A GEOGRAPHIC INFORMATION SYSTEMS ANALYSIS OF
ROCKFALL HAZARDS IN CLEAR CREEK CANYON, JEFFERSON
AND CLEAR CREEK COUNTIES, COLORADO
B.S., University of Colorado at Denver, 1998
A thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Basic Science
Melissa J. Crane
This thesis for the Master of Basic Science
Melissa J. Crane
has been approved
Crane, Melissa J. (M.B.S.)
A Geographic Information Systems Analysis of Rockfall Hazards in Clear
Creek Canyon, Jefferson and Clear Creek Counties, Colorado
Thesis directed by Professor John W. Wyckoff
The need to analyze the potential for rockfall hazards in Clear
Creek Canyon has increased due to the high volume of traffic on U.S.
Highway 6, which is a main route to the gambling towns of Black Hawk
and Central City. Clear Creek Canyon is located just west of Golden,
Colorado in the Precambrian core of the Front Range. The Colorado
Department of Transportation considers U.S. 6 in Clear Creek Canyon one
of the top rockfall hazard areas in the state due to the high volume of rock
debris and frequency of rockfalls in the canyon.
Numerous factors were identified that may contribute to rockfall
events including rock type, slope angle, aspect, vegetation, precipitation,
dip of foliation, and joint frequency. Once data were collected for each
factor, they were analyzed in ArcView and ARC/INFO to evaluate
potential rockfall hazards in Clear Creek Canyon. Statistical analyses
were also performed on each of these factors to identify which factors are
the most significant and are most highly correlated to rockfall events.
These results were then used to create a rockfall hazard map identifying
high, medium, and low risk areas.
Statistical analyses show that rock type is the most important factor
followed by slope angle and dip of foliation. The majority of the rockfalls
occur in biotite gneiss where slope is greater than 58 and where foliation
is dipping toward the road. Other factors looked at in this study do not
explain enough rockfall events to be considered significant.
Although, this study provides a good understanding of where the
potential for rockfall hazards are high and what contributes to those
hazards, there are two major factors that could improve this understanding.
First, is a closer look at precipitation amounts and trends. Second, is a
more detailed analysis of joint orientations.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
I would like to dedicate this thesis to my family and friends for their love
and support throughout my graduate studies.
I wish to thank my thesis committee for all their help, patience, and
support over the years. I would also like to thank the Colorado Scientific
Society for financially supporting my study.
1. Introduction................................... 1
1.1 Objectives..................................... 4
1.2 Scope.......................................... 5
1.3 The Clear Creek Canyon Study Area.............. 5
1.3.1 Topography..................................... 7
1.3.2 Climate........................................ 8
1.3.3 Geologic Setting................................ 11
2. Geology of Clear Creek Canyon.................... 14
2.1 Precambrian Rocks................................ 15
2.1.1 Biotite Gneiss................................... 15
2.1.2 Granite Gneiss................................. 16
2.1.3 Hornblende Gneiss and Amphibolite................ 17
2.2 Intrusions..................................... 18
2.2.1 Pegmatites..................................... 18
2.2.2 Lamprophyres................................... 19
2.3 Sedimentary Rocks.................................... 20
2.4 Surficial Deposits................................... 23
2.5 Structural Geology................................... 23
2.5.1 Folds................................................ 24
2.5.2 Faults............................................... 25
3. Presentation of Data................................. 27
3.1 Rockfall Location Data............................... 28
3.2 Rockfall Factors..................................... 28
3.2.1 Rock Type............................................ 29
3.2.2 Vegetation........................................... 29
3.2.3 Precipitation........................................ 31
3.2.4 Aspect............................................... 32
3.2.5 Slope............................................... 33
3.2.6 Foliation.......................................... 33
3.2.7 Joints............................................... 34
4. Methodology.......................................... 36
4.1 Integration of Data into GIS......................... 36
4.2 Analysis............................................. 37
4.3 Modeling............................................ 40
5. Results.............................................. 41
5.1 Numerical Results................................... 41
5.2 Graphical Results.................................... 43
5.3 Interpretations..................................... 45
5.3.1 Rock Type........................................... 45
5.3.2 Vegetation.......................................... 46
5.3.3 Precipitation....................................... 46
5.3.4 Aspect.............................................. 47
5.3.5 Slope.............................................. 47
5.3.6 Foliation........................................... 48
5.3.7 Joints.............................................. 48
6. Conclusions and Recommendations..................... 49
6.1 Conclusions......................................... 49
6.2 Recommendations..................................... 50
A. Spreadsheet......................................... 52
B. Maps................................................ 54
1-1 Colorado Map Showing Location of Study Area.... 3
1-2 Inset Map of Figure 1-1........................ 4
1-3 Quadrangle Map................................. 7
1-4 Average Monthly Precipitation..... .............. 10
1- 5 Average Monthly Temperatures..................... 10
2- 1 Stratigraphic Column............................. 22
3- 1 Photo of Joint Sets.............................. 35
B-l Geology Map...................................... 55
B-2 3D Geology Map................................... 56
B-3 Folds.......................................... 57
B-4 Faults........................................... 58
B-5 Vegetation Map................................... 59
B-6 3D Vegetation Map................................ 60
B-7 Precipitation Map................................ 61
B-8 3D Precipitation Map............................. 62
B-9 Slope Map........................................ 63
B-10 3D Slope Map..................................... 64
B-ll Digital Elevation Model.......................... 65
B-12 3D Digital Elevation Model......................... 66
B-13 Map Showing Location of Clear Creek................ 67
B-14 Rockfall Hazard Map................................ 68
B-15 3D Rockfall Hazard Map............................. 69
1 -1 Number of Rockfalls by Season...................... 11
1- 2 Geologic Time Scale................................ 13
2- 1 Mineral Stabilities................................ 17
3- 1 Data Used in this Study............................ 27
5-1 Statistical Results................................ 43
5-2 Definition of Rockfall Hazard Map Cell Values.... 44
A-l Spreadsheet of Data................................ 53
Clear Creek Canyon has long been a popular transportation route
dating back to the early 1860s. The canyon contained a trail connecting
supply towns, such as Denver and Golden to mountain gold mining towns
like Black Hawk and Central City. During the 1870s, Colorado Central
railroad built a narrow-gauge line up Clear Creek to Black Hawk, Central
City, and, eventually, to Georgetown (Noel and others, 1994). The
railroad was abandoned in the late 1930s, but is still visible from U.S.
Highway 6 today (Grose and others, 1988). The highway was originally
constructed in the 1940s (Savage and others, 1998).
In 1990, Colorado voters approved limited stakes gambling in
Central City and Black Hawk (Lane, 2000). By 1992, the casinos were up
and running (Harveys Casino, 2000) and the traffic volume on U.S.
