2016

Colorado Design Snow Loads

Prepared by the Structural

Engineers Association of
Colorado (SEAC) Snow Load
Committee
California State University,
 Colorado Design Snow Loads
Structural Engineers Association of Colorado

Prepared By
The SEAC Snow Load Committee
D. Jared DeBock, Ph.D.
James Robert Harris, P.E., Ph.D.
Abbie B. Liel, Ph.D., P.E.
Robert M. Pattillo, P.E.
Jeannette M. Torrents, P.E., Chair

April 2016

© Copyright 2016 by the Structural Engineers Association of Colorado

Colorado Design Snow Loads i

Executive Summary
The Structural Engineers Association of Colorado (SEAC) has developed new ground snow
loads recommended for use in structural design in Colorado. The recommended loads are based
on a suite of over 300 snow sites in the state of Colorado for which data have been assembled
from historical records of ground snow loads and depth. Philosophically, these loads represent a
transition from a uniform hazard approach to a uniform risk approach for determining design
snow loads. Previous design snow loads, including those in the 2007 SEAC report, are based on
a uniform hazard, i.e. the design value at a site is taken as the snow load that occurs, on average,
once every 50 years (in other words, the mean recurrence interval is 50 years). However, due to
the large variations in climate throughout Colorado, new studies conducted by this committee
have shown that using the 50-year load as the basis for design does not provide uniform
protection against the risk of roof failure. In particular, at sites on the plains, roof design based
on the 50-year ground snow load may not achieve the reliability levels targeted by the national
standard for loads, ASCE/SEI 7, meaning that the probability of failure may be higher than
expected. The converse is true in mountainous regions, where the reliability achieved when the
50-year load is used in design may actually be greater than required by ASCE 7. These patterns
stem from differences in snowfall and accumulation throughout the state. In the mountains, snow
builds up over the course of the winter, producing large ground snow loads, but fairly low
variability in the loads from year to year. Snow loads on the plains tend to be dependent on large
single-storm events and are highly variable from year to year.

The computation of loads for a target probability of failure results in loads appropriate for
strength-based structural design. However, for convenience of use with ASCE 7, the loads
presented in this report have been divided by the load factor (1.6) prescribed for snow loads in
the ASCE 7. Thus, the loads are also appropriate for use with allowable stress design methods
and for serviceability checks.

Compared to the 2007 recommended loads, there have been small decreases in design loads in
the mountains (approximately 10 percent), and increases in recommended design loads on the
plains (approximately 50 percent to100 percent). The increased design loads on the plains are
consistent with historic and current practice, where design roof loads of 20 to 30 psf have
typically been used instead of the 50-year values. A number of other improvements have been
made. First, the new analysis utilizes approximately seven additional years of weather data
generated since the development of the 2007 map. Second, improved relationships for snow
depth and snow weight have been used. Third, a method of using all the data for large regions of
the state to better predict rare events has been incorporated. Finally, the snow loads have been
correlated with the altitude of the site and maps have been produced that provide parameters to
compute the ground snow load given the altitude of the site. The mapping process introduces a
geographic smoothing, or averaging, on top of the reliability-targeted loads, which reduces some
of the “noise” inherent in the historical records. The recommended design loads are those
determined from the mapping process.

This report documents: (a) new mapped values, including comparison to previously used values,
(b) snow load and depth data used in the development of the maps, and (c) probabilistic
approaches used to evaluate roof reliability and conduct spatial smoothing. The report also
provides recommendations for use of these loads with ASCE 7 and for future improvements.

