fluctuated throughout the record. At the bottom of the core, from 97 cm to 85 cm depth, ratio remained high fluctuating between from 0.41 and 0.4 3 Rb/Sr ratio then decreased from 84 to 72 cm depth from 0.44 to 0.31. In th e middle of the core (from 65 to 55 cm depth), composed primarily of FDG sediment, Rb/Sr ratio s increased to 0.5. From 47 to 23 cm depth where darker coarse sediment layers oc curred Rb/Sr ratio s decreased to 0.33 with greater variability The top of the core (from 22 to 0 cm depth), composed primarily of FDG sediments displayed relatively higher levels of Rb/Sr ratio s ( 0.502 ) Mirror Lake (ML15EBC) The ML15EBC Rb/Sr rati os remained relatively high throughout the core with some variability The bo ttom of the ML15EBC core (99 to 86 cm depth) recorded Rb/Sr ratio of 0.52 to 0.56. Rb/Sr ratios decreased after 83 cm depth to 0.29 b efore abruptly increasing again after 78 cm de pth to an average of 0.48 The middle of the core ( 84 to 70 cm depth ) composed of FDG remained relatively high ( 0.51 ) from 78 to 68 cm depth before decreasing to 0.25 after 67 cm depth. The top of the core from 65 to 53 cm depth, composed primarily of FDG, ratios remained high ( 0. 52

PAGE 46

37 0.67 ) with little variability and experienced the highest ratios occurred at 58 (0.6) and 54 cm (0.67) depth. Particle Size Distribution Cottonwood Lake (CWL15) The Cottonwood Lake core was primarily composed of sand (56.8%) and lower amounts of clay ( 4.2 %) and silt ( 39.2 ) (Figure 5 a). Increased percentages of sand occurred in inorganic layers (83 to 67 and 47 to 23 cm depth), while higher clay and silt percentages occurre d in the FDG. Inorganic layers contained very poorly sorted particles (83 to 67 cm depth, and 48 to 23 cm depth) and became more sorted as the sediment transitioned back to F DG (67 to 57 cm depth) (Figure 6 a). The distribution of average FDG sediments ha d a mesokurtic to platykurtic symmetrical grain size. Distributions taken from specific darker coarse layers displayed a more leptokurtic and a coarsely skewed distribution (see Table 3) Event 1 represents the grain size distribution from 68 cm, even t 2 at 70 cm, and e vent 3 at 78 cm depth. (Figure 6 b). Mirror Lake (ML15EBC) The Mirror Lake 15EBC core was composed of silt (77%) and lower amounts of sand (13.3%) and clay (98%) (Figure 5 b). Sand remained low (< 10%) throughout the core, but at 84 cm depth, a sharp increase from 10 to 55% occurred. Mean grain size remained low and did not exceed 1 mm in the core. The ML15EBC core was relatively poorly sorted but at 84 cm depth became very poorly sorted. The di stribution of averaged FDG sediments displayed a mesokurtic symmet rical grain size curve (Figure 6 d). The area of interest at depth 84 cm represented by event 3 (Figure 6 d) became more leptokurtic with a coarsely skewed grain size distribution (Table 3) s uggesting a shift from autochthonous sediments to allochthonous.

PAGE 47

38 Mirror Lake (ML131A) The ML131A core was composed of sand (average 55.6%) with slightly less silt (average 42%) and lit tle clay (average 11%) (Figure 5 c). Areas of increased sand ( 81.4% ) and the larger grain size (166 m ) occurred in the inorganic sections, whereas silt was highest in the FDG ( 82.8% ) Mean grain size increased throughout the record towards the top of the core where sediments displayed the larg est mean sizes ( 230.6 m ) Grain sizes greater than 1 mm occurred five times throughout the record (27.5, 55, 65, 67, 106 cm depths). Layers (0 to 29.5 cm depth) of coarser materials were composed of poorly sorted material with larger mean grain sizes, contrasting with the more wel l sorted material with smaller grain size of FDG sediments. The distribution of average FDG sediments showed mesokurtic and slightly leptokurtic curves that were slightly skewed towards coarser grain sizes. Disturbances layers displayed more leptokurtic with slightly coars e skewed distributions (Figure 6 f). Event 1 represents the grain size distribution at 55 cm depth event 2 at 27.5 cm depth and event 3 at 17.5 cm depth. It is important to note that the ML131A core was processed for PSD as a seco ndary analysis of ML15EBC

PAGE 48

39 Figure 5 : Percent age compositions of three lake cores as indicated in legend above a. Cottonwood Lake (CWL15). b. Mirror Lake (ML15EBC). c. Mirror Lake (ML131A).

PAGE 49

40 Figure 6 : Particle size analysis for the CWL15, ML15EBC, and ML131A cores. a CWL15 Folk and Ward method bivariate plot of sorting (m) versus mean grain size (m). b Plot of CWL15 average grain size (m) distributions of background sedimentation versus grain siz e distributions of differing types of events; 1 (sample 46 at depth 68 cm), 2 (sample 48 at depth 70 cm), and 3 (sample 56 at depth 78 cm). c ML15EBC Folk and Ward method bivariate plot of sorting (m) versus mean grain size (m) d Plot of ML15EBC averag e grain size (m) distributions of background sedimentation versus a differing event (sample 91 at depth 84 cm). e ML131A Folk and Ward method bivariate plot of sorting (m) versus mean grain size (m). f Plot of CWL15 average grain size (m) distributio ns of background sedimentation versus grain size of differing events; event 1 (sample 106 at depth 55 cm), event 2(sample 100 at depth 27.5), and event 3 (sample 99 at depth 17.5 cm).

PAGE 50

41 CHAPTER VI DISCUSSION The following chapter discusses sediment depositi on characteristics of snow avalanche events in high elevation lakes of central Colorado, and how they can be distinguished from autochthonous sediments and oth er disturbance events that deposit allochthonous sediment (e.g. flood and debris flow). Cottonwo od and Mirror Lake sediment records are compared to demonstrate how these events can be separated within two separate watersheds in the Sawatch Mountains of central Colorado. In addition, two cores from Mirror Lake allow for a with in lake comparison Distinguishing between autochthonous from allochthonous disturbance Allochthonous sediments deposited through instantaneous disturbance events (e.g. snow avalanche or floods, or debris flows) are often visibly identifiable among more abundant autochthono us fine det ritus gyttja (FDG) and other internally redistributed allochthonous sediments through time Allochthonous layers deposited after a disturbance event, are often distinguished as coarser grained sediment (e.g. silts to small pebble s) with higher abundance of macrofossils brought into the lake (Nesje et al., 2007). The CWL15 and ML131A core s both contain distinct sections of disturbed sediment composed of coarser silt and sandy material with macrofossils and traces of mica not found in the FDG sediments (Figure 3a). The ML15EBC core, in contrast, is composed of almost all FDG with infrequent brown grey band and one dark brown sediment layer at 84 cm depth suggesting a core composed primarily of undisturbed autoc hthonous material (F igure 4 b.). Therefore, the differences in sediment characteristics in the Mirror Lake cores, suggests that the disturbance events are localized and are not recorded throughout the lake. Allochthonous sediments deposited by mass wasting events such as avalanches are known to display distinct geochemical signatures differing from FDG sediments subject to chemical weathering associated with the watershed (Vasskog et al., 2011; Croudace et al. 2006). The geochemical

PAGE 51

42 analysis of the sediments, using x ray f luoresce nce (XRF), provides a means of determining sediment provenance and parent material. For example, Rb/Sr ratios have been used as a proxy for evaluating chemical weathering and past climates within individual watersheds (Jin et al., 2001). Rubidium (Rb) is linked to detrital clays more resistant to weathering processes typically found in autochthonous sediments whereas Strontium (Sr), can be used to reflect activity during chemical and physical breakdown (Jin et al. 2001 ; Jin et al ., 2006 ). This ratio increases significantly in the presen ce of weathered rock, reflecting chemical erosion, versus fresh or newer rock (Dasch, 1969). Rb/Sr ratios are well reflected with in igneous (basalt and granite ) rock The Rb/Sr ratio can therefore be used to distinguish weathered sediments and material introduced by mass wasting events within the Mirror and Cottonwood lake records as local geology contains high concentrations of basalt and other igneous rock Autochthonous sediments display higher values o f RB/Sr ratios due to the slower organic and clay ac cumulation Mass wasting events (i.e. snow avalanches) display lower Rb/Sr ratios indicating an increased amount of detrital material and inorganics that are not subjected to gradual modes of weathering. The CWL15 sediment had an average Rb/Sr ra tio of 0.43, but displayed the lowest ratios ( 0.34 ) and great er variability in areas of coarse grained sediment At the bottom of the core between the depths of 83 to 67 cm depth the greatest variability occurred in the dark coarse g rained sediments. Locations of lower Rb/Sr ratios corresponding with coarse layers suggest sediment transported into lake rapidly and that has not sustain ed extensive chemical weathering. Contrastingly, highest ratio s of CWL15 of 0.57 occurred within areas of uninterrupted lighter FDG found at 59 cm depth (Figure 4 a). Similarly, ratios of ~0.5 occurred at 16 cm depth within FDG sedimen ts. ML15EBC composed primarily of FDG had an average Rb/Sr ratio of 0.45 as w ell as more consistently higher values than CWL15, with some fluctuation ( Figure 4 b). Although minor variability occurred within ML15EBC the higher ratios suggest a consistent and relatively u ndisturbed sediments that have undergone chemical weathering. The depths at which the higher and l ower values occurred within CWL15 and

PAGE 52

43 ML131A suggest differences between autochthonous sediments naturally broken down through time from allochthonous sediments abruptly deposited into the lake via geomorphic disturbanc e. Analysis of grain size distributions (e.g. core percent composition, mean grain size, sorting, kurtosis, and skewness) are useful indicator s of sediment transport and deposition mechanism s (Folk and Ward, 1957; Beierle et al. 2002). V isibly identified allochthonous sediments in core CWL15 record large grain sizes with higher percentages of sand ( 70.2 %) (Figure 4a). Similar layers in M L131A also coincided with the highest percentages of sand ( 53.6%) ( Figure 4c). A layer from 117 to 99 cm depth records an increased percent sand from 37% to 61% before decreasing to 33% in areas of FDG (Figure 5 c). The FDG sediments record high instances of finer silts ( 73%) and consolidated clays ( 8.6%) in CWL15 between of 65 to 49 cm depth (Figur e 4a). The ML15EBC cor e had the highest levels of medium silt ( 80.2 %) throughout the core, but at depth 84 cm, a sharp increase from 13% to 53 % in coarse grain sand occurred, before sharply declining back to 7.7% indicat ing the location of a geomorphic disturbance event (Figure 4b) and suggesting core composition of autochthonous sediment with a layer of allochthonous deposited material Particle size distributions can also distinguish between autochthonous and allochthonous disturbance s Distur bance events quickly introduce substantial amounts of detrital material of varying sizes into the lake. Occurrences of mean grain size larger than 1 mm are not seen within lake cores in areas of FDG, and largest grain diameters were only recorded wi thin disturbed sediments (Figure 5). Previous literature has determined that particles with D > 1 mm to be a strong indicator of snow avalanche events, but in the absence of particles below this threshold other indicators within particle size characteristics exist to be used to identify snow avalanche signals. ( Nesje et al. 2007; Vasskog et al ., 2011). For example, CWL15 mean grain size exceeding 1 mm occurred 12 times in areas of allochthonous sediments (83 to 67 cm and 47 to 23 cm depth ; Table 4 ) whereas the autochthonous sediments at the top of the CWL15 core (6 to 2 cm depth) were well sorted and did not exceed an average grain size of 41 m (Figure 6 a). The ML15EBC layer (84 cm depth)