Highway 6 increased to accommodate gamblers heading up to the gaming
towns. Since this highway is heavily traveled, numerous rockfall events
each year result in injuries and deaths. As a result of the frequency of
rockfalls, the quantity of rockfall debris, and the high volume of traffic on
U.S. Highway 6, the Colorado Department of Transportation considers
this highway one of the most dangerous rockfall areas in the state (Gamer,
In an attempt to improve motorist safety and understand the
mechanisms behind rockfall events, this study was launched to determine
where rockfalls are likely to occur and what factors contribute to those
rockfalls. The focus of this study is the area surrounding U.S. Highway 6,
in Clear Creek Canyon, from the intersection of State Highway 93 in
Golden to the junction of U.S. Highway 6 and Interstate 70 just a few
miles east of Idaho Springs. Portions of Jefferson, Clear Creek, and
Gilpin counties are included in the study area (Figures 1-1 and 1-2).
The emphasis of this study is the use of Geographic Information
Systems (GES) to integrate rockfall location data with rock type,
vegetation, precipitation, slope, and aspect; to spatially analyze each of
these factors in relation to the locations of rockfall events; and to
determine what portions of Clear Creek Canyon produce the most
rockfalls and what factors play a role in those rockfall events.
Figure 1-1 Map of Colorado showing location of study area in
Jefferson, Clear Creek, and Gilpin counties.
JEFFERSON s V7
US. * V & U.S. 6
Figure 1-2 Inset map of previous figure showing location of U.S.
Highway 6 and study boundary in relation to the counties.
The objectives of this study are:
1) to determine the major factors that contribute to rockfall events
in Clear Creek Canyon;
2) to create a rockfall hazard map outlining low, medium, and
high risk areas throughout the canyon based on those major
3) to demonstrate how a combination of GIS and statistical
analyses can assist in evaluating rockfall hazards in Clear
The scope of this study is summarized as follows:
1) this study is limited to Gear Creek Canyon along U.S.
2) geologic, vegetation, and precipitation data were obtained from
existing digital data;
3) existing data were used in conjunction with limited field
4) analysis and synthesis of data were carried out using
1.3 The Clear Creek Canyon Study Area
Clear Creek Canyon is located at the eastern edge of the Front
Range region of Colorado and is dominated by metamorphic rocks of
Precambrian age. The study area covers portions of six U.S. Geological
Survey 7.5-minute 1:24,000-scale topographic and geologic quadrangles:
Morrison, Golden, Evergreen, Ralston Buttes, Squaw Pass, and Black
Hawk. However, U.S. Highway 6 is only located in the southwestern
comer of the Golden quadrangle and at the northern edge of the Evergreen
and Squaw Pass quadrangles (Figure 1-3).
1 ! I
! j i
; BLACK HAWK RALSTON BUTTES | GOLDEN
SQUAW PASS EVERGREEN I MORRISON
Figure 1-3 Location of study area in relation to 7.5-minute
The study area ranges in elevation from approximately 1,737
meters (5,700 feet) near Golden to 2,438 meters (8,000 feet) near the
western portion of U.S. 6, but elevations over 3,048 meters (10,000 feet)
are found within 2.5 miles of the highway. Santa Fe Mountain is the
tallest peak in the vicinity of the study area with an elevation of 3,212
meters (10,537 feet) (Sheridan and Marsh, 1976). It is located at the
southwestern comer of the study boundary.
Slopes are steep throughout the canyon, averaging about 56. The
aspect of these slopes is north and south facing because U.S. Highway 6
runs east to west through Clear Creek Canyon. Aspect is addressed later
as a rockfall factor. Steeper slopes tend to be associated with biotite
gneiss while less steep slopes are predominantly associated with granite
Clear Creek Canyon receives an average of 16 to 20 inches of
precipitation annually (NOAA, 1999) while the annual mean temperature
ranges from 43 to 45 Fahrenheit (Colorado Climate Center, 2000). The
winters can be long and precipitation due to snow is usually heavy.
Runoff from snowmelt occurs during the late spring or early summer
when temperature extremes between day and night are great enough to
create a freeze-thaw cycle, melting snow and ice during the day and
refreezing it at night. The freeze-thaw effect is a major cause of rockfalls
because the water seeps into joints and fractures during the day
and expands when it freezes at night. Expansion of ice wedges the rock
loose after repeated freezing and thawing. This is also the time period
when the majority of rockfalls occur. Rockfall events also seem to
coincide with heavy rain and snowfalls throughout the year, although the
majority of the precipitation tends to fall in the spring and summer
Figures 1-4 and 1-5 show the average monthly precipitation and
temperatures from 1962 to 1997 in the Evergreen area (Colorado Climate
Center, 2000). Evergreen is approximately the same elevation and has
similar climatic conditions as the study area. Table 1-1 shows the number
of rockfalls per season. This table does not account for all rockfall events
because the date of occurrence is not known for every event.
Figure 1-4: Average Monthly Precipitation in Evergreen from 1962 to
Figure 1-5: Average Monthly Temperatures in Evergreen from 1962 to
1997 ( Fahrenheit)
Table 1-1: Number of Rockfalls by Season in Clear Creek Canyon
Season Total Number of Rockfalls Average Number of Rockfalls
Winter 22 3.67
Spring 24 4.00
Summer 6 1.00
Fall 10 1.67
1.33 Geologic Setting
Clear Creek Canyon is dominated by complexly deformed rocks of
Precambrian age. These rocks were metamoiphosed to upper amphibolite
facies during regional metamoiphism approximately 1.7 billion years ago.
Multiple episodes of structural deformation created the dominant
northwest-trending folds and faults in the study area that control much of
the geology in the canyon.
During the late Paleozoic, the region was uplifted and Lower
Paleozoic strata were eroded away to expose the underlying Precambrian
rocks. The Laramide orogeny took place in the Upper Cretaceous, which
uplifted the area and forced the intrusion of dikes, sills, and stocks around
the region. Following the close of the Laramide orogeny in the middle to
late Eocene, erosion leveled the recently uplifted surface. The region was
again uplifted and faulted by tectonic events occurring since the Oligocene
and much of the present topography is the result of uplift and dissection of
the late Eocene surface (Reed and others, 1988). Table 1-2 summarizes
the major events affecting the eastern portion of the Front Range.
Table 1-2: Geologic Time Scale of Major Events Affecting the Eastern
Era Period Millions of years ago Event
Cenozoic Quaternary 1.8 Development of present topography; Surficial deposits
Tertiary 65 Uplift; Erosion; Denver Formation; Arapahoe Formation
Mesozoic Cretaceous 144 Beginning of Laramide orogeny; Laramie Formation; Pierre Shale; Niobrara Formation; Benton Shale; Dakota Group
Jurassic 206 Morrison Formation; Ralston Creek Formation
Triassic 248 Lykins Formation
Paleozoic Permian 290 Erosion of Ancestral Rockies; Lyons Formation
Pennsylvanian 323 Uplift of Ancestral Rockies; Erosion of Lower Paleozoic strata; Fountain Formation
Precambrian 1400 1700 1775 Silver Plume Batholith; Boulder Creek Batholith; Biotite and granite gneisses
Modified from Colorado Geological Survey (2000) and Curtis (1997).