Colorado Design Snow Loads ii

Table of Contents
EXECUTIVE SUMMARY ........................................................................................................................................II
TABLE OF CONTENTS ......................................................................................................................................... III
PART 1. RECOMMENDED DESIGN SNOW LOADS ..........................................................................................1
1.1. INTRODUCTION AND MOTIVATION ......................................................................................................................1
1.1.1. The Need for Revised Design Snow Load Recommendations and a New Map ..........................................1
1.1.2. Objective of New Reliability-Based Design Snow Loads ...........................................................................1
1.2. RECOMMENDED DESIGN SNOW LOADS FOR COLORADO .....................................................................................2
1.2.1. Basis for Use ..............................................................................................................................................2
1.2.2. Design Snow Loads ....................................................................................................................................2
1.2.3. Additional Recommendations ................................................................................................................... 16
1.3. FUTURE WORK ................................................................................................................................................. 17
1.4. ACKNOWLEDGMENTS ....................................................................................................................................... 17
PART 2. DEVELOPMENT OF 2015 DESIGN SNOW LOADS FOR COLORADO ......................................... 18
2.1. DATA ACQUISITION AND PROCESSING ............................................................................................................... 18
2.1.1. Data Sources ............................................................................................................................................ 18
2.1.2. Initial Data Processing ............................................................................................................................ 19
2.2. METHODOLOGY TO COMPUTE RELIABILITY-TARGETED GROUND SNOW LOADS ............................................. 20
2.2.1. Test Structure Design ............................................................................................................................... 20
2.2.2. Reliability Analysis Overview .................................................................................................................. 21
2.2.3. Uncertain Demand Variables................................................................................................................... 21
2.2.4. Uncertain Capacity Variables .................................................................................................................. 22
2.2.5. Example Reliability Assessments ............................................................................................................. 22
2.3. MAPPING RELIABILITY-TARGETED GROUND SNOW LOADS ............................................................................. 26
2.3.1. Overview .................................................................................................................................................. 27
2.3.2. Importance of Correlations of Snow Loads with Altitude ........................................................................ 27
2.3.3. Correlations of Snow Loads with Altitude ............................................................................................... 28
2.3.4. Methods for Mapping the Parameter K ................................................................................................... 29
2.3.5. Expert Judgment....................................................................................................................................... 29
2.3.6. Evaluation of the Proposed Snow Map .................................................................................................... 30
2.4. COMPARISONS OF 2015 RECOMMENDED LOADS WITH PREVIOUS PROVISIONS ................................................. 31
2.4.1. Comparison of Reliability-Targeted Design Ground Snow Loads to 50-year Ground Snow Loads ........ 31
2.4.2. Comparison with Prior SEAC Recommendations .................................................................................... 32
2.5. CONCLUSIONS ................................................................................................................................................... 34
2.6. WORKS CITED................................................................................................................................................... 34
APPENDIX 1. DEVELOPMENT OF CONVERSIONS BETWEEN SNOW DEPTH AND SNOW WEIGHT
..................................................................................................................................................................................... 36
A1.1. INTRODUCTION .............................................................................................................................................. 36
A1.2. COMPACTED SNOW AND SETTLED SNOW....................................................................................................... 36
A1.3. COMPACTED SNOW SITES .............................................................................................................................. 36
A1.4. SETTLED SNOW SITES .................................................................................................................................... 37
A1.5. INTERMEDIATE SNOW SITES .......................................................................................................................... 38
A1.6. TREATMENT OF UNCERTAINTY/VARIABILITY IN THE DEPTH-TO-WEIGHT CONVERSIONS.............................. 39
A1.7. WORKS CITED ................................................................................................................................................ 40
APPENDIX 2. DEVELOPMENT OF GROUND TO ROOF WEIGHT CONVERSION FOR MONTE
CARLO SIMULATION ............................................................................................................................................ 41
A2.1. INTRODUCTION .............................................................................................................................................. 41
A2.2. ROOF AND GROUND SNOW LOADS DATA SET ............................................................................................... 41
A2.3. FITTING THE PROBABILISTIC MODEL ............................................................................................................. 42
A2.3.1. Step 1: Non-parametric Characterization of the Ground-to-Roof Snow Weight Conversion ................ 42