PAGE 53

44 had a mean grain size of 49 m (Figure 6 c) and the FDG unit did not exceed a mean grain size of 29.25 m, suggesting a core composed primarily of FDG sediments. The ML131A core recorded mean grain size exceeding 1mm four times (Table 4) with areas of visible distur bance contrasting to undisturbed FD G with a mean grain size not exceeding 48 m (Figure 6 e). Due to rapid transport and deposition, all ochthonous disturbance sediment is poorly sorted and are distinguishable from well sorted autochthonous sediment. Highest sorting values (indicative of poorer sorting) coinciding with higher mean grain size are strong indicators of allochthonous disturbance events (Vasskog et al 2011) Although sorting of a ll three cores was poor FDG sections were not as poorly sor ted as allochthonous disturbance events with medium to large grain sizes. FDG sediments within CWL15 were poorly sorted, but with only a low to medium mean grain size ( 13 to 47 m ) RSE at depths of 83 to 67 cm and 47 to 23 cm depths recorded both large grain si zes and poor sorting compared to autochthonous FDG sediments (Figure 6 a). The difference between particle size characteristics of autochthonous and allo ch th o nous disturbance sediment can be visualized using bivariate plots of particle size distribution descriptors (Beierle et al. 2002). Background sediments of all three cores have symmetrical grain size distributions, wi th a slight skew towards coarse silts (Figure 6 b,d ,e Table 3 ). Disturbance events from the cores are coarsely skewed (Table 3) and distinct from the autochthonous sediment curves of CWL15 and ML15EBC The disturbance events have higher percentages of larger grains that create s t sized particles, evident within the CWL15 and ML1 31A cores. The slight bimodal distribution suggests a more heterogeneous particle matrix that is introduced to the system through a quick and fluid pulse event (Spencer, 1963). Autochthonous FDG distributions of both CWL15 and ML15EBC display ed lower mes o or platykurtic (less peaked) distributions (Table 3) The meso and platykurtic ( less peaked ) distributions of autochthonous sediments contrast with disturbance event distributions of all three cores that are far more leptokurtic (more peaked) and s kewed to the left by the presence of coarse silts and sands. The ML131A autochthonous sediment

PAGE 54

45 distribution is slightly more leptokurtic than the other two cores, which may be due to the fewer number of samples analyzed in proximity to visibly coar ser event layers. In summary, the visual, geochemical, and particle size analyses allow for th e characterization between autocht honous derived FDG sediments and those deposited through rapid deposition during a disturbance event. The vi sible presence of more detrital material and higher occurrence of mica infers rapidly mixed sediments introd uced into the record in comparison to sediment which has been gradual ly decomposed and weathered through time. The visual differences of background sediments and di sturbance layers further correspond with the geochemistry in both CWL15 and ML15EBC The p resence of higher Rb/Sr ratio ( >0.42 ) suggests that background sediments have experienced a higher degree of chemical weathering, while the lower ratios ( <0.42 ) fou nd in disturbance layers indicates the entrainment of less chemically weathered sediments through strong physical weathering (Vasskog et al., 2011; Boggs, 2001). Allochthonous disturbance sediments contain very poorly so rted material (coarser silts and sands) with a higher mean grain size ( 52 172 m ) which suggests an increased depositional energy entering the lake (Nesje et al. 2007). Furthermore, sediment that is chemically weathered through time were defined by Gaussian distributions, contrasting di stinctly with than those derived from a disturbance which were leptokurtically peaked and coarse ly skewed negatively to the right due to the presence of larger grains (Sp encer, 1963) ( Figure 6 b,d,e ) The differences in distributions suggests allochthonous disturbance material deposited by mass wasting events contain larger, newer, and more detrital material that has not been subjected to the degree of chemical weathering found in autochthonous FDG sedi ments. The results suggest that several proxies should be used to charact erize geomorphic disturbance within a lake basin Visual inspection of core lithology is useful for quickly identifying areas of disturbance, but not all events are visible. For example, at the top of the CWL15 core, what appears to be uninterrupted autochthonous sediments (1 cm depth) (Figure 4 a), are disturbance events identified by their poorly sorte d sediments with a higher mean grain size of 84 m. This is also

PAGE 55

46 evident within ML131A where a darker FDG (5 cm depth) was found to contain poorly sorted materials with a mean grain size of 49 m Conversely the ML1EBC had visible dark gray banding, whi ch was initially interpr eted to be disturbed sediments, but lacked particles with medium to high mean grain size. Similar problems arise when interpreting Rb/Sr ratios. Current literature has noted that using elemental analyses prov id es the strongest sep aration of autochthonous and allochthonous sediments (Vasskog, et al., 2011; Jin et al ., 2006); however, ML15EBC composed of FDG sediments with very low mean grain size still had fluctuations and the presence of relatively low Rb/Sr ratios. Given all the measures conducted in this study to characterize geomorphic disturbance events, particle size analysis provides the most powerful results ; howeve r, other proxies help to verify disturbance from non disturbance events. The e ffectiveness of other proxy indicators examined in this study are influenced by local and regional characteristi cs such as climate and topography. For example, m easurable spikes in magnetic susceptibility have been well documented within the literature to indicate pulses of iron bearing clastic sediments entering the lake (Wondzell and King, 2003 ; Blake et al ., 2006 ) This proxy is limited to watersheds with iron bearing sediments or where fire events were frequent and intense enough to influence the magnetic s of sediments B oth Mirror and Cottonwood lakes however, record peaks during events identified t hrough particle size analysis and geochemistry. The geology of the watersheds did not contain iron bearing bedrock. Loss on ignition, which is used to measure lake productivity, also did not decrease during a geomorphic disturbance This is likely due to large macrofossils being brought into the lake along with the inorganic sediments. Characterizing different all ochthonous events (e.g. snow avalanches floods debris flows) The grain size properties of allochthonous disturbance layers can be used to distinguish differences in geomorphic mass wasting events common in high elevation lakes such as: snow avalanches, floods, and density currents (Vasskog et al. 2011). Statistical differences in particle size distributions can be used to infer distinct methods of transport and depositional characteristics of

PAGE 56

47 particles in to the watershed (Tanner, 1991). Using statistics calculated in GRADISTAT software, grain size properties were used to describe and interpret different RSE (Folk and Ward, 1957; Blott and Pye, 2001). By plotting derived grain size parameters together, si gnatures of three disturbance events can be identified within both Cotto nwood and Mirror l ake s ediment cores (Figure 6 ). Wi thin this sub section, differences in disturbance event signatures are discussed. Figure 7 : Disturbances the watershed. Summary of disturbance characteristics and transport in to the lake.

PAGE 57

48 Snow Avalanche Events Snow avalanche events (Event 1) w ithin Cottonwood and Mirror lakes are associated with poorly sorted medium to high mean grain size with higher concentrations of larger particles. The presence of grain sizes greater than 1 mm is a strong indicator of avalanche activity; however, this depends on the type of the slide (Vasskog et al. 2011). Both CWL15 and ML131A showed avalanche even ts to be the most poorly sorted of the three disturbances with the larges t range of grain sizes (Figure 6 a). In both cases, instances of coarse sand exceeding 1 mm were evident; however layers were described as coarse silts and fine sands ( Table 3 ). A va lanche events ( Figure 6b d e; Event 1) were identified by higher concentrations of coarse sands. This contrasts with the other types of disturbances that while more highly leptokurtic ( Figure 6 b,d,e; Events 2 and 3; Table 3) do not display the same concentration or abundance of coarse sands (Figure 6 b). The poorly sorted sediment deposited from a s now avalanche is inferred from the intense mixing of debris transported down an avalanche path and deposited in a fan shape feature at the end of the runout zone, or lake. Further variation in sorting and sediment mixing exist s due to the differing types of snow avalanche activity (Hungr 1985). For example, slush flows are associat ed with inconsistent early spring weather conditions, such as higher temperatures, and fluctuations are thought to transport more material due to decreasing stability of snow pack and decaying crystal layering (Schweiser et al. 2003). On the south wester n edge of Cottonwood Lake, near where the sediment core was taken, the avalanche path suggests slush flow slides based on topography and changes in vegetation including large woody detrital debris deposited at the lakes edge When a snow avalanche slide over the water before settling out and allowing debris to sort through the water column. The quick pulse like nature of the snow avalanche event is evident by the poor sortin g and mixed deposition of grain sizes. Such transport variability is reflected in all three core s, which are more leptokurtic and coarsely skewed than the FDG sediments It is worth noting that as Mirror Lake is at a higher

PAGE 58

49 elevation and experiences colder average temperatures, so avalanche material is more likely to encounter lake ice leaving a fan of sediment on the ice that enter s the record upon melt. This allows the deposition of larger particles further off shore whi ch is likely why the ML15EBC core lacks higher resolution in its event signals than in ML13 1A It has been suggested that rapid precipitation events, common to the central Rocky Mountains, could cause a mud flow into the lake record, as indicated for the southern Sierra Nevada Mountains (Anderson, 1990). Would such an event leave similar signatures within the lake sediment as snow avalanches? The signals would be different because w inter precipitation events are uncommon to central Colorado Rocky Mountains, with most winter precipitation occur ring in the form of snowfall, contrasting with the wetter winter s (snow and winter rains) that characterize the southern Sierra Nevada Mountain s (Anderson, 1990). This sh ould not be confused with sediments deposited by spring and summer thunderstorms. Such deposits would be more likely to cause hypernycal flood pulses or slope failures depositing sediments with different grain size character istics closer to those of a flood or density current. Flood Events Flood events (Event 2) can be visually differentiated from other disturbance events within lake sedime nt records through higher minerogenic composition and presence of macrofossils. Identified flood event s within the ML131A core contained high concentrations of macrofossils such as pine needles and other woody debris. Flooding events are associated with larg e mean grain size (very fine sand to fine sand), but are recognized by their well sorted deposits. Flood event deposits are more leptokurtic and coarsely skewed than avalanches, but lack the characteristic bimodal fat tail reflecting coarse grain size. This suggest s a difference in tran sport and deposition that is characteristic of a higher influx of water and sediment to a lake that lacks the energy and velocity of a snow avalanche. For example, avalanche deposits are more poorly sorted then flood deposits due to the higher transport e nergy of slide activity.