2. Geology of Clear Creek Canyon
The geology of Clear Creek Canyon consists predominantly of
Precambrian metamorphic rocks with a few scattered Tertiary gravel
deposits and Quatemaiy surficial deposits. Sedimentary rocks lie to the
east of the canyon. Figure B-l is a map of bedrock geology for the study
area. The units used for this study reflect a synthesis of rock
classifications and mapping defined by numerous investigators working in
the region at different times and at different scales.
The Golden 7.5-minute quadrangle (1:24,000-scale) was mapped
and described by Van Horn (1972); the Morrison quadrangle was mapped
by Scott (1972); the Evergreen quadrangle by Sheridan and others (1972);
the Ralston Buttes quadrangle by Sheridan and others (1967); the Squaw
Pass quadrangle by Sheridan and Marsh (1976); and the Black Hawk
quadrangle by Taylor (1975). Bryant and others mapped and described
the Denver 1 x 2 quadrangle, which includes the entire study area, at a
scale of 1:250,000 (1981). Twetos 1:500,000-scale geologic map of
Colorado (1979) was ultimately used and modified because it was
obtained in digital format and ready to use in a GIS.
2.1 Precambrian Rocks
Precambrian rocks in the study area consist of biotite gneiss of
metasedimentary origin, granite gneiss of metavOlcanic origin, and
amphibolite of metasedimentary and metavolcanic origin. These rocks
were deformed and metamorphosed during multiple stages of Precambrian
folding. The metamorphic grade of the gneisses in the area is upper
amphibolite facies as indicated by the presence of sillimanite, garnet, and
migmatite. Although the Boulder Creek Batholith is not found in the
study area, it is believed that this pluton added heat to the terrain and helps
explain the low pressure-high temperature regional metamorphism
(Swayze and Holden, 1985). Throughout the study area, contacts are
gradational between the biotite gneiss and the granite gneiss. There are
also variations and gradations within each of the gneissic units.
2.1.1 Biotite Gneiss
The biotite gneiss is characterized for this study as a rock with a
predominance of biotite over feldspars and hornblende. This unit may
contain sillimanite and garnet as well as quartz, muscovite, and
plagioclase. Large portions of this unit are migmatitic and heavily
foliated. It is fine- to medium-grained and light- to dark-gray. The biotite
gneiss is considered to be metasedimentary in origin (Sheridan and Marsh,
Biotite gneiss has natural planes of weakness, which is
characteristic of biotites platy structure. Additionally, biotite breaks
down faster than quartz, which is much more abundant in the granite
2.1.2 Granite Gneiss
The granite gneiss for this study is defined as having a
predominance of feldspars, quartz, and hornblende over biotite. Although
biotite gneiss is interlayered with feldspar-rich gneiss, hornblende gneiss,
and amphibolite, it generally makes up less than 10 to 15 percent of the
rock in this unit. Granite gneiss is light- to medium-gray, tan, or pink in
color and fine- to medium-grained. Large parts of this unit are probably of
metavolcanic origin, but other portions may be metamorphosed
sedimentary rocks (Sheridan and Marsh, 1976).
The granite gneiss is the most competent unit in the study area due
to the predominance of resistant quartz, which seems to hold the rock
together and prevents the unit from rapid weathering (Table 2.1).
Table 2-1: Relative Chemical Stabilities of Some Rock-Forming Minerals
in the Weathering Cycle
i L Muscovite
Modified from Manual of Mineralogy (Klein and Hurlbut, 1993).
2.1.3 Hornblende Gneiss and Amphibolite
This unit is found throughout the study area, however, the
outcrops are too small to map separately at the scale used in this study. It
is composed of hornblende and plagioclase as well as biotite, quartz,
clinopyroxene, and epidote-group minerals. This unit is fine- to medium-
grained and gray to dark-green to black in color. Some parts of this unit
may be metavolcanic, while others might be metasedimentaiy. It is also
possible that some amphibolites are intrusive as well (Sheridan and Marsh,
This unit is brittle and easily broken up. Areas with outcrops of
this unit tend to have a lot of rock debris along side the highway. The
weakness and instability is a result of the rapid chemical breakdown of
hornblende, plagioclase, and pyroxene.
Intrusions in the study area include pegmatites and lamprophyres
of Precambrian age. Although they are common, they are not large
enough to map at the scale used in this study. Both the pegmatites and
lamprophyres are younger than the biotite and granite gneisses because
they intruded into the gneissic complexes.
Pegmatites are common throughout Clear Creek Canyon. They are
coarse-grained, white to pink to light-gray granite. They occur as dikes,
lenses, and irregularly shaped bodies. Compositionally, the pegmatites are
quartzofeldspathic with lesser amounts of biotite, muscovite, and
magnetite. Although the pegmatites are of Precambrian age, some are
possibly related to the Boulder Creek Batholith while others may be
related to the Silver Plume Batholith. The older pegmatites related to the
Boulder Creek Batholith show gneissic foliation, while the younger
pegmatites related to the Silver Plume Batholith are nonfoliate (Sheridan
and Marsh, 1976).
Pegmatites are resistant to erosion due to the abundance of stable
quartz over less stable minerals like biotite, muscovite, and magnetite.
The lamprophyres are of Precambrian age and are composed of
green hornblende, biotite, microcline, and albite-oligoclase. They are
dark-gray to greenish-black in color and form short dikes a few
centimeters to a few meters thick. Lamprophyres may have been intruded
at several different times during the Precambrian; the more foliated the
dike is, the older it is likely to be (Sheridan and Marsh, 1976). They are
mainly concentrated around milepost 266.
Lamprophyres are non-resistant due to the lack of quartz and the
predominance of unstable minerals, such as hornblende, biotite, and
2.3 Sedimentary Rocks
The rock units to the east of the Front Range Precambrian rocks
are of sedimentary origin. However, due to the presence of the Golden
fault, the thickness and sequence of rock units is variable. At the junction
of U.S. Highway 6 and State Highway 93, several units have been cut out
by the fault. The units present in that area include the Pennsylvanian and
Permian Fountain Formation; Upper Cretaceous Pierre Shale, Fox Hills
Sandstone, Laramie, and Arapahoe Formations; and the Upper Cretaceous
and Paleocene Denver Formation. This sequence is missing numerous
formations from the Permian through the Upper Cretaceous including the
Lyons Sandstone, Lykins, Ralston Creek, Morrison, Dakota Group, and
Niobrara Formations. South of this junction, in the Morrison area, the
stratigraphic sequence is more complete. Rock units in this area include
the Fountain through the Denver Formations.
These sedimentary units are composed of conglomerate, sandstone,
siltstone, claystone, shale, and limestone (Figure 2-1). In addition to
sedimentary components, the Denver Formation also consists of basaltic
flows dating back to approximately 62-64 million years ago. This
sedimentary sequence is shown on Figure B-l, however, not all of the
formations are represented on the map due to the extremely small scale of
the original map. Although these units lie within the study area boundary,
they are not located within Clear Creek Canyon and were not used in the
analysis of rockfall hazards.