Colorado Design Snow Loads iii

 A2.3.2. Step 2: Fitting the Analytical Ground-to-Roof Snow Weight Conversion Model .................................. 43
A2.4. DISCUSSION AND LIMITATIONS ...................................................................................................................... 44
A2.5. CONCLUSIONS ................................................................................................................................................ 45
A2.6. WORKS CITED ................................................................................................................................................ 45
APPENDIX 3. DETERMINING PROBABILITY DISTRIBUTIONS FOR SNOW SITES .............................. 47
A3.1. INTRODUCTION .............................................................................................................................................. 47
A3.2. SITE-SPECIFIC DISTRIBUTION FITTING ........................................................................................................... 47
A3.2.1. Background ............................................................................................................................................ 47
A3.2.2. Tail-Fitting Approach ............................................................................................................................ 48
A3.2.3. Limitations of the Site-Specific Tail-Fitting Approach .......................................................................... 49
A3.3. DISTRIBUTION TAIL SHAPES INFORMED BY MULTIPLE SNOW SITES .............................................................. 52
A3.3.1. Accounting for Systematic Biases in the Distributions that are Fitted to the Clustered Data Sets........ 55
A3.3.2. Approach for Clustering Sites with Similar Distribution Shapes .......................................................... 61
A3.4. DISCUSSION AND CONCLUSIONS .................................................................................................................... 63
A3.5. WORKS CITED ................................................................................................................................................ 64
APPENDIX 4. RECOMMENDATIONS FOR SERVICEABILITY LOADS AND IMPORTANCE FACTORS
..................................................................................................................................................................................... 65
A4.1. INTRODUCTION .............................................................................................................................................. 65
A4.2. SERVICEABILITY LOADS ................................................................................................................................ 65
A4.3. SNOW LOAD IMPORTANCE FACTORS ............................................................................................................. 66
APPENDIX 5. TABULATED DATA SUMMARY ................................................................................................ 69
A5.1. INTRODUCTION .............................................................................................................................................. 69
A5.2. RAW DATA SUMMARY ................................................................................................................................... 69
A5.3. ASSEMBLED SNOW SITE DATA SUMMARY ..................................................................................................... 85
APPENDIX 6. COMPARISON OF 2015 DESIGN GROUND SNOW LOADS TO PREVIOUS EDITIONS . 99

Colorado Design Snow Loads iv

Part 1. Recommended Design Snow Loads
1.1. Introduction and Motivation
This report documents the basis for and motivation behind the 2015 SEAC design snow load
recommendations, as well as the methods by which they are developed. The new snow load map
is based on a detailed statistical analysis that aims to achieve uniform resistance against snow
loads throughout the state of Colorado.

1.1.1. The Need for Revised Design Snow Load Recommendations and a New
Map
The design snow loads described in this report revise the existing maps for Colorado in a number
of ways. First, it incorporates new snow data not available in previous analyses. Second, the new
map smooths design values between nearby sites, eliminating large jumps between adjacent
communities, and provides updated guidance on the determination of snow loads as a function of
altitude that is based on the unique climatological conditions in our state.

Finally, and most importantly, the new map revises design snow loads to ensure that the risk of
roof failure due to snow loading, or, alternatively, the structural safety and reliability for snow
loading, are consistent across the state. This report quantifies safety in terms of the reliability
index (or safety index) β, which is inversely related to the probability of failure; a higher safety
index corresponds to lower probability of failure. The national standard for loads, ASCE/SEI7
(hereafter ASCE 7), targets a safety index of 3.0 (0.13 percent probability of failure in 50 years)
for failure mechanisms that are not sudden and do not lead to widespread progression in ordinary
occupancy (Risk Category II) buildings under snow loads1. Previously, designing roofs for a 50-
year snow load was assumed to satisfy these criteria, but analyses show that a 50-year load is
conservative in some areas of Colorado and non-conservative in others. The SEAC 2007 design
snow load recommendations, which are based on a uniform hazard of a 50-year mean recurrence
interval (MRI) snow load, produce safety indices ranging from approximately β = 2 on the plains
(2.3 percent probability of failure in 50 years) to β = 3.5 in the mountains (0.02 percent
probability of failure in 50 years) for ordinary occupancy buildings. The probability of failure is
higher on the plains because of the larger variability in annual maximum snow loads on the
plains as compared to the mountains, where snowfall and accumulation are more consistent from
year to year. The large variability in annual maximum snow loads means that plains sites are
more likely to experience loads that are significantly greater than their 50-year loads during the
life of the structure, causing roof failures. These discrepancies in safety and building
performance across the state of Colorado are in need of resolution.

1.1.2. Objective of New Reliability-Based Design Snow Loads

The objective of the new snow load recommendations is to achieve uniform safety for snow
loading across the state of Colorado. The target safety objective of a reliability index of β = 3 is
appropriate for these recommendations; material design standards address “failure limit states

1
The target of β = 3.0 is found in Table C3.1.3.1a of ASCE 7-10. These targets will be moved to the body of the
standard in future editions.