PAGE 59

50 Density Current Events Density currents (Event 3) are interpreted as the least frequent disturbance events identified within Cottonwood and Mirror l ake s Density currents or turbidity currents appear to be discernable from the other geomorphic disturbances by large mean grain size s (very fine to fine sand) that are the most well sorted of the three d isturbance events (Figure 6) (Vasskog et al ., 2011). Observed d e nsity currents display the most peaked leptokurtic distributions tha t are the most coarsely skewed to the right (Figure 6 b,d,e). These characteristics reflect the responsive transport initiated by sediment failure and the fast fluid movement down a slope into and through the lake (Meiburg and Kneller, 200 9) Density current s appeared to be the only disturbance event distinguishable within all three lake sediment cores. These deposits are due to slope failure and downslope transport carrying coarser eroded m aterials that contrast with the finer auto chthonous sediments that have degraded over time (Jo hansen et al ., 2001; Wondzell and King, 2003). Within the CWL15 and ML15EBC cores, density current events appear to have had the largest mean grain siz e (110 179.2 m) (Table 4 ). The observed larger par ticles within the suspended sediment filter out into the record leaving behind layers or turbidities that appear to be well sorted with a higher mean grain size. Flood caused turbidities that can be recorded in watersheds such as Cottonwood and Mirror lakes can be further used to determine hydrological changes to the area (Zhang, 2014). It is important to acknowledge that current literature accepts the identification of density currents and hyperpycnites through a variety of methods including those of this study, but with the addition of others including stratigraphic and horizon characteristics, spatial distributions, and multiple core scales ( Lamb and Mohrig, 2009; Zhang, 2014a; Zhang et al. 2014b). Further investigation into the signals found withi n the Cottonwood and Mirror cores is needed to increase confidence of the interpretations of this study.

PAGE 60

51 Table 5 : Summary of event characteristics within the CWL15, ML15EBC, and ML131A core including type, mean grain size, sorting value, and Folk and Ward descriptions. Event Core Depth (cm) Event Type Magnetic Susceptibility % Organics Rb/Sr Ratio 1 CWL15 1 A 0.84 8.84 2 CWL15 8 10 A 0.68 9.5 0 3 CWL15 18 20 A 0.78 8.34 0.45 4 CWL15 23 28 A 2.5 10.1 0.29 5 CWL15 32 37 A 1.83 11.03 0.36 6 CWL15 39 45 F 0.92 10.3 0.19 7 CWL15 47 A 0.71 10.7 8 CWL15 68 A 0.67 12.46 0.44 9 CWL15 69 DC 0.62 8.6 0.51 10 CWL15 70 F 0.78 11.8 0.46 11 CWL15 71 DC 2.5 11.8 0.45 12 CWL15 72 74 F 1.83 13.98 0.37 13 CWL15 75 A 0.92 11.13 0.42 14 CWL15 76 F 0.71 11.75 0.42 15 CWL15 77 A 0.67 11.2 0.37 16 CWL15 78 DC 0.62 14.1 0.27 17 CWL15 79 A 0.55 22.77 0.3 18 CWL15 80 81 F 0.8 14.28 0.4 19 CWL15 83 91 A 0.78 11.5 0.43 20 CWL15 93 DC 11.7 0.32 21 CWL15 95 97 A 8.9 0.45 22 ML15EBC 84 85 DC 0.12 7.6 0.45 23 ML131A 5 A 2.52 7.65 24 ML131A 17.5 DC 1.18 16.96 25 ML131A 27.5 F 2.2 8.65

PAGE 61

52 26 ML131A 29.5 DC 2.2 27 ML131A 32 F 5.03 28 ML131A 55 A 3.3 3.89 29 ML131A 68 F 1.5 10.31 30 ML131A 70.5 A 0.63 31 ML131A 106 A 3.14 8.95 Event abbreviations: A Snow avalanche; F Flood; DC Density current Event Event Type Mean GS (m) Sorting (m) Grain Size Description Sorting Description Kurtosis Description 1 A 54.5 3.75 Very Coarse Silt Poorly Sorted Leptokurtic 2 A 54.7 3.88 Very Coarse Silt Very Fine Sand Poorly Sorted Meso Leptokurtic 3 A 63.9 3.7 Very Coarse Silt Very Fine Sand Poorly Sorted Leptokurtic 4 A 58.4 3.91 Very Coarse Silt Very Fine Sand Poorly Very Poorly Sorted Meso Leptokurtic 5 A 64.3 3.78 Very Coarse Silt Very Fine Sand Poorly Very Poorly Sorted Meso Leptokurtic 6 F 101.7 2.75 Very Fine Sand Poorly Sorted Leptokurtic Very Leptokurtic 7 A 60.2 3.66 Very Coarse Silt Poorly Sorted Leptokurtic 8 A 101.5 4.28 Very Fine Sand Very Poorly Sorted Very Leptokurtic 9 DC 110.1 2.48 Very Fine Poorly Very

PAGE 62

53 Sand Sorted Leptokurtic 10 F 98.91 2.92 Very Fine Sand Poorly Sorted Very Leptokurtic 11 DC 129.7 2.48 Fine Sand Poorly Sorted Leptokurtic 12 F 113 2.77 Very Fine Sand Poorly Sorted Leptokurtic 13 A 84.7 3.53 Very Fine Sand Poorly Sorted Leptokurtic 14 F 138.8 2.5 Fine Sand Poorly Sorted Very Leptokurtic 15 A 89.02 3.12 Very Fine Sand Poorly Sorted Leptokurtic 16 DC 172.4 2.34 Fine Sand Poorly Sorted Leptokurtic 17 A 91.4 3.17 Very Fine Sand Poorly Sorted Leptokurtic 18 F 120.9 2.86 Very Fine Sand Poorly Sorted Very Leptokurtic 19 A 96.26 3.71 Very Fine Sand Poorly Sorted Leptokurtic 20 DC 148 2.59 Fine Sand Poorly Sorted Very Leptokurtic 21 A 72.05 3.6 Very Fine Sand Poorly Very Poorly Sorted Meso Leptokurtic 22 DC 49.25 3 Very Coarse Silt Poorly Sorted Leptokurtic 23 A 153.4 3.16 Fine Sand Poorly Sorted Mesokurtic 24 DC 230.6 2.49 Fine Sand Poorly Sorted Mesokurtic 25 F 225.1 2.62 Fine Sand Poorly Sorted Platykurtic 26 DC 114.9 2.36 Very Fine Poorly Leptokurtic

PAGE 63

54 Sand Sorted 27 F 55.84 2.9 Very Coarse Silt Poorly Sorted Mesokurtic 28 A 58.3 3.71 Very Coarse Silt Poorly Sorted Leptokurtic 29 F 73.82 2.86 Very Fine Sand Poorly Sorted Very Leptokurtic 30 A 77.81 4.8 Very Fine Sand Very Poorly Sorted Mesokurtic 31 A 78.34 3.24 Very Fine Sand Poorly Sorted Leptokurtic Event Abbreviations: A Snow avalanche; F Flood; DC Density current Localized differences within lake records The disturbance events recorded at Cottonwood and Mirror Lake are highly localized and to reconstruct the mass wasting history of the basin requires a network of cores. The differences between ML15EBC and ML131A provide the best example of event detection Both cores were taken from the same location off the north eastern shore, with ML15EBC closer to shore to align with existing avalanche paths, and the ML131A taken further off shore near the center of the lake (Figure 3). This change in location result ed in cores with visibly different sediment records. The ML131A core has visible detrital debris throughout the core that corresponds with layers of disturbance. The sediment analyses on the ML131A core identified three types of disturbances; however, on ly one type in the ML15EBC. The overall lack of event detection in the ML15EBC core may be the result of the closer proximity to the existing avalanche path and subsequent run out zone on the north eastern shore. Snow avalanches reach their highest velo city and slide energy at the bottom of the avalanche track and run out zone, before fanning out into the lake (or over the flatter landscape) as they lose kinetic energy (Hungr, 1995). Therefore, the ML15EBC core location may have been in the zone of

PAGE 64

55 aval anche translation rather than deposition, particularly during conditions of a frozen lake surface. Under this scenario, the energy of slide activity would be too great to leave record in ML15EBC sediments, while the location of the ML131A core was far eno ugh from shore to catch debris settled out into the water or on the ice However, several types of events other than avalanches can register closer to the shore Sediment failures, often resulting in mud debris flows and turbidities deposit local distributions that are associated with identifiable steep topography (Mulder et al., 2001). For example, in the ML15EBC core the only event was the density current at 84 cm depth, that suggests slope instability from the energy of the slide and increased mixing of deposits. In addition, o n the southern bank of Cottonwood Lake (Figure 3) a well defined semi annual debris flow channel exists. In September of 2016, a sedime nt core was taken within proximity to the flow channel. When extruded and inspected back at the laboratory, the core lost its shape immediately and broke apart, suggesting a higher level of oxygen trapped within the sediments from the turbid disturbance o f sediment. Larger rocks and pebbles, far exceeding the 1 and 2 mm threshold, were present throughout the core. The core physically contrasts from the CWL15 core, and those closer to avalanche tracks. The differing lithologies suggests the core taken nea r the debris flow was significantly disturbed and would not serve as a good core to reconstruct debris flow event frequency; however, the CWL15 core picked up traces of density currents that could have been a result of debris flow activity from the channel significant rains in conjunction with snow melt. This creates fast moving surficial erosion with high velocity down steep topography. The channels derived from debris flows have deep er channels than avalanche tracks, are line d with large rocks and lack flag trees and new shrub growth associated with frequent disturbance

PAGE 65

56 Figure 8 : Example of sediment core extracted from base of mud debris flow channel off the south eastern bank of Cottonwood Lake. The sediment is very unconsolidated and maintain shape. In summary, core sampling locations are an important consideration for future research and multiple cores need to be taken to reconstruct pas t mass wa sting event disturbance. If the ML15EBC core was the only core taken from Mirror Lake little to no acti vity would have been identified producing a low resolution record For future studies cores should be taken around the basin, from central locations, a nd especially near mass wasting landscape features or incoming steams. Cores taken from the center of the lake tend to capture most events with the least amount of mixing due to transp ort Building on existing literature from Norway and future research The lake sediment records examined within this thesis corroborate with findings in current literature ( Nesje et al ., 2007; Vasskog et al 2011). Snow avalanche activity can be identified from autochthonous background sediments most effectively using parti cle size analysis and geochemistry. For example, N orwegian field sites have different environmental characteristics (e.g. local geology,