GENERALIZED STRATIGRAPHIC COLUMN
East of Clear Creek Canyon
Conglomerates, sandstones, and slltstones
Sandstones, shales, and conglomerates;
volcanic material common
Sandstones, shales, and thin coal layers in
Shales and sandstones; occasional fossils
Shales, sandy shales, and a few thin
Shales and limestones
Shales with tew thin limestones; platy siltstone
Sandstones with some shales; filled stream
Sandstones, shales, and thin limestones;
Shales, sandstones, siltstones, and gypsum
Siltstones, shales, and thin limestones
Permian Lyons i2ooFr.rl-a' __ . ,__.. .
- ----- ,-------r -t-O^Sandstones and conglomerates; cross bedded
(1 to 2 billion years)
l J Sandstones, siltstones, and conglomerates
Gneisses, schisls, and intrusions
Modified from Weimer and Ray (1997).
2.4 Surficial Deposits
Tertiary gravel deposits and Quaternary surficial deposits are
scattered throughout the study area. However, because they are not the
focus of this study and because they are too small to map at the scale used
in this study, the surficial deposits are not shown on the map of bedrock
geology. Some of the Quaternary surficial deposits include talus,
alluvium, alluvial fan deposits, and a landslide deposit located at the
junction of U.S. Highway 6 and Colorado State Highway 119. Some of
the alluvium may have been derived from the Pinedale or Bull Lake
Glaciations during the Pleistocene. The Tertiary gravel and probably
some non-glacially derived alluvium are related to ancestral Clear Creek
drainage (Sheridan and Marsh, 1976). Also, some of these surficial
deposits have formed as a result of rockfalls and other mass wasting
2.5 Structural Geology
The structural features of the area are a result of several episodes
of regional deformation. The main deformational episode affecting Clear
Creek Canyon predates the emplacement of the Boulder Creek Batholith
approximately 1.7 billion years ago. Deformation associated with the
intrusion of the Boulder Creek Batholith and the intrusion of the Silver
Plume Batholith, approximately 1.4 billion years ago, may have also
affected the structures in the study area (Taylor, 1975). The results of
these deformational episodes produced dominant northwest-trending
folding and faulting in the Clear Creek Canyon area.
The study area consists of a series of isoclinal folds all of which
trend to the northwest (Figure B-3). Most of the folds in the study area are
antiformal. Drag folds are common along the limbs of these folds (Gable,
1968). The most dominant fold affecting the study area is located at the
southern edge of the Ralston Buttes quadrangle and the northern edge of
the Evergreen quadrangle. This antiform is upright and moderately tight.
The trace of its axial surface trends nearly due west to approximately N.
75 W. and dips vertically to about 85 N. The axis plunges at a low to
moderate angle to the northwest (Sheridan and others, 1967).
As a result of this antiform, the northern edge of the Evergreen
quadrangle is affected by the southern limb of the fold. The portion of
U.S. Highway 6 that lies on the Squaw Pass 7.5-minute quadrangle is
mostly affected by the northern limb of several other folds. Due to the
orientation of these folds, the dominant foliation in the study area strikes
to the northwest and dips to the northeast on the northern limb of the folds
and to the southwest on the southern limb of the folds.
The study area is bounded by the Floyd Hill fault to the west and
the Golden fault just to the east of Clear Creek Canyon. Other faults
within Clear Creek Canyon include the Black Hawk, Junction Ranch, and
Windy Saddle faults (Figure B-4). All of the faults in Clear Creek Canyon
are part of a family of northwest-trending faults that cut the Proterozoic
igneous and metamorphic rocks in this part of the Colorado Front Range.
The faults all dip steeply and many show apparent left-lateral offset of the
Proterozoic rocks of a few hundred to several thousand feet. Movement
along the faults may have started as early as the Middle Proterozoic, but
were reactivated during the Laramide (Reed, 1991).
The Golden fault dips to the west and trends north-northwest in the
vicinity of the study area. The western block of the fault was upthrown in
relation to the eastern block. This fault cuts sedimentary rocks that post-
date the Precambrian metamorphic rocks and results in variable thickness
and sequence of rock units in the area. The cutting of these rock units
occurred during the Laramide (Van Horn, 1957).
None of these faults seem to be directly correlated with the
location of rockfall events.
Presentation of Data
Many types of data were used for this study including rockfall
locations and data for each rockfall factor. All of these data were analyzed
statistically. Table 3-1 lists the format and source for each type of data
and Appendix A is the spreadsheet used in analysis.
Table 3-1: Data Used in this Study
Data Format Source Analyzed Statistically
Rockfall Locations Spreadsheet State Patrol and CDOT Yes
Rock Type GIS CGS/Tweto Yes
Vegetation GIS CNDIS website Yes
Precipitation GIS CGS/NOAA Yes
Slope Angle GIS Derived in GIS Yes
Aspect GIS Derived in GIS Yes
Dip of Foliation Spreadsheet 7.5 quadrangles Yes
Joint Frequency Spreadsheet Collected in field Yes
Joint Orientation Notes Collected in field No
3.1 Rockfall Location Data
The most critical part of this study was locating rockfall events in
Clear Creek Canyon. Data were obtained from several sources. Data
from approximately 1976 through 1989 were taken from the Colorado
Rockfall Hazard Rating System published by the Colorado Department of
Transportation (Andrew, 1994). Rockfall location data from 1995 to May
2000 were obtained from Colorado State Patrol accident reports.
The data from the Colorado Department of Transportation (CDOT)
only listed the number of rockfalls for each mile in Clear Creek Canyon.
No information was obtained on the date, time, aspect, etc. Data from the
Colorado State Patrol included milepost and date. These data are stored in
3.2 Rockfall Factors
Many factors were analyzed in this study that were thought to be
related to the occurrence of rockfalls. Data were collected on rock type,
vegetation, precipitation, aspect, slope angle, foliation, and joints.
Statistical analyses were run on all these data to determine which factors
are correlated with the most rockfall events.
3.2.1 Rock Type
Twetos Colorado State geologic map was obtained in GIS format
from the Colorado Geological Survey (CGS) and, as a result, was used as
a base and modified to fit the rock units defined for this study. Several
U.S. Geological Survey 1:24,000-scale geologic maps of the study area
were used to help determine the rock types defined in this study. These
maps include the Golden, Morrison, Ralston Buttes, Evergreen, Black
Hawk, and Squaw Pass 7.5-minute quadrangles (Figure 1-3). For the
purpose of this study, two main rock types were considered: biotite gneiss
and granite gneiss. Both of these rock units were described in the previous
chapter (Figure B-l).
Several types of vegetation are present throughout Clear Creek
Canyon including midgrass prairie, mesic upland shrubs, xeric upland
shrubs, big sagebrush, Douglas fir, and ponderosa pine. Additionally,
spruce fir and lodgepole pine are found within the study area, but are not
immediately surrounding U.S. Highway 6. In general, the shrubs and
sagebrush are found on south facing slopes, which tend to be drier than
north facing slopes. The conifers and the midgrass prairie are
predominantly found on wetter, cooler north facing slopes.'