1
that are sudden or lead to wide-spread progression of damage through resistance factors and
other provisions.” ASCE 7 targets other safety levels for different Risk Categories and includes
load adjustment factors for other Risk Categories, and this report includes recommendations to
modify those adjusting factors to achieve consistent reliability targets. The design loads
recommended in this report are developed to achieve the target safety objective, an approach that
is fundamentally different than current practice, which is to design for a uniform hazard (e.g. a
50-year snow load). As a result, the recommended design ground snow loads are decreased in the
mountains and increased on the plains of Colorado to meet the uniform safety objective. The
concept of targeting a consistent safety or reliability rather than consistent hazard is not new to
the Structural Engineering profession. For example, Ellingwood and Tekie (1999) address the
need to target uniform reliability with regard to wind loading and the ASCE 7-10 Maximum
Considered Earthquake (MCE) values are risk-based rather than hazard-based (Luco et al. 2007).

1.2. Recommended Design Snow Loads for Colorado

1.2.1. Basis for Use
The recommended design ground snow loads provided in this report are intended to be used with
the ASCE 7 procedures for structural design. As such, they provide design ground snow loads
which must then be converted to roof snow loads according to the ASCE 7 procedures, with
appropriate treatment of exposure to wind, thermal properties of the roof, roof slope, unbalanced
snow loads, drifting, and risk categories. For convenience, the recommended loads are provided
in both map and tabular form. It is the responsibility of the engineer to check with the local
authority having jurisdiction to verify code-required design snow loads before using these
recommendations. This report is a recommendation, not a building code.

1.2.2. Design Snow Loads

Tabulated Design Values
The recommended design ground snow loads are tabulated for many locations in the state of
Colorado, see Table 1.1. Many communities in relatively flat areas are covered by entries in the
table defining portions of counties, suburban areas, or the Eastern Plains. These values may be
used to compute design loads at nearby sites, i.e. those sites that are (a) within five miles, (b)
within 1000 ft. of altitude, and (c) not separated by a mountain ridge from the tabulated location,
according to the procedures described below. If a site is not near enough to the tabulated site to
use these adjustments, the mapped design values should be used (described below).

In most parts of the state (except for locations east of the Rocky Mountains and below 6500 ft.
altitude), if the site of interest is higher altitude than a nearby tabulated site, the design ground
snow load should be adjusted from the tabulated values using Equation 1.1.

, = , ∗ 1.1,

where pg,site and pg,tabulated are the design ground snow load at the building site and at the
tabulated location, respectively, and Asite and Atabulated are the altitudes of the building site and the
tabulated location in thousands of ft., respectively. If the site of interest is lower in altitude than

Colorado Design Snow Loads 2

the tabulated location, the tabulated load may be used without modification. Many communities
extend over land with greatly varying altitudes, so it is important to compare the altitude of the
site with the tabulated altitude.

For locations East of the Rocky Mountains and below 6500 ft. altitude, the tabulated loads may
be used without modification for sites lower in altitude than the tabulated location and also up to
250 ft. higher in altitude than the tabulated location. Otherwise, refer to Equation 1.2 in the
section entitled Mapped Design Values to determine the design ground snow load.

The development of the design snow load values is described in detail in Part 2 of this report.

Colorado Design Snow Loads 3

Table 1.1a Tabulated design ground snow loads (pg) for the state of Colorado (Cities/Towns)
Design Ground
Latitude Longitude Altitude Snow Load (pg)
City/Town County (deg) (deg) (ft) (psf)*
Aguilar Las Animas 37.40 -104.65 6390 55
Air Force Academy El Paso 39.01 -104.89 7000 55
Alamosa Alamosa 37.47 -105.87 7540 25
Allenspark Boulder 40.19 -105.53 8500 70
Alma Park 39.28 -106.06 10360 65
Antonito Conejos 37.08 -106.01 7890 40
Arapaho ski area base Summit 39.64 -105.87 10850 125
Arvada Jefferson 39.80 -105.09 5350 40
Aspen Pitkin 39.19 -106.82 7890 75
Aurora Adams 39.73 -104.83 5400 40
Avon Eagle 39.63 -106.52 7430 60
Bailey Park 39.41 -105.47 7740 80
Basalt Eagle 39.37 -107.03 6610 55
Battlement Mesa Mesa 39.44 -108.03 5490 35
Bayfield La Plata 37.23 -107.60 6900 55
Beaver Creek Eagle 39.60 -106.52 8080 75
Bellvue Larimer 40.63 -105.17 5130 40
Beulah Pueblo 38.08 -104.99 6380 55
Black Forest El Paso 39.01 104.70 7370 65
Black Hawk Gilpin 39.80 -105.49 8540 85
Blue River Summit 39.38 -106.05 10500 90
Boulder Boulder 40.01 -105.27 5330 40
Branson Las Animas 37.02 -103.88 6270 55
Breckenridge Summit 39.50 -106.04 9600 80
Brighton Adams 39.99 -104.82 4980 35
Broomfield Broomfield 39.92 -105.09 5390 40
Buena Vista Chaffee 38.84 -106.13 7960 35
Canon City Fremont 38.44 -105.24 5350 35
Carbondale Garfield 39.40 -107.21 6170 50
Cascade El Paso 38.90 -104.97 7380 60
Castle Pines Douglas 39.47 -104.89 6370 50
Castle Rock Douglas 39.37 -104.86 6220 45
Cedaredge Delta 38.90 -107.93 6230 35
Centennial Arapahoe 39.58 -104.88 5830 40
Central City Gilpin 39.80 -105.51 8510 85
Cimarron Hills El Paso 38.86 -104.70 6450 45
Coal Creek Canyon Boulder 39.92 -105.40 8600 105
Collbran Mesa 39.24 -107.96 5980 45