PAGE 66

57 topography or climate ) than those of central Colorado influence disturbance signals and thresholds within the core. For example, steeper topography could result in greater transport energy, and would carry and deposit large particle sizes. The studies in Norwa y were conducted on a larger lake ( Oldevatnet; area 8.03 km 2 ); however, the signals were picked up on much smaller kettle lakes characteristic of subal pine environments of Colorado. This study suggests that geomorphic disturbance s from mass wasting events are highly localized, and can potentially be missed from cores extracted from large lakes such as Oldevatnet. The extraction of multiple cores from smaller lakes can be used to show the relationship between spatial location and proxy resolution and confidence. For example, Mirror Lake cores a re distinctly different due to the proximity of avalanche paths on the north eastern shore and resulting snow avalanche transport and debris deposit causing the ML15EBC to have low resolution, and the ML131A to more accurately pick up event signals. This is reaffirmed by the CWL15 and the September 2016 core taken at the base of the debris flow channel. Large lakes and cores taken large distance s apart increase the potential for record error and decrease overall resolution for disturbance to the watershed. Using a GIS, the study identif ied landscapes and lakes that had the greatest potential for capturing mass wasting/avalanche histories. Current liter ature utilized field sites situated in lar ge glacial valleys where the probability of snow avalanche activity is high. North America has several possibilities for avalanche/mass wasting reconstruction and the use of a GIS significantly helps narrow down potential sites, but also the most accessible and cost effective sites. Future research would involve develop ing an age model for the sites so that a paleo disturbanc e reconstruction can be developed Radiocarbon dating of macrofossils found in the cores has been problematic due to the number of older macrofossils washed into the lake sediments during disturbance events. Moving forward it would be useful to date material (i.e. pollen or charcoal),

PAGE 67

58 av oiding large macrofossils, within the FDG/autochthonous sediments. Alternatively, optically stimulated luminescence (OSL) dating c ould be used ( Chen and Pagonis, 2011 ) Once a chronology can be established, specifically for Cottonwood Lake which has no ag e model the record can be compared to regional paleoclimatic records for central Colorado ( Anderson 2011; Cook et al ., 2004; Mann et al ., 2008 ) Current paleoecological studies from c entral Colorado suggest experienced a shift from a summer dominated precipitation regime to a winter precipitation regime ~2000 yr BP with less effective summer moisture (Del Piore, 2015). The increase in winter precipitation suggest s increase in historic snow avalanche event frequency due to higher snow pack (Martin et al. 2001; Lazar et al. 2008; Vasskog et al., 2011). Frequency of snow avalanche events, and other geomorphi c disturbances can be compared to historic climatic records to determine if trends exist regarding changes in climate. Paleoflood deposits are particularly sensitive to climate change (i.e. higher frequency of precipitation events and snow melt) (Knox, 2000). Hyperpycnal, and related density currents are linked to climate by flood frequency and impact can record changes to climate (Mulder et al., 2001). Linking flood disturbance layers to regional paleoclimate data is useful can also be useful to further interpret apparent density currents and turbidities as indicators of past hydrologic changes and activity within the two watershe ds with potential climate implications (Zhang, 2014; Molnar, 2004) Other directions for future research would be to examine the impacts of subalpine and alpine wildfire on disturbance event, such as debris flow frequency. Wildfire influences levels of lo cal erosion by removing vegetation from slopes susceptible to failure (Wondzell and King, 2003). The absence of vegetation allows already erodible soils to be exposed to increased surface erosion. Wildfire debris flow relationships have been well documented in southern California and the western slope of Colorado, but little is known about fire debris flow relationships within central Colorado (Wells, 1987; Cannon and DeGraff, 2009; Cannon et al ., 2001). Fire increases infiltration soil slip to the landscape, creating ideal conditions for debris flows during precipitation events (Cannon et al .,

PAGE 68

59 2001). Established debris flow channels, such as those of Cottonwood and Mirror lakes, are susceptible to overland flows and further accentuated by th e influence of wildfire (Wondzell and King, 2003). The combination of sediment chronology and multiple proxies (magnetic susceptibility, loss on ignition, geochemistry, and particle size distributions) would be ideal to further determine this relationship in central Colorado.

PAGE 69

60 CHAPTER VII CONCLUSION In conclusion, the sediment cores from Cottonwood and Mirror l ake offer new, high resolution data on avalanche and other mass wasting disturbance common to high elevations of lakes of central Colorado. This data distinguishes between autochthonous and allochthonous sediments deposited by various disturbance including snow avalanche, flooding, and deposits left from density currents. The research was str uctured around the following three questions: 1. What characteristics distinguish autochthonous and alloc h thonous/disturbance sediments in a lake sediment record? Allochthonous sediments deposited within sediment cores can often be visibly identifiable among autochthonous sediment as da rker colored bands composed of coarser material (e.g. silts to small pebbles) with higher abundance of macrofossils Although certain events are not visibly distinguishable and must be identified using other proxy data analys e s (e.g. particle size distrib utions). Autochthonous sediment is visibly identifiable as lighter FDG composed of consolidated clays and medium silts. A llochthonous sediments deposits display geochemical trends differing from autochthonous background sedimentation Allochthonous sediments display lower Rb/Sr ratio values (<0.42) with higher value variability. Autochthonous sediments yield higher ratio values (>0.42) with less variability. Allochthonous sediments show larger mean grain sizes with higher percenta ges of sand and coarser silts as disturbance events quickly introduce substantial amounts of detrital material into the lake in varying degrees of decomposition. G rain size > 1mm can be used as a strong indicator of mass wasting disturbance, but requires f urther characteristics to determine the type of geomorphic disturbance. It should also be noted that some allochthonous sediment s contain grains ranging from 50 200 m. The rapid depositional energy of disturbance results in allochthonous sediments that are

PAGE 70

61 coarse grain ed and poorly sorted compar ed to the relatively well sorted autochthonous sediment with smaller grain sizes. D ifference s between autochthonous and allochthonous sediments are visualized using bivariate plots Autochthonous sediments display a symme trical grain size distribution with only a slight skew from coarser silts (Table 3). RSE layers are coarsely skewed (Table 4 ) distribution s with sand siz ed particles Autochthonous distributions display lower meso or platykurtic less peaked curves (Table 3) due to smaller standard deviations like that of a Gaussian distribution curve. Disturbance e vent distributions are fa r more leptokurtic skewed by th e presence of coarser silts and sands. The results suggest that the use of several proxies to distinguish geomorphic disturbance events. Visual inspection of core lithology can be useful but not all events are visible. Simila rly, interpreting Rb/Sr rati os can be used as an indicator of autochthonous and alloc h thonous material through chemical weathering ; however areas of allochthonous material show high variability in ratio values and can include higher ratios within areas of visibly identified disturbance. Disturbance events are characterized by their high mean grain size and poorly sorted sediment due to high energy and rapid transport methods. Local environmental variability can impact proxy confidence. Mag netic susceptibility is used to identify disturbance, specifically fire events, within lake sediments ; however, if geologic bedrock does not contain iron bearing material, disturbance can go undetected. Loss on ignition, used to infer lake productivity, d id not decrease during a geomorphic disturbance event likely due to large macrofossils being brought into the lake along with the inorganic sediments. 2. What sediment properties define snow avalanches from other mass wasting events (flood, debris flows, fire, etc) in lake sediments? Snow avalanche events are recognizable within lake sediments as poorly sorted (higher sorting values) material with a medium to high mean grain size. The wide range of mean grain size is likely

PAGE 71

62 due to several types of snow avalanche slides and characteristics (Hungr, 1985, Table 1). Avalanche deposits have relatively low Rb/Sr ratios that contrast with surrounding FDG sediments. Avalanche disturbanc es are the most poorly sorted of the three disturbances which includ ed floods and density currents (Figure 6a, Table 4). Distributions of snow avalanche events are identifiable by the presence sands and their grain size distributions are more leptokurtic and coarsely skewed than background autochthonous sediments differing from both floods and density currents (Table 1, Table 4) Flood events can be visually inferred by higher mineroge nic composition (mica and quartz) and increased macrofossil concentration. They are also associated with larger me an grain size (very fine sand to fine sand) (Table 3), but differ from avalanche deposits as they are better sorted (lower sorting values). F lood event distributions are more leptokurtic and coar sely skewed (Table 3, Table 4) than avalanche distributions, but lack the fat tail of larger grain sizes (Figure 7). This suggests different transport and deposition of water pulses into a lake versus the energy and velocity of a snow avalanche depositing material at the bottom of the run out zone. Density c urrent s appear to differ from avalanche events as material that have the largest mean grain size (very fine to fine sand) (Table 3, Table 4) and are the best sorted of the three geomorphic disturb ance events (Figure 7). Density currents or turbidities display more leptokurtic distributions than avalanche events which are most coarsely skewed ( Table 3, Table 4 ).

PAGE 72

63 REFERENCES Alestalo J. (1971) Dendrochronological interpretation of geomorphic processes. Fennia 105 p 1 140 Anderson, R., Smith, S., Lynch, A., and Geils, B. (2010). The pollen record of a 20 th century spruce beetle (Dendroctonus rufipennis) outbrea in a Colorado subalpine forest, USA. Forest Ecology and Management 260 p 448 455. Anderson, R., Allen, C., Toney, J., Jass, R., and Bair, A. (2008) Holocene vegetation and fire regimes in subalpine and mixed conifer forests southern Rocky Mountains, USA International Journal of Wildland Fire 17 p 96 114 Aplet, G., Laven, R., and Smith, F. (1988) Patterns of community dynamics in Colorado Engelma nn spruce subalpine fir forests. Ecological 69 : 2 p 312 319 Baggi, S., Schweizer J. (2009) Characteristics of wet snow avalanche activity: 20 years of observations from a high alpine valley (Dischma, Switzerland). Natural Hazards 50 p 97 108 Baker, W. (1992) The landscape ecology of large disturbances in the design and management of nature rese rves. Landscape Ecology 7 : 3 Barsch, D. (1977) Nature and importance of mass wasting by rock glaciers in alpine permafrost environments. Earth Surficial Processes 2 p 231 245 Beierle, B., Lamoureux, S., Cockburn, J., Spooner, I. (2002) A new method for visualizing particle size distributions. Paleolimnol 27 p 279 283 Blake, W., Walbrink, P., Doerr, S., Shakesby, R., and Humphreys, G. (2006) Magnetic enhancement in wildfire affected soil and its potential for sediment source ascription. Earth Surface Processes and Landforms 31 p 249 264 Prentice Hall Publishing Bocet, E., Poesen, J., and Rubio, J. (2006) Runof and soil loss under individual plants o a semi arid Mediterranean scrubland: Inluence o plan m orpholo and rainfall intensity. Earth Surface Processes and Landforms 31 p 536 549 Brugger, K. (2006) Late Pleistocene climate inferred from the reconstruction of the Taylor River glacier complex, so uthern Sawatch Range, Colorado. Geomorphology 75 p 318 329