Midgrass prairie is a blend of shortgrass and tallgrass varieties.
This vegetation type is found at the lowest elevations and in the eastern
portion of the study area. Xeric upland shrubs are dominated by mountain
mahogany and are found in dry environments. Xeric shrubs are
predominantly located above midgrass prairie and below mesic upland
shrubs. Mesic upland shrubs include a variety of shrub communities, such
as Rocky Mountain maple, serviceberry, and/or chokecherry. Mesic refers
to an intermediate environment that is not extremely wet or extremely dry.
Mesic shrubs are located above xeric upland shrubs and below big
sagebrush. Big sagebrush is found predominantly on south facing slopes
above shrubs and below coniferous trees. Douglas fir is often the
dominant tree species of north facing slopes and steep ravines. This
vegetation type is generally located between big sagebrush and the other
types of coniferous tree species. Ponderosa pine is usually found on north
facing slopes and has the highest fire frequency of any Rocky Mountain
forest type. It is found between Douglas fir and lodgepole pine.
Lodgepole pine and spruce fir are found at the highest elevations and the
western portion of the study area (Colorado Natural Diversity Information
Figure B-5 is a vegetation map for the study area, which shows the
distribution of the various plant communities in the canyon. The
vegetation data for the study area were downloaded from the Colorado
Natural Diversity Information Source (CNDIS) website and are stored in
There are two precipitation ranges affecting Clear Creek Canyon,
16 to 18 inches per year and 18 to 20 inches per year. However,
precipitation ranges of 14 to 16 and 20 to 24 inches per year are found in
the study area (Figure B-7). The precipitation range of 16 to 18 inches per
year is found at the western and eastern portions of U.S. Highway 6 and
the precipitation range of 18 to 20 inches per year is found between these
two areas. Precipitation range varies with elevation. Lower elevations get
less precipitation while the highest elevations get the most precipitation.
It is important to note that these data were obtained from a
statewide precipitation map and lack enough detail to accurately represent
the precipitation trends in the canyon. A more detailed study and
collection of data was not undertaken due to time constraints, but may
give better insight into the timing of rockfall events.
Statewide digital precipitation data were obtained from the
Colorado Geological Survey, but originated from the National Oceanic
and Atmospheric Administration (NOAA) website. The precipitation map
is stored in GIS format.
North and south facing slopes are very different in terms of their
characteristics. South facing slopes receive more sunlight and, therefore,
tend to be drier, have less vegetation, and are characterized by large
temperature extremes from day to night, which provoke frost wedging
from frequent freeze-thaw. Frost wedging is a form of physical
weathering, which dominates south facing slopes.
North facing slopes are just the opposite of south facing slopes.
They receive less sunlight, which inhibits evaporation and creates a wetter
environment, are more densely vegetated, and temperature extremes
between day and night are not as great. Root wedging is more frequent
than frost wedging due to the abundance of vegetation and lack of
temperature extremes. Additionally, chemical weathering is more
dominant than physical weathering due to the wetter environment.
The majority of slopes in Clear Creek Canyon that are next to the
highway are south facing while north facing slopes tend to be on the river
side of the canyon (B-13). This is important because most of the debris
affecting U.S. Highway 6 originates from south facing slopes whereas
debris falling from north facing slopes usually falls into the river. Aspect
was determined using GIS analysis techniques.
Slope angles for the entire study area were derived using GIS. The
average slope for each mile was then determined for analysis purposes.
The goal of analyzing the average slope for each mile was to determine
what angles are associated with the most rockfall events. Figure B-9 is a
slope map of the study area.
The main consideration when dealing with foliation is whether it is
dipping toward or away from the road. To make this determination, it is
best to consider the aspect of a specific area and the direction of foliation
dip. If the dip of foliation is to the south on a south facing slope, then the
dip is toward the road or parallel to slope. If, on the other hand, the dip is
to the north on a south facing slope, then the dip is away from the road.
Foliation strike and dip data were taken from the 7.5-minute
geologic maps for the study area and are stored in spreadsheet format.
In this study, joints were considered in two different ways. First, a
joint frequency was taken for every milepost by determining the number
of joints over a distance of 7.6 meters (25 feet). Second, the dip of the
joints in relation to the highway was also considered in a similar fashion to
foliation dip. However, only joint frequency was analyzed statistically.
Dip of joints was only noted as observations due to the complexity of joint
sets in the study area and due to time constraints. Figure 3-1 shows the
complexity of these joint sets.
Joint frequency data were collected in the field at random rock
outcrops for each milepost. These data are stored in spreadsheet format.
Observations on joint orientations were also noted in the field.
Figure 3-1 Photo showing multiple joint sets at mile 264.5.
Locating, acquiring, and converting existing data for the study was
critical for successful analysis and modeling of the problem. A three-step
methodology was followed to analyze and model rockfall hazards. The
steps include integrating data into the GIS, analysis, and modeling.
4.1 Integration of Data into GIS
This project combined several different types of data into a GIS.
Most of the data were obtained in a digital format and did not have to be
scanned or digitized. For instance, Twetos geologic map and the
precipitation data were obtained in a format that automatically opened in
ArcView. Twetos map was then modified based on the geologic units
used for this study. Precipitation data were clipped in ArcView to extract
the data for the study area. The vegetation data were downloaded from the
internet and then imported directly into ArcView.
The digital elevation models (DEM) were imported and merged in
ARC/1NFO (Figure B-l 1). Slope and aspect maps were then created from
the DEM. Eventually, these coverages were brought into ArcView for
display and analysis purposes.
Rockfall locations, foliation, and joint data are not in GIS format,
but were handled in Microsoft Excel spreadsheets.
Analyses of the data were first performed using Environmental
Systems Research Institutes (ESRI) ARC/INFO version 7.2 for Unix
workstations and ArcView 3.2 for Windows geographic information
systems software. These software packages were used to work with
geographic data that is spatially linked to a map and related to the space
around us (Clarke, 2000). ARC/INFO is a powerful software program that
has the capability to perform complex analyses and has many tools that
allow the user to manipulate data. ArcView is an easy to use desktop GIS
that allows for quick display, query, and analysis of the data. For more
information on these two GIS software packages and a more detailed
discussion of geographic information systems, see Getting Started with
Geographic Information Systems by Keith Clarke (2000).
One of GISs capabilities is to overlay the different types of data to
see how they relate to one another spatially. Overlaying is a technique
where one type of data is displayed on top of another type of data, but both
data types are visible. For instance, the highway and the river can be
displayed together so the user can see what side of the road the river lies
on (Figure B-13). From this information, the user can determine if
rockfalls from south facing slopes will fall into the river or onto the road.
This was one method of determining aspect in this study. Other overlay
examples include displaying the geology and the highway together to see
where the biotite gneiss is located along U.S. Highway 6 (Figure B-l).
Overlaying the different types of data is an easy way to analyze spatial
relationships between the data.