Colorado Design Snow Loads 4

Table 1.1a Tabulated design ground snow loads (pg) for the state of Colorado (Cities/Towns)
Design Ground
Latitude Longitude Altitude Snow Load (pg)
City/Town County (deg) (deg) (ft) (psf)*
Colorado Springs
El Paso 38.83 -104.82 6010 45
(downtown)
Colorado Springs (Chapel
El Paso 38.95 -104.79 6500 50
HIlls)
Commerce City Adams 39.81 -104.93 5160 35
Conifer Jefferson 39.52 -105.31 8280 100
Copper Mountain ski area
Summit 39.51 -106.14 9700 80
base
Cortez Montezuma 37.35 -108.59 6190 30
Craig Moffat 40.52 -107.55 6200 30
Crawford Delta 38.70 -107.61 6560 45
Creede Mineral 37.85 -106.93 8800 65
Crested Butte Gunnison 38.87 -106.98 8910 125
Crestone Saguache 38.00 -105.70 7930 35
Cripple Creek Teller 38.75 -105.18 9490 70
De Beque Mesa 39.33 -108.22 4950 30
Del Norte Rio Grande 37.68 -106.35 7880 30
Denver Denver 39.74 -104.98 5280 35
Dillon Summit 39.63 -106.04 9110 65
Dinosaur Moffat 40.24 -109.01 5920 35
Dolores Montezuma 37.47 -108.50 6940 55
Dotsero Eagle 39.65 -107.07 6150 40
Dove Creek Dolores 37.77 -108.91 6840 45
Durango La Plata 37.28 -107.88 6530 55
Durango Mountain ski area
La Plata 37.63 -107.81 9000 125
base
Eagle Eagle 39.66 -106.83 6600 45
Edwards Eagle 39.64 -106.59 7220 55
El Jebel Eagle 39.40 -107.09 6480 50
Elbert Elbert 39.22 -104.54 6720 55
Empire Clear Creek 39.76 -105.68 8620 60
Estes Park Larimer 40.38 -105.52 7520 65
Evergreen Jefferson 39.63 -105.32 7050 70
Fairplay Park 39.22 -106.00 9950 55
Fort Carson (HQ) El Paso 38.74 -104.79 5800 40
Fort Collins Larimer 40.59 -105.08 5000 35
Fort Garland Costilla 37.43 -105.43 7940 25
Fort Morgan Morgan 40.25 -103.80 4330 30
Fountain El Paso 38.68 -104.70 5550 35