PAGE 73

64 Brun, E., Martin, E., Simon, V., Gendre, C., Coleou, C. (1989) An Energy and Mass Model of Snow Cover Suitable for Op erational Avalanche Forecasting. Journal of Glaciology 35 : 121 p 333 342 Bui, E., Mazullo J., Wilding L. (1990) Using quartz grain size and shape analysis to distinguish between Aeolian and fluvial deposits in the Dallol Boss of Niger (West Africa). Earth Surface Processes and Landforms 14, p 157 166 Butler, D., Walsh, S. (1 990) Lithologic, Structural, and Topographic Influences on Snow Avalanche Path Location, Eastern Glacier National Park, Montana. Annals of the Ass ociation of American Geographers 80 : 3 p 362 378 Cannon, S., Kirkham, R., and Parise, M. Wildfire related debris flow initiation processes, Storm King Mountain, Colorado. Geomorphology 39 p 171 188 Cannon, S., and DeGraff J. The increasing wildfire and Post fire Debris Flow threat in western USA, and implications of consequences of climate change. Landslides Disaster Risk Reduction P 177 190 Carrara, P. (1979) The determination of snow avalanche frequency through tree ring analysis and histor ical records at Ophir, Colorado. The Geological Society of America 90 : 8, p 773 780 Chen, R., and Pagonis, V. (2011) Optically Stimulated Luminescense: A Simulation Approac h. John Wiley & Sons, Ltd Chichester, UK. Connell, J. (1978) Diversity in Tropical Rain Forests and Coral Reefs. Science 199 (4335) p 1302 1310 Cook, E., Woodhouse, C., Eakin, M., Meko, D., and Stahle, D. (2004). Long term aridity changes in the weste rn United States. Science 306 p 1015 1018. Corona, C., Saez, J., Stoffel, M., Rovera, G. (2012) Seven centuries of avalanche activity at Echalp (Queyras massif, southern French Al ps) as inferred from tree rings. The Holocene 23 : 2 p 292 304 Croudace I., Rindby, A., Rothwell, G. (2006) ITRAX: description and evalutation of a new multi function X Ray core scanner. The Geological Society 267 p 51 63 Dasch, E. (1969) Strontium isotopes in weathering profiles, deep sea sediments, and sedimentary rocks. Geochemica Cosmachimica Acta 33 p 1521 1552

PAGE 74

65 Dean, W. (1974) Determination of carbonate and organic matter in calcareous sediments by loss on ignit ion comparison to other methods. Journal of Sedimentary Petrology 44 p 242 248 Del Piore (2015) 4, 000 Years of Environmental Change in Central Colorado: A Paleoecological Perspective. University of Colorado Denver, p 1 117. Doesken, N., Pielke, R., and Bliss, O. (2003). Climate of Colorado. Climatography of the United States 60 http://climate.atmos.colostate.edu/climateofcolorado.php Fall, P. (1997a) Fire history and composition of the subalpine forest of western Colorado during the Holocene Journal of Biogeography 24 p 309 325 Finkenbinder, M., Abbott, M., Edwards, M., Langdon, C., Steinman, B., Finney, B. (2014) A 31,000 year record of paleoenvironmental and lake level change from Harding lake, Alaska, USA. Quaternary Science Reviews 87 p 98 113 Friedman, G. and Sanders, J. (1978) Principles of Sedimentology. Wiley, New York Garrow, H. (1982) A theoretical model of source and transport effects on grain size distributions. Eleventh International Congress on Sedimentology p 80 Gedye, S., Jones, R., TInner, W. Ammann, B. and Oldfield, F. (2000) The use of mineral ma gnetism in the reconstruction of fire history: a case study fr om Lago di Origlio, Swiss Alps. Paleogeography, Paleoclimatology, Palaecology 164 p 101 110 Gray, A., Pasternack, G., and Watson, E. (2010) Hydrogen peroxide treatement effects on the partic le size distribution of alluvial and marsh sediments. The Holocene 20 : 2 Gregory, K. and McIntosh, W. Paleoclimate and paleoelevation of the Oligocnene Pitch Pinnacle flora, Sawatch Range, Colorado. Geological Society of America Bulletin 108 : 5 p 545 561 Higuera, P., Briles, C., and Whitlock, C. (2014) Fire regime complacency and sensitivity to centennial through millennial scale climate change in Rocky Mountain su balpine forests, Colorado, USA. Journal of Ecology 102 p 1429 1441 Hungr, O. (1995) A Mod el of the runout analysis of rapid flow slides, debris flows, and avalanches Canadian Geotechnical Journal 32 p 610 623 Jin, Z., Cao, J., Wu, J. and Wang, S. (2006) A Rb/Sr record of catchment weathering response to Holocene climate change in Inner Mongolia. Earth Surface Processes and Landforms

PAGE 75

66 Johansen, M., Hakonson, T., Breshears, D. (2001) Post fire runoff and erosion from rainfall simulation: contrasting forests with shrublands and grasslands. Hydrological Pro cesses 15 p 2953 2965 Krumbein, W. (1938) Size frequency distribution of sediments and normal phi curve. Journal of Sedimentary Petrology 8 p 84 90 Kluth, C. and Coney, P. (1981) Plate tectoni cs of Ancestral Rocky Mountains. Geology Knox, J. (2000) Sensitivity of modern and Holocene floods to climate change. Quaternary Science Reviews 19 p 439 457 LaChapelle, E. IV. 1. Principles of Avalanche Forecasting Associate Committee on Geotechnical Research Lamb, M. and Mohrig D. (2009) Do hyperpycnal flow deposits record river flood dynamics? Geological Society of America 37 :12 p 1067 10170 Lowe, J., and Walker, M. (1997) Reconstructing Quaternary Environments. Longman London. 2 nd Ed. 446 Lubchenco, J. Plant Species Diver sity in a Marine Intertidal Community: Importance of Herbivore Food Preference and Algal Competitive Abilities. The American Naturalist 112 : 983, p 23 29 Luckman, B. (1977) The Geomorp hic Activity of Snow Avalanches. Geografiska Annaler 59 : 1 p p 31 48 Mann, M., Zhang, A., Hughes, M., Bradley, R., Miller, S., and Rutherford, S. (2008). Proxy based reconstructions of hemispheric and global surface temperature variations over the past two millennia. Proceedings of the National Academy of Sciences 105 p 13252 13257. Mast, J., Veblen, T., and Linhart, Y. (1998) Disturbance and climatic influences on age structure of Ponderosa Pine at the Pine/Grassland Ecotone, Colorado Front Range. Journal of Biogeography 25 : 4 p 743 755 Martin C., Pohl, M., Alewell, C., Korner, C., Rixen, C. (2010) Inerrill erosion a isturbed alpine sies: Effects of plant functional diversit y and vegetation cover. Basic and Applied Ecology 11 p 619 626 McLaren, P., and Bowles, D. (1985) The effects of sediment transport on grain size distributions. Journal of Sedimentary Petrology 55: 4 p 457 470

PAGE 76

67 Meiburg, E. and Kneller, B. (2009) Turbidity Currents and Their Deposits. Annual Review Fluid Mechanics 42 p 135 156 Meusburger, K., Konz, N., Schaub, M., Alewell, C. (2010) Soil erosion modelled with USLE and PERSERA using QuickBird derived vegetation parameters in an al pine catchment. International Journal of Applied Earth Observation and Geoinformation 12 p 208 215 Molnar, P. (2004) Late Cenozoic Increase in Accumulatio n Rates of Terrestrial Sediment: How Might Climate Change Have Affected Erosion Rates? Annual Review Earth Planet Science 32 P. 67 89 Mulder, T., Migeon, S., Savoye, B., Faugeres, J. (2001) Inversley graded turbidite sequences in the deep Meditteranean : a record of deposits from flood generated turbidity currents? Geo Marine Letters 21 :2 p 86 93 Paine, R., Tegner, M., and Johnson, E. (1998) Compounded Perturbati ons Yield Ecological Surprises. Ecosystems P 535 545 Pazzaglia, F. and Kelley, S. (1998) Large scale geomorphology and fission track thermochronology in topographic and exhumation reconstructions of the Southern Rocky Mountains. Rocky Mountain Geology. 33 : 2 p 229 257 Refsnider, K., and Brugger, K. (2007) Rock Glaciers in Central Colorado, U.S.A., as Indica tors of Holocene Climate Change. Arctic, Antarctic, and Alpine Research 39 : 1 p 127 136 Ryan, S., Bishop, E., and Daniels, M. (2014) Influence of large wood on channel morphology and sediment storage in headwater mountain streams, Frase r Experimental Forest, Colorado. Geomorphology 217 :15 p 73 88 Schillereff, D., Chiverrell, R., Macdonald, N., Hooke, J. (2014) Flood stratigraphies in lake sediments: A review. Earth Science Reviews P 1 63 Schoennagel, T., Veblen, T., Romme, W., Sibold, J., and Cook, E. (2005) ENSO and PDO variability affect drought induced fire occurrence in Ro cky Mountain subalpine forests. http://spot.colorado.edu/~schoenna/images/ Schoennagel2005EcoApps.pdf Schweizer, J. (1999) Review of dry snow slab avalan che release. Cold Regions Science and Technology 30 : 1 3, p 43 57

PAGE 77

68 Sherriff, R., Veblen, T., and Sibold, J. (2001) Fire History in high el evation subalpine forests in the Ecoscience 8 p 369 380 Stohlgren, T., Owen, A., and Lee, M. (2000) Monitoring shifts in plant diversity in response to climate c hange: a method for landscapes. Biodiversity and Conservation 9 p 65 86 Szeic, J., Zeeb B., Bennett, K., and Smol, J. (1998) High resolution paleoecological analysis of recent dist urbance in a southern Chilean Nothofagus forest. 20 p 235 252 Tweto, O. (1975) Laramide (Late Cretaceous Early Tertiary) Orogeny in the Southern Rocky Mountains. Geological Society of America Memoirs 144 p 1 44 Tweto, O. (1987) Rock units of the Precambrian basement in Colorado U.S. Geological Survey, professional paper. 1321 Va sskog, K., Nesje, A., Nage, E. (2011) A Holocene record of snow avalanche and flood activity reconstructed from a lacustrine sedientary sequenc e in Oldevatnet, western Norway. The Holocene Vasskog, K., Paasche, O., N esje, A., Boyle, J., Birks, H. (2012) A new approach for reconstructing gla cier variability based on lake sediments recording i nput from more than one glacier. The Holocene P 192 204 Vivoni, E.; Bowman, R., Wyckoff, R., Jakubowski, R., Richards, K. (2006) Analysis of a monsoon flood event in an ephemeral tributary and it s down stream hydrologic effects. Water Resources Research 42 : 3 Watson, E., Pasternack, G., Gray, A., Goni, M. and Woolfolk, A. (2013) Particle size characterization of historic sediment deposition from a closed estua rine lagoon, Central California. Estuarine, Coastal and Shelf Science 126 p 23 33 Wells, W. The effects of fire on the generation of debris flows in southern California. Geological Society of America 7 p 105 114 Weltje, G., von Eynatten, H (2004) Quanitative provenance analysis of sediments: review and outlook. Sedimentary Geology 171 p 1 11 White, P., and Pickett, S. (1985) White Academic Press p 3 13

PAGE 78

69 Whitlock, C., Higuera, P., McWethy, D., and Bri les, C. (2010) Paleoecological perspectives on fire ecology: revis iting the fire regimes concept. The Open Ecology Journal 3 p 6 23 Wilhelm, B., Arnaud, F., Sabatier, P. (2013) Paleoflood activity and climate change over the last 1400 years recorded by lake sediments in the north west European Alps. Jour nal of Quaternary Science. p 189 199 Wondzell, S., and King, J. (2003) Postfire erosional processes in the Pacific Northwest and Rocky Mountain regions. Forest Ecology and Management 178 p 75 87 Zhang, X. (2014 a ) Lacustrine Turbidites from Tropical African Lakes as Indicators of Hydrologic and Climatic Changes. Syracuse University Dissertation ALL p 182 Zhang, X., Scholz, C., Hecky, R., Wood, D., Zal, H., and Ebinger, C. (2014) Clim atic control of the late Quaternary turbidite sedimentology of Lake Kivu, East Africa: Implications for deep mixing and geologic hazards. Geology 42 :2 p 811 814