ArcViews Spatial Analyst and 3D Analyst were also used to aid in
analysis. Spatial Analyst is the extension of ArcView that can work with
grid data. A grid is made up of a series of rows and columns. These rows
and columns form cells that store some type of value or attribute. The
advantage of grid data is that each cell in the grid has its own value so that
small areas or a set of cells can be differentiated from the rest of the data.
Digital elevation models (DEMS) are one of the most common examples
of grid data where each cell has its own elevation value. DEMs are stored
in a grid because elevation is continuous and every point has an elevation
value, which is stored in a grid cell.
3D Analyst is the extension of ArcView that can display data in
three dimensions. This extension was used to drape the different types of
data over the DEM. For example, when the geology is draped over the
DEM, the user can see what type of rock makes up the highest peaks and
how geology changes with elevation and topography. Figure B-2, B-6, B-
8, B-10, B-12, and B-15 are three dimensional maps of the study area.
The second phase of analysis utilized a statistical package called
STATGRAPHICS Plus for Windows 4.0 produced by Statistical Graphics
Corporation. This software was used to perform statistical analyses on
rock type, slope, aspect, vegetation, precipitation, joint frequency, and
foliation data to determine which factors are statistically significant and
explain the most variability in the rockfall data.
The statistical methods used to determine which factors are most
highly correlated with the number of rockfalls per milepost include
regression analysis and correlation coefficients. Regression analysis
quantifies the relationship between a dependent variable and one or more
independent variables. The dependent variable used in the analyses was
the number of rockfalls per milepost and the independent variables were
rock type, slope, aspect, vegetation, precipitation, joint frequency, and
foliation. Correlation coefficients measure the linear association between
two variables. The coefficient values fall between -1 and +1. A positive
correlation indicates that the variables vary in the same direction while a
negative correlation indicates that the variables vary in the opposite
direction. Statistically independent variables have an expected correlation
of 0 (Statistical Graphics Corporation, 1999). Table 5-1 shows the results
of these analyses.
Once the significant factors were identified, a rockfall hazard map
was created to show which areas of Clear Creek Canyon are considered
high risk and which areas are considered low risk based on the presence or
absence of the significant factors.
The rockfall hazard map was produced in ArcView using the
Spatial Analyst extension to work with the grid data. First, the geology
layer was converted to a grid and reclassified. Reclassifying the geology
grid changed the cell values, which held the rock type, to a numeric value.
All cells holding a value of biotite gneiss were given a value of 200 and all
other cells were given a value of 2. The values of 200 and 2 were
arbitrarily chosen to separate rock type based on biotite content. Second,
the slope grid was reclassified to give all cells with a slope greater than
58 a value of 100 and all other cells a value of 1. Again, the values of
100 and 1 were chosen arbitrarily to separate the two slope ranges
determined statistically. Third, the reclassified geology and slope grids
were added together to produce cell values of 300,201,102, and 3.
The main result of this study is a rockfall hazard map that was
created by determining which factors explain the most rockfall events.
The factors that were found to be statistically significant were then used to
create the hazard map, which shows high, medium, and low risk areas
based on the presence or absence of those factors.
5.1 Numerical Results
Statistical analyses indicate that rock type is the most important
factor in determining where rockfalls are likely to occur. The correlation
coefficient between rock type and total rockfall accidents per milepost is
0.7856 and the P-value is 0.0005. The bivariate r2 value is 61.7 meaning
that rock type explains 61.7% of the variability in rockfall events. Low
risk areas mainly occur in granite gneiss while high risk areas are
predominantly made up of biotite gneiss.
The second most important factor in predicting rockfall occurrence
is average slope per milepost. The correlation coefficient between average
slope per milepost and total rockfall accidents per milepost is 0.7511, the
P-value is 0.0012, and r2 is 56.4%. Low risk areas have an average slope
per milepost ranging from 42 to 56 whereas high risk areas have an
average slope that ranges from 58 to 65. It is important to note that there
is possible collinearity between rock type and slope angle meaning that the
two are not truly independent.
The final factor that is statistically significant in terms of
explaining rockfall events is the dominant dip of the foliation of the rock
within each milepost. The correlation coefficient between dip of foliation
and total rockfall accidents per milepost is -0.6295, the P-value is 0.0119,
and r2 is 39.6%. In low risk areas, the rock is mostly dipping away from
the road while it is mainly dipping toward the road in high risk areas.
Table 5-1 summarizes these results and shows that rock type and
slope angle are significant at the 99% confidence level with P-values less
than 0.01 while foliation is significant at the 98% confidence level. Rock
type has the strongest relationship with rockfall events and explains the
most variability; this is closely followed by slope angle and foliation.
Together, rock type, slope angle, and foliation explain 65.5% of the
variability in rockfall events. The other factors studied in this project are
not statistically significant at the 95% confidence level meaning that their
P-values are greater than 0.05.
Table 5-1: Correlation Coefficients, P-values, and R Values for
Factor Correlation Coefficient P-value Bivariate R2
Rock type 0.7856 0.0005 61.7%
Slope Angle 0.7511 0.0012 56.4%
Dip of Foliation -0.6295 0.0119 39.6%
Precipitation 0.4593 0.0850 21.1%
Vegetation -0.3422 0.2119 11.7%
Aspect 0.1119 0.6913 1.3%
Joint Frequency 0.0540 0.8484 0.3%
5.2 Graphical Results
The combination of the geology and slope grids, discussed in
section 4.3, shows the areas that are in the biotite gneiss and have a slope
greater than 58. These are the areas that are considered to be high risk.
Medium risk areas either have a slope greater than 58 in the granite gneiss
or have a slope less than 58 in the biotite gneiss. Low risk areas have a
slope less than 58 and are located in the granite gneiss (Table 5-2). In
general, the high risk miles include 263 to 270 while the lower risk miles
include 257 to 262 and mile 271 (Figure B-14).
Table 5-2: Definition of Rockfall Hazard Map Cell Values
Cell Value Definition Risk Level
300 Two high risk criteria: biotite gneiss and slope >58 High
201 One high risk criteria and one low risk criteria: biotite gneiss and slope <58 Medium
102 One high risk criteria and one low risk criteria: granite gneiss and slope >58 Medium
3 Two low risk criteria: granite gneiss and slope <58 Low
Foliation data were not handled in GIS format and, therefore, were
not integrated into the rockfall hazard map. However, the high risk
foliation areas are nearly identical to those determined using GIS and are
found from mileposts 263 to 271. This means that areas dominated by
biotite gneiss with a slope greater than 58 also have a foliation dip
parallel to slope. Low risk foliation areas are predominantly found in die
granite gneiss from mileposts 257 through 262 and closely match the
lower risk areas determined using GIS.
The following sections discuss the results of the statistical analyses
and offer explanations for the significance of each rockfall factor.
5.3.1 Rock Type
Statistical analyses indicate that rock type is the most important
factor contributing to rockfall hazards because it explains the most rockfall
occurrences. Miles 257 through 262 and mile 271 are dominated by
granite gneiss, while miles 263 through 270 are dominated by biotite
gneiss. The miles with the most rockfalls are 263 through 270, which is
also where the biotite gneiss occurs. From a geological standpoint, biotite
has natural planes of weakness and is less resistant to erosion than the
quartz-rich granite gneiss. These characteristics help explain the higher
frequency of rockfalls in the biotite gneiss.