Colorado Design Snow Loads 5

Table 1.1a Tabulated design ground snow loads (pg) for the state of Colorado (Cities/Towns)
Design Ground
Latitude Longitude Altitude Snow Load (pg)
City/Town County (deg) (deg) (ft) (psf)*
Franktown Douglas 39.39 -104.75 6160 45
Fraser Grand 39.94 -105.82 8580 75
Frisco Summit 39.57 -106.10 9080 65
Genesee Jefferson 39.70 -105.27 7650 80
Georgetown Clear Creek 39.71 -105.70 8520 60
Gleneagle El Paso 39.05 -104.83 6930 55
Glenwood Springs Garfield 39.55 -107.32 5760 40
Granby Grand 40.09 -105.94 7980 55
Grand Junction Mesa 39.06 -108.55 4590 25
Grand Lake Grand 40.25 -105.82 8390 70
Greeley Weld 40.42 -104.71 4680 30
Green Mountain Falls Teller 38.93 -105.02 7760 70
Gunnison Gunnison 38.55 -106.93 7700 45
Gypsum Eagle 39.65 -106.95 6310 40
Hartsel Park 39.02 -105.80 8870 35
Hayden Routt 40.50 -107.26 6350 55
Highlands Ranch Douglas 39.54 -104.97 5900 45
Hot Sulphur Springs Grand 40.07 -106.10 7730 45
Howard Fremont 38.45 -105.84 6720 40
Idaho Springs Clear Creek 39.74 -105.51 7530 60
Ignacio La Plata 37.12 -107.63 6450 45

Keystone ski area base Summit 39.60 -105.95 9170 70

Kremmling Grand 40.06 -106.39 7310 40

La Junta Otero 37.99 -103.54 4080 30
La Veta Huerfano 37.51 -105.01 7040 50
Lake City Hinsdale 38.03 -107.32 8660 55
Lakewood Jefferson 39.70 -105.08 5520 40
Lamar Prowers 38.09 -102.62 3620 30
Larkspur Douglas 39.23 -104.89 6730 55
Leadville Lake 39.25 -106.29 10160 75
Littleton Arapahoe 39.61 -105.02 5350 40
Livermore Larimer 40.79 -105.22 5900 45
Longmont Boulder 40.17 -105.10 4980 35
Loveland Larimer 40.40 -105.07 4980 35
Loveland ski area base Clear Creek 39.68 -105.90 10850 125
Lyons Boulder 40.22 -105.27 5370 40
Mancos Montezuma 37.34 -108.29 7030 45
Manitou Springs El Paso 38.86 -104.92 6360 50

Colorado Design Snow Loads 6

Table 1.1a Tabulated design ground snow loads (pg) for the state of Colorado (Cities/Towns)
Design Ground
Latitude Longitude Altitude Snow Load (pg)
City/Town County (deg) (deg) (ft) (psf)*
Marble Gunnison 39.07 -107.19 7990 90
Meeker Rio Blanco 40.04 -107.91 6240 40
Mesa Mesa 39.17 -108.14 5640 35
Mesa Verde Montezuma 37.15 -108.52 6770 50
Minturn Eagle 39.59 -106.43 7860 70
Monarch ski area base Chaffee 38.51 -106.33 10800 135
Monte Vista Rio Grande 37.58 -106.15 7660 25
Montezuma Summit 39.58 -105.87 10310 105
Montrose Montrose 38.48 -107.88 5810 25
Monument El Paso 39.09 -104.87 6980 60
Mount Crested Butte ski
Gunnison 38.90 -106.97 9900 155
area base
Mountain Village San Miguel 37.93 -107.86 9600 120
Nederland Boulder 39.96 -105.51 8230 70
Newcastle Garfield 39.57 -107.54 5600 35
Norwood San Miguel 38.13 -108.29 7010 35
Nucla Montrose 38.27 -108.55 5790 25
Oak Creek Routt 40.28 -106.96 7430 70
Ophir San Miguel 37.86 -107.83 9700 125
Ouray Ouray 38.02 -107.67 7790 65
Pagosa Springs Archuleta 37.27 -107.01 7130 75
Palmer Lake El Paso 39.11 -104.91 7300 65
Paonia Delta 38.87 -107.59 5680 35
Parker Douglas 39.52 -104.76 5870 45
Pitkin Gunnison 38.61 -106.52 9220 105
Poncha Springs Chaffee 38.51 -106.08 7470 45
Ponderosa Park Elbert 39.40 -104.64 6680 55
Pueblo Pueblo 38.25 -104.61 4690 30
Pueblo West Pueblo 38.33 -104.74 4960 35
Rangely Rio Blanco 40.09 -108.80 5230 35
Red Cliff Eagle 39.51 -106.37 8750 85
Rico Dolores 37.69 -108.03 8830 100
Ridgway Ouray 38.15 -107.76 7050 40
Rifle Garfield 39.53 -107.78 5350 40
Roxborough Park Douglas 39.46 -105.08 6000 45
Rye Pueblo 37.92 -104.93 6800 60
Salida Chaffee 38.53 -106.00 7080 45
San Luis Costilla 37.20 -105.42 7980 30
Sawpit San Miguel 37.99 -108.00 7590 55