PAGE 79

70 A PPENDIX A MAGNETIC SUSCEPTIBILITY FOR COTTONWOOD LAKE (CWL15)

PAGE 80

71 Depth (cm) MS (cgs) 1 2.04E 01 2 2.11E+00 3 2.17E+00 4 1.04E+00 5 1.80E+00 6 2.48E+00 7 4.42E 01 8 3.05E+00 9 2.56E+00 10 3.02E 01 11 2.56E+00 12 2.58E+00 13 2.67E+00 14 1.47E+00 15 1.38E+00 16 1.17E+00 17 1.36E+00 18 1.14E+00 19 7.89E 01 20 1.08E+00 21 1.98E+00 22 2.97E+00 23 1.42E+00 24 2.23E+00 25 1.67E+00 26 1.91E+00 27 1.76E+00

PAGE 81

72 28 1.31E+00 29 2.20E+00 30 1.57E+00 31 2.33E+00 32 1.93E+00 33 1.12E+00 34 1.34E+00 35 1.04E+00 36 1.33E+00 37 1.59E+00 38 1.27E+00 39 1.45+00 40 1.07E+00 41 1.35E+00 42 1.57E+00 43 2.21E+00 44 1.61E+00 45 1.60E+00 46 1.20E+00 47 3.77E 01 48 1.20E+00 49 8.37E 01 50 1.05E+00 51 7.75E 01 52 5.40E 01 53 5.82E 01 54 6.12E 01 55 5.06E 01 56 5.44E 01

PAGE 82

73 57 5.23E 01 58 8.18E 01 59 4.88E 01 60 4.08E 01 61 5.42E 01 62 3.95E 01 63 6.67E 01 64 1.08E+00 65 8.35E 01 66 6.17E 01 67 7.44E 01 68 8.37E 01 69 6.80E 01 70 7.78E 01 71 2.50E+00 72 2.50E+00 73 1.53E+00 74 1.47E+00 75 9.15E 01 76 7.02E 01 77 6.65E 01 78 6.20E 01 79 5.40E 01 80 4.81E 01 81 1.11E+00 82 8.57E 01 83 4.02E 01 84 7.52E 01 85 9.62E 01

PAGE 83

74 86 1.01E+00 87 7.94E 01 88 9.68E 01 89 1.33E+00

PAGE 84

75 APPENDIX B MAGNETIC SUSCEPTIBILITY FOR MIRROR LAKE (MLL15EBC)

PAGE 85

76 Depth (cm) MS (cgs) 53 1.17E 01 54 1.30E 01 55 1.37E 01 56 1.77E 01 57 1.45E 01 58 1.53E 01 59 1.35E 01 60 7.59E 02 61 1.20E 01 62 1.39E 01 63 1.71E 01 64 1.69E 01 65 2.38E 01 66 1.45E 01 67 2.05E 01 68 1.71E 01 69 1.82E 01 70 1.11E 01 71 2.56E 01 72 2.68E 01 73 5.94E 01 74 1.75E01 75 1.95E 01 76 1.42E 01 77 1.74E 01 78 2.13E 01 79 2.22E 01

PAGE 86

77 80 1.23E 01 81 6.53E 02 82 1.83E 01 83 1.81E 01 84 9.96E 02 85 1.51E 01 86 1.66E 01 87 1.97E 01 88 1.32E 01 89 1.01E 01 90 4.63E 02 91 5.73E 02 92 1.69E 01 93 1.47E 01 94 2.06E 01 95 1.07E 01 96 9.53E 02 97 1.20E 01

PAGE 87

78 APPENDIX C MAGNETIC SUSCEPTIBILITY FOR MIRROR LAKE (ML131A)

PAGE 88

79 Depth (cm) MS (cgs) 6.5 1.82E+00 7 2.86E+00 7.5 1.19E+00 8 1.06E+00 8.5 3.18E+00 9 3.41E+00 9.5 1.48E+00 10 6.14E 01 10.5 1.63E+00 11 2.52E+00 11.5 2.02E+00 12 2.10E+00 12.5 2.62E+00 13 2.48E+00 13.5 2.36E+00 14 2.18E+00 14.5 2.51E+00 15 1.54E+00 15.5 1.83E+00 16 3.44E+00 16.5 4.13E+00 17 4.57E+00 17.5 2.63E+00 18 1.97E+00 18.5 1.82E+00 19 3.00E+00 19.5 2.59E+00

PAGE 89

80 20 2.05E+00 20.5 2.95E+00 21 1.08E+00 21.5 1.18E+00 22 1.66E+00 22.5 2.15E+00 23 1.74E+00 23.5 1.14E+00 24 1.27E+00 24.5 3.45E+00 25 4.59E+00 25.5 3.83E+00 26 1.92E+00 26.5 1.75E+00 27 3.09E+00 27.5 1.25E+00 28 1.08E+00 28.5 1.99E+00 29 2.51E+00 29.5 2.10E+00 30 2.05E+00 30.5 2.03E+00 31 8.06E 01 31.5 1.09E+00 32 1.24E+00 32.5 2.02E+00 33 2.31E+00 33.5 1.66E+00 34 3.13E+00

PAGE 90

81 34.5 1.33E+00 35 2.19E+00 35.5 2.90E+00 36 3.72E+00 36.5 2.24E+00 37 1.66E+00 37.5 2.49E+00 38 2.87E+00 38.5 4.26E+00 39 4.08E+00 39.5 3.62E+00 40 2.88E+00 40.5 4.60E+00 41 4.95E+00 41.5 3.21E+00 42 2.31E+00 42.5 2.58E+00 43 2.56E+00 43.5 1.40E+00 44 9.28E 01 44.5 2.09E+00 45 1.82E+00 45.5 3.63E+00 46 1.47E+00 46.5 2.04E+00 47 3.39E+00 47.5 3.43E+00 48 4.95E+00 48.5 5.10E+00

PAGE 91

82 49 3.28E+00 49.5 2.18E+00 50 2.52E+00 50.5 3.38E+00 51 3.73E+00 51.5 1.08E+00 52 8.42E+00 52.5 3.83E+00 53 4.54E+00 53.5 3.36E+00 54 1.79E+00 54.5 3.02E+00 55 3.28E+00 55.5 3.84E+00 56 4.30E+00 56.5 3.27E+00 57 3.34E+00 57.5 3.07E+00 58 2.87E+00 58.5 3.14E+00 59 3.89E+00 60 2.04E+00 60.5 9.31E 01 61 3.38E+00 61.5 5.00E+00 62 3.20E+00 62.5 1.73E+00 63 1.63E+00 63.5 2.32E+00

PAGE 92

83 64 9.11E 01 64.5 2.68E+00 65 4.30E+00 65.5 7.00E 01 66 6.93E 01 66.5 1.23E+00 67 1.76E+00 67.5 1.31E+00 68 1.40E+00 68.5 2.40E+00 69 2.03E+00 69.5 2.10E+00 70 1.48E+00 70.5 1.37E+00 71 1.31E+00 71.5 1.51E+00 72 1.15E+00 72.5 1.82E+00 73 1.24E+00 73.5 1.19E+00 74 2.18E+00 74.5 2.72E+00 75 4.10E+00 75.5 4.09E+00 76 2.79E+00 76.5 2.26E+00 77 1.42E+00 77.5 1.68E+00 78 1.47E+00

PAGE 93

84 78.5 1.38E+00 79 2.37E+00 79.5 1.37E+00 80 1.55E+00 80.5 1.30E+00 81 9.22E 01 81.5 5.70E 01 82 5.44E 01 82.5 5.75E 01 83 7.65E 01 83.5 2.13E+00 84 2.22E+00 84.5 2.86E+00 85 2.58E+00 85.5 2.26E+00 86 1.26E+00 86.5 1.61E+00 87 2.49E+00 87.5 2.46E+00 88 2.38E+00 88.5 2.71E+00 89 1.99E+00 89.5 2.10E+00 90 1.66E+00 90.5 2.75E+00 91 1.91E+00 91.5 1.15E+00 92 2.21E+00 92.5 2.03E+00

PAGE 94

85 93 3.63E+00 93.5 3.32E+00 94 2.46E+00 94.5 2.21E+00 95 2.74E+00 95.5 2.50E+00 96 2.10E+00 96.5 2.27E+00 97 1.40E+00 97.5 2.28E+00 98 1.95E+00 98.5 2.06E+00 99 1.01E+00 99.5 1.05E+00 100 2.09E+00 100.5 3.14E+00 101 1.98E+00 101.5 1.81E+00 102 1.61E+00 102.5 1.80E+00 103 1.56E+00 103.5 2.05E+00 104 1.85E+00 104.5 2.88E+00 105 3.53E+00 105.5 3.56E+00 106 3.34E+00 106.5 1.71E+00 107 2.08E+00

PAGE 95

86 107.5 2.55E+00 108 2.63E+00 108.5 1.96E+00 109 9.43E 01 109.5 1.31E+00 110 1.01E+00 110.5 1.79E+00 111 1.49E+00 111.5 2.51E+00 112 1.92E+00 112.5 1.67E+00 113 2.33E+00 113.5 1.91E+00 114 1.91E+00 114.5 1.86E+00 115 1.88E+00 115.5 1.98E+00 116 2.95E+00 116.5 2.25E+00 117 7.77E 01 117.5 1.03E+00 118 2.11E+00 118.5 1.78E+00 119 1.81E+00 119.5 1.72E+00 120 1.21E+00 120.5 6.71E 01 121 1.99E+00 121.5 2.08E+00

PAGE 96

87 122 2.94E+00 122.5 2.50E+00 123 2.64E+00 123.5 2.12E+00 124 2.41E+00 124.5 1.18E+00 125 1.17E+00 125.5 1.81E+00 126 1.95E+00 126.5 1.54E+00 127 1.87E+00 127.5 1.25E+00 128 1.62E+00 128.5 1.42E+00 129 8.81E 01 129.5 1.21E+00 130 1.38E+00 130.5 7.81E 01 131 1.15E+00 131.5 1.22E+00 132 7.34E 01 132.5 1.12E+00 133 1.19E+00 133.5 1.62E+00 134 1.74E+00 134.5 1.59E+00 135 1.27E+00 135.5 1.31E+00 136 9.80E 01

PAGE 97

88 136.5 1.01E+00 137 1.06E+00 137.5 1.03E+00 138 1.63E+00 138.5 1.80E+00 139 2.27E+00 139.5 2.93E+00 140 1.99E+00 140.5 1.24E+00 141 1.93E+00 141.5 1.06E+00 142 1.90E+00 142.5 1.16E+00 143 2.14E+00 143.5 2.11E+00 144 2.02E+00 144.5 2.90E+00 145 2.11E+00 145.5 2.02E+00 146 2.90E+00 146.5 2.47E+00 147 2.30E+00 147.5 3.14E+00 148 2.42E+00 148.5 1.67E+00 149 3.97E+00 149.5 1.80E+00 150 9.43E 01 150.5 1.31E+00