Figure B-5 is a vegetation map for the study area. This map shows
that the dominant vegetation type between mileposts 263 and 271 is mesic
upland shrubs with smaller areas of xeric shrubs, Douglas fir, and
ponderosa pine. This is also the area that is considered to be high risk in
terms of rockfalls. However, vegetation is not a statistically significant
rockfall factor because mesic shrubs are found in lower risk areas and
because the lower risk areas are dominated by a variety of vegetation types
instead of just one vegetation type.
Miles 257,258,270, and 271 have a dominant precipitation range
of 16 to 18 inches per year while miles 259 through 269 are in the 18 to 20
inches per year range. Since the majority of Clear Creek Canyon falls
within the 18 to 20 inches per year precipitation range, precipitation is not
a statistically significant rockfall factor. However, there does seem to be
more rockfall events during high precipitation periods than during drier
periods. Additionally, there are more rockfall events during the winter
and spring (Table 1-1), which may be a result of both precipitation and
temperature extremes creating the freeze-thaw effect.
Aspect is not a statistically significant factor because the majority
of the slopes directly affecting the highway are south facing in Clear
Creek Canyon. Only miles 259, 265, and 266 have a dominant north
facing aspect. Most of the rockfalls originating from north facing slopes
will fall into the river and not onto the road.
Slope angle is a statistically significant factor because the slope
grid calculated from the digital elevation model indicates that mileposts
263 through 270 all have areas with a slope greater than 73.5 and an
average slope over the entire mile greater than 58 (Figure B-9) This is
the same area where biotite is the dominant rock type and, consequently,
where the highest rockfall hazards occur. Miles 257 through 262 and mile
271 have an average slope of less than 58 and are dominated by granite
Foliation dips oriented toward the road are more dangerous
because they are more likely to slip in the direction of the road. Dip of
foliation is statistically significant because foliation generally dips away
from the road from mileposts 257 through 262 and milepost 266 and dips
toward the road from mileposts 263 through 271 excluding milepost 266.
The former is predominantly located in the low risk granite gneiss while
the latter is predominantly found in the high risk biotite gneiss. Folding in
the study area controls much of this foliation.
Joint frequency is not a statistically significant factor because there
is no obvious correlation with the number of rockfall accidents per
milepost. The joint frequency for the low risk miles of 257 to 262 and
mile 271 ranges from 15 to 55 joints while the range for the high risk
miles of 263 to 270 is 24 to 43. However, observation indicates that joint
orientation does seem to play an important role in the distribution of
rockfalls, but a more detailed analysis of the complex system of joints
should be undertaken to determine its true significance.
Conclusions and Recommendations
This project was a reconnaissance study of Clear Creek Canyon
that brought together many different types of data that were analyzed in
relation to rockfall occurrences. The objective of the study was to use GIS
and statistical analyses to determine which factors contribute the most to
rockfalls. This information was then used to create a rockfall hazard map
of Clear Creek Canyon that shows high and low risk areas. The results of
this analysis draw some very important conclusions and also lead to
recommendations for future work.
The data for the study area provide an important representation of
the conditions present throughout Clear Creek Canyon, which is located in
the Precambrian core of the Front Range just west of Golden, Colorado.
From these data, it was possible to create a rockfall hazard map showing
areas that are likely to produce rockfalls and areas that are less likely to
produce rockfalls. This hazard map is based on the presence or absence of
factors that were found to be statistically significant in terms of explaining
the occurrence of rockfalls. The factors considered in this study include
rock type, slope inclination, aspect, vegetation, precipitation, dip of
foliation, and joint frequency. Results indicate that rockfall events are
more likely to occur in areas where biotite gneiss is the dominant rock
type, where the slope is greater than 58, and where the dip of foliation is
parallel to slope. In general, these areas are located between miles 263
These results should be used as guidelines for mitigation purposes
to help prevent injuries and deaths due to rockfalls along U.S. Highway 6.
Mitigation should focus first on the high risk areas then to medium and
low risk areas as funds become available. The need for this analysis and
for mitigation has increased over the years due to the high volume of
traffic along the highway to the gaming towns of Black Hawk and Central
City. Hopefully, the combination of this analysis and mitigation efforts
will help make U.S. Highway 6 safer for motor vehicle traffic.
The results of this study provide a basis for understanding the
cause and location of rockfall events. However, further research in two
different areas may give better insight into when and where rockfalls may
occur. The first recommendation would entail a closer look at
precipitation trends as they relate to rockfall events. Tracking daily
precipitation amounts and comparing them to the timing of rockfalls
would lead to a better understanding of when rockfalls may occur and how
rockfall occurrences relate to seasonal precipitation patterns.
The second recommendation would involve a more detailed study
on joint orientation. The joint sets in the study area are complex mainly
due to the fact that there are multiple joint sets at any given rock outcrop.
Collecting more field data on the orientation of these joint sets and
comparing the data to the location of rockfalls may help explain the
distribution of rockfall events.
Milepost Rockfall Accidents Dominant Rock Type Slope Angle Dominant Foliation Dominant Precipitation Dominant Vegetation Dominant Aspect Joint Frequency
257 3 Granite gneiss 42.6 Away 16-18 Douglas fir South 40
258 4 Granite gneiss 55.7 Away 16-18 Douglas fir South 15
259 4 Granite gneiss 54.7 Away 18-20 Big sagebrush North 23
260 3 Granite gneiss 52.0 Away 18-20 Big sagebrush South 55
261 4 Granite gneiss 49.0 Away 18-20 Big sagebrush South 27
262 6 Granite gneiss 54.5 Away 18-20 Mesic shrub South 30
263 18 Biotite gneiss 62.0 Toward 18-20 Mesic shrub South 26
264 26 Biotite gneiss 62.2 Toward 18-20 Mesic shrub South 43
265 19 Biotite gneiss 64.9 Toward 18-20 Mesic shrub North 24
266 10 Biotite gneiss 59.9 Away 18-20 Mesic shrub North 39
267 14 Biotite gneiss 58.9 Toward 18-20 Mesic shrub South 30
268 10 Biotite gneiss 63.4 Toward 18-20 Mesic shrub South 29
269 11 Biotite gneiss 61.3 Toward 18-20 Mesic shrub South 40
270 8 Biotite gneiss 60.6 Toward 16-18 Xeric shrub South 25
271 2 Granite gneiss 45.6 Toward 16-18 Xeric shrub South 31
Table A-l: Data used in statistical analyses.
Bedrock geologic map of the study area.
, v U.S.6
| Dakota/Monison/Ralston Creek
^ Denver Basalt
i Blotlte Gneiss
3D bedrock geologic map of the study area.
Map showing major folds in the study area.
Map showing major faults in the study area.