Colorado Design Snow Loads 7

 Table 1.1a Tabulated design ground snow loads (pg) for the state of Colorado (Cities/Towns)
Design Ground
Latitude Longitude Altitude Snow Load (pg)
City/Town County (deg) (deg) (ft) (psf)*
Security-Widefield El Paso 38.74 -104.72 5800 40
Sedalia Douglas 39.44 -104.96 5840 50
Silt Garfield 39.55 -107.66 5460 35
Silver Cliff Custer 38.14 -105.45 7990 55
Silver Plume Clear Creek 39.70 -105.73 9100 70
Silverthorne Summit 39.63 -106.07 8760 65
Silverton San Juan 37.81 -107.66 9310 105
Snowmass Village Pitkin 39.21 -106.94 8210 90
South Fork Rio Grande 37.67 -106.64 8210 70
Steamboat Springs Routt 40.48 -106.83 6730 85
Sterling Logan 40.63 -103.21 3940 30
Telluride San Miguel 37.94 -107.81 8790 75
The Pinery Douglas 39.44 104.74 6250 50
Thornton Adams 39.87 -104.97 5350 40
Trinidad Las Animas 37.17 -104.50 6030 45
Vail Eagle 39.64 -106.37 8190 90
Vail, Mid- Mountain Eagle 39.61 -106.37 10300 175
Victor Teller 38.71 -105.14 9710 80
Walden Jackson 40.73 -106.28 8100 45
Walsenburg Huerfano 37.62 -104.78 6170 45
Ward Boulder 40.07 -105.51 9150 75
Westcliffe Custer 38.13 -105.47 7870 50
Westminster Adams 39.84 -105.04 5380 40
Winter Park Grand 39.89 -105.76 9050 100

Wolf Creek ski area base Mineral 37.47 -106.79 10650 295

Woodland Park Teller 38.99 -105.06 8480 85

Woodmore El Paso 39.11 -104.85 7240 65
Yampa Routt 40.15 -106.91 7880 60
*Tabulated loads must be adjusted by Equation 1.1 where site altitude is greater than the tabulated
altitude

Colorado Design Snow Loads 8

Table 1.1b Tabulated design ground snow loads (pg) for the state of Colorado (groups of communities)

Groups of Communities Design Ground Snow Load (pg) (psf)*

Communities east of Interstate Highway 25 and not otherwise tabulated
- below 4750 feet 30
- 4750 feet or above, but below 5250 feet 35
- 5250 feet or above, but below 5750 feet 40
- 5750 feet or above, but below 6250 feet 45
- 6250 feet or above, but below 6500 feet 50
Denver suburban communities
- in Adams County, more than 1 mile west of S. Platte River 40
- in Adams County, more than 2 miles east of S. Platte River 40
- in Adams County, elsewhere 35
- in Arapahoe County, except east of Tower Road 40
- in Broomfield County 40
- in Douglas County, within 3 miles of Arapahoe County 45
- in Jefferson County more than 1 mile east of foothills 40
Boulder County Communities not otherwise tabulated
- more than 1 mile east of foothills and north of Baseline Rd 35
- more than 1 mile east of foothills and south of Baseline Rd 40
Conejos County Communities not otherwise tabulated
- below 7700 feet 25
Costilla County Communities not otherwise tabulated
- below 7800 feet 25
Delta County Communities not otherwise tabulated
- below 5500 feet 25
Fremont County Communities not otherwise tabulated
- below 5400 feet 35
Larimer County Communities not otherwise tabulated
- more than 1 mile east of foothills and south of Co. Rd 66 35
Mesa County Communities not otherwise tabulated
- below 5000 feet 25
Montrose County Communities not otherwise tabulated
- below 5500 feet 25
Pueblo County Communities not otherwise tabulated
- below 4750 feet 30
Saguache County Communities not otherwise tabulated
- below 7800 feet 25
Weld County Communities not otherwise tabulated
- West of Interstate Highway 25 35
- East of Interstate Highway 25 refer to the general rules above
*Tabulated loads must be adjusted by Equation 1.1 where site altitude is greater than the tabulated
altitude

Colorado Design Snow Loads 9

 Mapped Design Values
In addition to the tabulated design ground snow loads, this report also provides a parameter map
for determining design ground snow loads, which is included at the end of this section.