PAGE 98

89 151 1.58E+00 151.5 3.04E+00 152 2.75E+00 152.5 2.95E+00 153 2.63E+00 153.5 3.09E+00 154 2.76E+00 154.5 2.94E+00 155 3.13E+00 155.5 2.90E+00 156 3.14E+00 156.5 4.02E+00 157 3.81E+00 157.5 2.33E+00 158 2.20E+00 158.5 2.38E+00 159 2.14E+00 159.5 8.52E 01 160 7.37E 01 160.5 8.04E 01 161 1.18E+00 161.5 2.39E+00 162 1.99E+00 162.5 2.12E+00 163 2.10E+00 163.5 2.85E+00 164 2.29E+00 164.5 1.56E+00 165 1.84E+00

PAGE 99

90 165.5 2.79E+00 166 2.23E+00 166.5 1.62E+00 167 1.30E+00 167.5 2.23E+00 168 2.00E+00 168.5 2.43E+00 169 1.74E+00 169.5 1.39E+00 170 1.92E+00 170.5 2.44E+00 171 1.77E+00 171.5 2.09E+00 172 2.21E+00 172.5 2.23E+00 173 2.39E+00 173.5 2.27E+00 174 2.56E+00 174.5 1.86E+00 175 1.78E+00 175.5 2.69E+00 176 2.53E+00 176.5 1.91E+00 177 1.70E+00 177.5 2.14E+00 178 1.72E+00 178.5 8.99E 01 179 9.13E 01 179.5 6.52E 01

PAGE 100

91 180 8.04E 01 180.5 2.14E+00 181 2.09E+00 181.5 1.94E+00 182 7.71E 01 182.5 1.35E+00 183 2.36E+00 183.5 1.09E+00 184 1.14E+00 184.5 1.69E+00 185 1.82E+00 185.5 2.52E+00 186 1.88E+00 186.5 1.58E+00 187 2.23E+00 187.5 1.95E+00 188 2.52E+00 188.5 2.69E+00 189 2.55E+00 189.5 2.98E+00 190 3.17E+00 190.5 3.92E+00 191 3.13E+00 191.5 3.00E+00 192 3.51E+00 192.5 5.02E+00 193 3.38E+00 193.5 1.59E+00 194 2.14E+00

PAGE 101

92 194.5 2.62E+00 195 2.01E+00 195.5 2.10E+00 196 1.28E+00 196.5 2.01E+00 197 2.37E+00 197.5 1.50E+00 198 1.35E+00 198.5 2.73E+00 199 2.61E+00 199.5 2.12E+00 200 2.25E+00 200.5 1.59E+00

PAGE 102

93 APPENDIX D LOSS ON IGNITION FOR COTTONWOOD LAKE (CWL15)

PAGE 103

94 Depth (cm) % Bulk Density % Organics % Carbonates 1 48% 8.841% 2% 2 47% 8.966% 2% 4 48% 8.853% 2% 6 49% 9.441% 2% 8 50% 9.858% 1% 10 48% 9.078% 2% 12 50% 9.601% 2% 14 53% 9.407% 2% 16 51% 8.768% 2% 18 47% 8.726% 1% 20 48% 7.946% 2% 22 50% 9.342% 2% 24 52% 9.788% 2% 25 47% 8.021% 2% 28 58% 12.407% 2% 29 55% 10.299% 1% 30 57% 12.021% 2% 32 48% 11.581% 1% 33 50% 10.920% 2% 35 50% 11.062% 1% 37 47% 10.481% 1% 38 49% 10.519% 1% 39 48% 10.248% 1% 42 45% 9.389% 1% 44 50% 11.191% 1% 46 46% 9.449% 2% 48 48% 11.968% 0%

PAGE 104

95 50 59% 12.569% 2% 52 62% 12.840% 3% 54 64% 14.014% 4% 56 61% 12.607% 3% 58 61% 12.373% 3% 59 63% 15.625% 2% 62 46% 8.178% 1% 65 59% 14.076% 2% 68 58% 12.464% 1% 69 48% 8.587% 1% 70 48% 11.794% 0% 71 49% 11.807% 1% 72 57% 18.416% 0% 73 49% 13.284% 1% 74 47% 10.224% 1% 75 47% 11.130% 1% 76 49% 11.745% 1% 77 48% 11.181% 0% 78 54% 14.079% 1% 79 61% 22.765% 0% 80 53% 14.280% 1% 83 49% 11.093% 1% 85 53% 14.789% 0% 87 45% 9.657% 1% 89 50% 12.317% 1% 91 45% 9.530% 1% 93 48% 11.703% 1% 95 40% 8.830% 1% 97 41% 8.914% 1%

PAGE 105

96 APPENDIX E LOSS ON IGNITION FOR MIRROR LAKE (ML15EBC)

PAGE 106

97 Depth (cm) Bulk Density (g/cm3) % Organics % Mineral Residue 53 0.4754 12.80417 1 23.9926205 55 0.3055 11.16107 9 29.3059347 57 0.4808 10.62326 5 30.0840239 7 59 0.2568 7.747543 4 38.3596264 60 0.3915 11.40042 2 26.2588895 61 0.4003 14.76147 6 15.5092705 7 63 0.2929 14.19122 0 16.0073396 65 0.5444 13.26037 9 17.8780921 9 67 0.3111 17.08945 2 8.88467131 69 0.2293 17.93286 2 10.2970312 9 70 0.1739 13.83595 6 21.4828211 3 72 0.3152 10.33834 5 30.812641 74 0.3156 13.52169 5 19.7759265 9 75 0.3157 16.45338 2 12.066833 76 0.3519 17.50339 2 4.92132986 2 77 0.4396 16.81643 1 11.8603451 2 78 0.3306 11.44486 27.9326219

PAGE 107

98 6 3 79 0.2763 15.29010 2 12.2196455 3 81 0.3579 13.72997 7 18.2455519 5 83 0.1545 9.570650 0 34.0237804 8 84 0.2181 7.589478 2 38.7467011 6 86 0.3571 15.05811 0 14.0498759 5 88 0.185 12.78305 3 23.1014164 2 89 0.3297 12.50566 3 20.9981557 8 91 0.2842 14.52420 7 18.8977316 1 93 0.2661 15.20467 8 18.1985585 2 95 0.3089 19.06424 5 4.98781613 8 97 0.2935 21.89254 2 0.28036795 7 99 0.1385 14.51776 6 20.5792525

PAGE 108

99 APPENDIX F LOSS ON IGNITION FOR MIRROR LAKE (ML131A)

PAGE 109

100 Depth (cm) % Organics 7 9.799554566 9 8.285052144 11 7.65294772 13 8.87060916 15 8.348932984 17 7.281553398 19 10.50993181 21 10.63084112 23 16.96145125 25 7.728842832 27 7.70008668 29 8.440830062 31 7.824331146 33 8.648648649 35 8.053818457 37 8.469635628 39 5.912921348 41 6.078064237 43 4.049466537 45 8.214793741 47 5.878510777 49 4.098657962 51 4.403669725 53 3.086940793 55 4.688083676 57 5.404938698 59 7.361142092 61 9.906588004

PAGE 110

101 63 13.03566846 65 5.414319088 67 10.31055901 70 8.149386845 72 7.969478593 74 7.547662575 76 3.629283489 78 15.05657093 80 2.241359644 82 21.93158954 84 6.888694128 86 8.36071559 88 6.927411336 90 6.946125908 92 7.332892124 94 6.490541422 96 6.238554511 98 7.903968428 100 9.499389499 102 8.420320111 104 12.53996448 105 9.084604716 109 10.31092214 111 17.88964733 113 9.353574927 115 7.623196297 117 15.16748212 119 15.08862949 121 17.01846966

PAGE 111

102 123 10.55017515 125 17.61416589 127 17.70458055 129 18.5827552 131 21.54490461 133 12.58871871 135 14.76833977 137 11.39021221 139 11.27078629 141 11.99098024 143 23.46232179 145 12.61131859 147 7.791902072 149 9.451385117 151 9.884736291 153 9.9756691 155 11.47859922 157 10.84538376 159 13.42369917 161 12.82843895 163 20.15957447 165 18.71838111 167 15.00796178 169 13.4164892 171 12.83482143 173 13.5529608 175 15.76197388 177 9.588268472 179 20.37566354

PAGE 112

103 181 12.039801 183 20.11652867 184.5 14.16234888 186.5 12.33339964 188.5 11.67072841 190.5 12.64573991 192.5 8.240582192 194.5 15.8278909 196.5 15.93701997 198.5 15.36341715 200.5 16.96252465

PAGE 113

104 APPENDIX G GEOCHEMISTRY FOR COTTONWOOD LAKE (CWL15)

PAGE 114

105 Depth Rb Sr Rb/Sr 0 0.0023 0.0057 0.403508772 2 0.01 0.0199 0.502512563 4 0.0014 0.0041 0.341463415 8 0 0.0025 0 12 0.0009 0.0025 0.36 16 0.0034 0.0065 0.523076923 20 0.0079 0.0175 0.451428571 22 0.0021 0.005 0.42 24 0.0098 0.0237 0.41350211 25 0.0076 0.017 0.447058824 28 0 0.0025 0 30 0.0093 0.0181 0.513812155 33 0.001 0.003 0.333333333 37 0.0014 0.0036 0.388888889 39 0.0058 0.0156 0.371794872 42 0 0.003 0 46 0.0037 0.0097 0.381443299 50 0.0034 0.0068 0.5 52 0.0029 0.0052 0.557692308 59 0.0039 0.0068 0.573529412 62 0.0066 0.0131 0.503816794 65 0.0028 0.0056 0.5 68 0.0016 0.0036 0.444444444 69 0.0021 0.0041 0.512195122 70 0.003 0.0065 0.461538462 71 0.0013 0.0029 0.448275862 72 0.0006 0.0019 0.315789474 73 0.0019 0.0045 0.422222222

PAGE 115

106 74 0.0035 0.0091 0.384615385 75 0.0037 0.0089 0.415730337 76 0.0035 0.0083 0.421686747 77 0.0013 0.0035 0.371428571 78 0.0004 0.0015 0.266666667 79 0.0007 0.0023 0.304347826 80 0.003 0.0075 0.4 83 0.0074 0.0168 0.44047619 85 0.0033 0.007 0.471428571 87 0.0044 0.0102 0.431372549 89 0.0044 0.0111 0.396396396 91 0.0045 0.0109 0.412844037 93 0.0051 0.0141 0.361702128 95 0.0042 0.0084 0.5 97 0.0044 0.0108 0.407407407