1 Denver Basalt
I 1 Granite Gneiss
i___i Blotlte Gneiss
Vegetation map of the study area.
f\/ U.S. 6
Maalc Upland Shrub
Xartc Upland Shrub
3D vegetation map of the study area.
Precipitation map of the study area.
/V U.S. 6
Precipitation Range (inches)
3D precipitation map of the study area.
Slope map of the study area.
| | No Data
3D slope map of the study area.
s ns s
Digital elevation model of the study area.
3D digital elevation model of the study area.
Map showing the location of Clear Creek in relation to U.S. Highway 6.
A / U.S. 6
Rockfall hazard map of the study area.
3D rockfall hazard map of the study area.
Andrew, R.D., 1994, Colorado rockfall hazard rating system: Colorado
Department of Transportation Report CTI-CDOT-2-94, p. 1-29,
Blatt, H. and Tracy, R.J., 1996, Petrology: New York, W.H.
Freeman and Company, p. 365-463.
Bryant, Bruce, McGrew, L.W., and Wobus, R.A., 1981, Geologic map of
the Denver 1 x 2 quadrangle, north-central Colorado: U.S.
Geological Survey Miscellaneous Investigations Map 1-1163, scale
Clarke, K.C., 2000, Getting started with Geographic Information Systems:
New Jersey, Prentice Hall, 353 p.
Colorado Climate Center, 2000, Climate data. Accessed September 12,
2000 at URL http://ulysses.atmos.colostate.edu.
Colorado Geological Survey, 2000, Geologic time scale highlighting
events in Colorado. Accessed September 12,2000 at URL
Colorado Natural Diversity Information Source, 1999, Vegetation data.
Accessed November 8,1999 at URL http://ndis.nrel.colostate.edu.
Colorado State Patrol, 1995-2000, Accident reports: Unpublished Internal
Curtis, G.R., 1997, The Colorado mineral belt, in Bolyard, D.W. and
Sonnenberg, S.A., eds., Geologic histoiy of the Colorado Front
Range: Rocky Mountain Association of Geologists Field Trip #7,
Gable, D J., 1968, Geology of the crystalline rocks in the western part of
the Morrison quadrangle, Jefferson County, Colorado: U.S.
Geological Survey Bulletin 1251-E, 45 p.
Gamer, Joe, 2000, Rockslides made priority: Denver Rocky Mountain
News, Jan. 20,2000, p. 7A.
Grose, T.L.T., Bryant, Bruce, and Reed, J.C., Jr., 1988, View of the Blue
River valley and west flank of the Front Range, in Holden, G.S.,
ed., Geological Society of America field trip guidebook:
Professional Contributions of the Colorado School of Mines, n. 12,
Harveys Wagon Wheel Hotel and Casino, 2000: Oral Communication,
Klein, C. and Hurlbut, C.S., Jr., 1993, Manual of mineralogy: New York,
John Wiley and Sons, Inc., p. 571.
Lane, George, 2000, Access road nearer for reeling casino town.
Accessed August 17,2000 at URL http://www.denverpost.com.
National Oceanic and Atmospheric Administration, 1999, Precipitation
map. Accessed November 20,1999 at URL http://www.noaa.gov.
Noel, T.J., Mahoney, P.F., and Stevens, R.E., 1994, Historical atlas of
Colorado: Norman, OK., University of Oklahoma Press, p. 26-31.
Reed, J.C., Jr. and others, 1988, Geology and mineral resources of central
Colorado, in Holden, G.S., ed., Geological Society of America
field trip guidebook: Professional Contributions of the Colorado
School of Mines, n. 12, p. 68-73.
Reed, J.C., Jr., 1988, Precambrian rocks of Clear Creek Canyon, in
Holden, G.S., ed., Geological Society of America field trip
guidebook: Professional Contributions of the Colorado School of
Mines, n. 12 p. 116-118.
Reed, J.C., Jr., 1991, Geology of a fault-controlled cave in Precambrian
crystalline rocks in Clear Creek Canyon, Jefferson County,
Colorado: U.S. Geological Survey Open File Report 91-144, p. 1-
Reed, J.C., Jr., Sheridan, D.M., and Bryant, Bruce, 1973, Map showing
faults, joints, foliation, and surficial deposits in the Evergreen
quadrangle, Jefferson County, Colorado: U.S. Geological Survey
Miscellaneous Investigations Series Map I-786-F, scale: 1:24,000.
Savage, W.Z., Coe, J.A., Vames, D.J., Streufert, R.K., and Godt, J.W.,
1998, History of movement of the Clear Creek Forks landslide,
Clear Creek County, Colorado, U.S. Geological Survey
Open File Report 98-150, p. 1.
Scott, G.R., 1972, Geologic map of the Morrison quadrangle, Jefferson
County, Colorado: U.S. Geological Survey Miscellaneous
Geologic Investigations Map I-790-A, scale 1:24,000.
Sheridan, D.M. and Marsh, S.P., 1976, Geologic map of the Squaw Pass
quadrangle, Clear Creek, Jefferson, and Gilpin Counties, Colorado:
U.S. Geological Survey Geologic Quadrangle Map GQ-1337, scale
Sheridan, D.M., Maxwell, C.H., and Albee, A.L., 1967, Geology and
uranium deposits of the Ralston Buttes District, Jefferson County,
Colorado: U.S. Geological Survey Professional Paper 520, 121 p.
Sheridan, D.M., Reed, J.C., Jr., and Bryant, Bruce, 1972, Geologic map of
the Evergreen quadrangle, Jefferson County, Colorado: U.S.
Geological Survey Miscellaneous Geologic Investigations Map I-
786-A, scale 1:24,000.
Statistical Graphics Corporation, 1999, Online help: STATGRAPHICS
Plus for Windows 4.0 Software.
Swayze, G.A. and Holden, G.S., 1985, Middle Proterozoic metamorphism;
central Front Range, Colorado: Geological Society of America
Abstracts with Programs, v. 17, n. 4, p. 266.
Taylor, R.B., 1975, Geologic map of the Black Hawk quadrangle, Gilpin,
Jefferson, and Clear Creek Counties, Colorado: U.S. Geological
Survey Geologic Quadrangle Map GQ-1248, scale 1:24,000.
Tweto, Ogden, 1979, Geologic map of Colorado, U.S. Geological Survey
State Geologic Map SG6-2, scale 1:500,000.
Van Horn, R., 1972, Surficial and bedrock geologic map of the Golden
quadrangle, Jefferson County, Colorado: U.S. Geological Survey
Miscellaneous Geologic Investigations Map 1-761-A, scale
Van Horn, R., 1957, Bedrock geology of the Golden quadrangle,
Colorado: U.S. Geological Survey Geologic Quadrangle Map GQ-
103, scale 1:24,000.
Weimer, R.J. and Ray, R.R., 1997, Laramide mountain flank deformation
and the Golden fault zone, Jefferson County, Colorado, in Bolyard,
D.W. and Sonnenberg, S.A., eds., Geologic history of the Colorado
Front Range: Rocky Mountain Association of Geologists Field
Trip #7, p. 53.