PAGE 116

107 APPENDIX H GEOCHEMISTRY FOR MIRROR LAKE (M L131A)

PAGE 117

108 Depth Rb Sr Rb/Sr 53 0.0016 0.0038 0.421052632 54 0.0022 0.0033 0.666666667 55 0.0014 0.0033 0.424242424 56 0.0012 0.0023 0.52173913 57 0.0021 0.0052 0.403846154 58 0.0024 0.004 0.6 59 0.002 0.0042 0.476190476 60 0.0012 0.0026 0.461538462 61 0.0012 0.0026 0.461538462 63 0.0008 0.0018 0.444444444 65 0.0014 0.0035 0.4 67 0.0004 0.0016 0.25 69 0 0.0014 0 70 0.0019 0.0037 0.513513514 72 0.0006 0.0018 0.333333333 74 0.0006 0.0017 0.352941176 75 0.0007 0.0019 0.368421053 76 0.0007 0.0015 0.466666667 77 0.0009 0.0019 0.473684211 78 0.0011 0.0026 0.423076923 79 0.0004 0.0014 0.285714286 81 0.0017 0.0035 0.485714286 83 0.0007 0.0018 0.388888889 84 0.0015 0.0033 0.454545455 86 0.002 0.0036 0.555555556 88 0.0014 0.0031 0.451612903 89 0.0026 0.0047 0.553191489 91 0.0022 0.0038 0.578947368

PAGE 118

109 93 0.0018 0.0037 0.486486486 95 0.0013 0.0027 0.481481481 97 0.0015 0.0027 0.555555556 99 0.0013 0.0025 0.52 101 0.0012 0.0026 0.461538462 103 0.0012 0.0026 0.461538462

PAGE 119

110 APPENDIX I PARTICLE SIZE ANALYSIS FOR COTTONWOOD LAKE (CWL15)

PAGE 120

111 Depth %clay %silt %sand Grain Size Mean (m) Sorting (m) Skewness (m) Kurtosis (m) 1 3.8 38.9 57.4 54.52 3.746 0.471 1.114 2 5.3 51.7 43 37.18 4.197 0.356 0.973 3 5.7 53.4 40.9 34.07 4.119 0.383 0.960 4 5.2 49.6 45.1 38.34 4.089 0.409 0.984 5 4.9 49.5 45.6 39.41 3.933 0.420 1.019 6 4.8 48.3 46.9 40.75 4.040 0.407 1.013 7 4.2 43.6 52.3 47.58 3.829 0.444 1.068 8 3.9 40.4 55.7 52.10 3.738 0.468 1.102 9 4.3 42.9 52.8 48.48 3.987 0.438 1.016 10 3.4 34 62.6 63.47 3.607 0.498 1.199 12 6.4 57.4 36.2 29.84 4.443 0.284 0.919 14 5.6 50.4 44 37.19 4.469 0.345 0.932 16 5.4 51 43.5 37.02 4.267 0.365 0.968 18 3.2 31.3 65.6 72.30 3.630 0.480 1.227 20 3.9 38.5 57.7 55.48 3.738 0.468 1.143 22 5.6 51.1 43.3 36.00 4.379 0.358 0.934 23 3.8 36.1 60.1 58.57 3.885 0.496 1.099 25 3.3 31.8 64.9 68.45 3.616 0.511 1.212 27 4.1 42 53.9 51.73 4.212 0.401 0.991 28 3.9 39.7 65.5 54.57 3.919 0.447 1.050 29 4.4 45.6 50 45.00 3.989 0.413 1.017 30 5.8 55.3 38.9 32.85 4.200 0.344 0.950 31 5.6 52 42.4 36.27 4.439 0.320 0.947 32 4.1 40.5 55.4 52.67 4.058 0.436 1.046 34 3.9 37 59.1 58.68 3.970 0.461 1.097 35 2.6 26.5 70.9 84.70 3.066 0.459 1.490

PAGE 121

112 37 3.8 36.9 59.4 60.89 3.990 0.434 1.113 39 1.8 21.2 76.9 105.2 2.722 0.417 1.559 40 2.3 23.5 74.1 94.92 2.921 0.455 1.539 41 2.4 25.8 71.8 88.54 2.924 0.441 1.487 43 1.9 19.6 78.5 111.3 2.709 0.443 1.601 45 1.4 20.4 78.2 108.8 2.472 0.359 1.488 47 3.6 36 60.3 60.62 3.662 0.469 1.194 49 7.2 62 30.8 25.21 4.649 0.168 0.904 51 8.7 70.1 21.2 18.52 4.496 0.119 0.960 53 9.2 74.8 16 15.49 4.185 0.102 0.981 55 10.2 78.9 10.9 13.01 3.890 0.120 0.998 57 8.7 74.8 16.4 16.75 4.036 0.173 0.956 58 9.1 75.8 15.1 15.65 4.049 0.139 0.989 59 8.9 76 15.1 15.82 4.061 0.128 1.034 60 7.1 62 30.9 25.57 4.752 0.125 0.910 61 5.7 53.3 41 33.62 4.509 0.268 0.877 62 7.4 65.3 27.4 23.42 4.532 0.165 0.941 64 6.5 59.4 34 28.27 4.565 0.223 0.914 66 6.4 54.7 38.8 31.34 5.108 0.178 0.871 68 2.6 25.5 71.9 101.5 4.278 0.165 1.779 69 1.6 19.4 79 110.1 2.484 0.389 1.565 70 2.1 23 74.9 98.91 2.922 0.424 1.536 71 1.3 16.6 82 129.7 2.479 0.331 1.471 72 1.5 19.5 79 115.9 2.581 0.369 1.469 73 2.1 22.9 75 103.9 2.994 0.386 1.474 74 1.7 19.2 79.1 119.1 2.730 0.388 1.497 75 2.7 27.8 69.5 84.66 3.528 0.455 1.317 76 1.2 15.4 83.4 138.8 2.504 0.318 1.509 77 2.4 26.6 71.1 89.02 3.121 0.435 1.398

PAGE 122

113 78 1 11.6 87.4 172.4 2.339 0.348 1.486 79 2.4 24.8 72.8 91.39 3.169 0.452 1.479 80 1.8 17.9 80.4 124.8 2.771 0.434 1.622 81 2 19.8 78.2 117.0 2.958 0.457 1.531 83 3.1 28 68.9 87.18 3.884 0.476 1.215 85 2.8 26.1 71.1 94.37 3.869 0.474 1.260 87 3.1 28.1 68.8 82.70 3.907 0.506 1.212 89 2.6 23.3 74.1 103.8 3.555 0.483 1.374 91 2.4 21.7 76 112.6 3.338 0.467 1.427 93 1.6 14.3 84.1 148.0 2.592 0.423 1.748 95 3.2 31.3 65.5 75.70 3.650 0.443 1.233 97 3.7 33.8 62.4 68.39 4.269 0.450 1.085

PAGE 123

114 APPENDIX J PARTICLE SIZE ANALYSIS FOR MIRROR LAKE (ML15EBC)

PAGE 124

115 Depth %clay %silt %sand Grain Size Mean (m) Sorting (m) Skewness (m) Kurtosis (m) 53 10.53112131 81.534633 7.934261757 13.78 3.754 0.283 1.040 55 8.880570372 78.201939 12.91748018 17.60 3.756 0.320 1.063 56 8.31969133 74.493523 17.18679633 19.38 3.845 0.312 1.014 57 8.01217973 77.188979 14.79886631 19.50 3.648 0.332 1.084 59 7.319625929 80.203903 12.47645539 20.63 3.389 0.399 1.188 60 8.44752091 79.229309 12.32317861 17.61 3.645 0.312 1.024 61 9.370920756 81.195348 9.433750383 15.50 3.666 0.296 1.004 62 11.5757595 80.742441 7.681788146 12.55 3.865 0.230 0.962 64 8.331852919 75.794878 15.87325572 19.24 3.756 0.360 0.981 68 12.07825712 81.92124 6.000514681 11.53 3.764 0.222 0.971 69 10.67082702 82.75232 6.57682104 13.13 3.659 0.271 0.986 70 9.87815231 79.461848 10.65998891 15.53 3.824 0.295 1.022 72 8.482981307 75.151725 16.36528939 18.61 3.828 0.302 0.984 74 10.6393933 78.158726 11.20187749 14.66 3.959 0.264 0.958 75 12.0434582 80.64962 7.306929408 12.21 3.900 0.243 0.991 76 12.10514856 81.10786 6.786962522 11.89 3.849 0.224 0.965 77 11.68986839 80.833047 7.477062709 12.41 3.884 0.233 0.975 78 10.39099187 80.02065 9.588352219 14.08 3.848 0.247 0.962 79 13.73612604 80.68137 5.582518571 10.21 3.889 0.181 0.952 80 12.68270055 78.8544 8.46291653 11.70 4.047 0.196 0.984 82 11.01843557 76.103451 12.87809631 14.59 4.149 0.231 0.962 84 3.837825492 42.609909 53.55226442 49.25 3.006 0.541 1.385 85 7.111587873 85.188881 7.699525791 18.34 3.168 0.363 1.161 87 11.75865084 79.003585 9.237778179 13.04 4.005 0.239 0.975 89 8.49985855 74.83417 16.66596811 18.98 3.857 0.314 1.007

PAGE 125

116 90 10.11740107 71.895312 17.98728746 17.26 4.287 0.255 0.955

PAGE 126

117 APPENDIX K PARTICLE SIZE ANALYSIS FOR MIRROR LAKE (ML131A)

PAGE 127

118 Depth %clay %silt %sand Grain Size Mean (m) Sorting (m) Skewness (m) Kurtosis (m) 1165 1.6% 21.20% 77.2% 153.4 3.156 0.429 0.962 1177.5 90.0% 10.50% 88.6% 230.6 2.492 0.506 0.959 1187.5 80.0% 12.50% 86.7% 225.1 2.620 0.430 0.782 1189.5 1.2% 20.70% 78.1% 114.9 2.363 0.086 1.172 1190 1.5% 30.10% 68.4% 82.08 2.311 0.297 1.375 1191 1.8% 55% 42.8% 48.31 2.413 0.296 1.027 1192 2.1% 49.20% 48.7% 55.84 2.875 0.208 1.110 1200 2.3% 63.50% 34.1% 43.27 2.496 0.195 1.231 1215 3.3% 43.10% 53.6% 58.29 3.707 0.235 1.319 1227 2.6% 56.60% 40.8% 47.56 3.010 0.168 1.333 1228 1.8% 26.10% 72.1% 93.52 2.571 0.244 1.442 1229.5 1.9% 35.50% 62.5% 84.82 2.949 0.094 1.265 1230.5 3.2% 34.90% 61.9% 77.81 4.797 0.288 0.948 1255 2.6% 60.90% 36.5% 40.72 2.705 0.333 1.113 1266 2.6% 32.70% 64.7% 78.34 3.244 0.267 1.458 1267 2.9% 62.20% 34.8% 39.57 2.584 0.409 1.201 1268 2.2% 37.50% 60.3% 73.82 2.863 0.165 1.525 1290 2.7% 64.80% 32.4% 41.65 2.550 0.278 1.376 1335 4.7% 82.80% 12.5% 23.45 2.837 0.286 1.271