In this map, the state is divided into three regions: (1) east of the Rocky Mountains below 6500
ft. altitude, (2) the eastern slopes of the Rocky Mountains above 6500 ft., and (3) the remainder
of the state.

East of the Rockies, below 6500 ft. altitude (i.e. the first region), Equation 1.2 should be used to
compute the ground snow loads:

, = 10 − 15 , 30 1.2,

In the second and third regions, a parameter K is mapped for determining design ground snow
loads. Design ground snow loads are determined based on K by Equation 1.3:

, = ∗ , 1.3,
100
where pg,site is the design ground snow load and Asite is the altitude of the building site in
thousands of feet. MIN is 50 psf for the eastern slopes of the Rocky Mountains above 6500 ft.
(i.e. region 2) and 25 psf for areas not on the eastern slope. Equation 1.3 is only applicable
below 11,500 ft. altitude, where contours of the parameter K are mapped. For sites between
contours K should be interpolated between the nearby contours. When determining the value of
K from the map, it is important to note that the map is partitioned along mountain ridges, where
the contour lines for the parameter K are interrupted. Therefore, when interpolating between
contours on the map for the value of K at a location, treat such discontinuities as a boundary and
extrapolate from contours on the same side of the ridge. Many, but not all, of these
discontinuities are locations above 11,500 ft. The detailed development of these equations and
the parameter map is discussed in Part 2 of this report.

Mapped vs. Tabulated Design Values

Most of the tabulated loads are determined directly from the parameter map and then rounded to
the nearest 5 psf. However, there are locations where the tabulated value is different than the
rounded value from the map. These differences occur where analysis of local snow data yields
results significantly different than the spatial averaging required to construct the map. Higher
values are shown in bold type and smaller values are shown in italic type. Spatial averaging was
allowed to control on the Eastern Plains, but local data was given extra weight in the
mountainous areas. For sites where the tabulated load is more conservative than the load
determined from the parameter map, the tabulated load (bold type) takes precedence over the
load from the parameter map. Rounding to the nearest 5 psf is not mandatory, and at locations
not governed by the boldfaced type in the table, values computed from the map may be used.

Colorado Design Snow Loads 10

 Figure 1.1a: Overview Map

Colorado Design Snow Loads 11

Colorado Design Snow Loads Figure 1.1b: North Central Colorado Parameter Map 12
for Determining Ground Snow Loads
Colorado Design Snow Loads Figure 1.1c: Northwest Colorado Parameter Map 13
for Determining Ground Snow Loads
Colorado Design Snow Loads Figure 1.1d: South Central Colorado Parameter Map 14
for Determining Ground Snow Loads
Colorado Design Snow Loads Figure 1.1e: Southwest Colorado Parameter Map 15
for Determining Ground Snow Loads
 1.2.3. Additional Recommendations
Loads for Use in Serviceability Checks
Roof snow loads that are determined from the design ground snow loads in this study are
considered sufficient for checking serviceability limit states. However, it should be noted that
given the focus on safety (strength) in development of these loads, the loads may have some
limitations for serviceability checks. Appendix 4 addresses serviceability loads in more detail.

Structures Assigned to Higher Risk Categories

ASCE 7 requires snow load importance factors of Is = 1.1 and Is = 1.2 for Risk Category III and
IV buildings, respectively. However, the increases in safety provided by these safety factors is
not consistent throughout Colorado, even when they are applied to reliability-targeted design
snow loads. ASCE 7 states that Risk Category III and IV buildings are intended to have safety
indices of β = 3.25 and β = 3.5, respectively (ASCE 2010). A study of snow load importance
factors that is described in Appendix A4.3 shows that the ASCE 7 importance factors are non-
conservative at low altitude locations and slightly conservative at high altitude locations in
Colorado. This trends occur because distributions of annual maximum snow loads at low altitude
sites (such as sites on the plains) tend to have longer/heavier tails. As are result, larger increases
in the design loads are required at low altitude sites to increase safety against snow-induced
failure than for high elevation sites where snow loading is less variable from year to year.

Therefore, the committee recommends that the snow loading importance factor (Is) be computed
by Equation 1.4 for Risk Category IV buildings. For Risk Category III buildings, Is should be
taken as the average of Equation 1.4 and the value 1.0. No change is recommended for the
importance factors for Risk Category I and II buildings.

! = 1.15 ≤ 1.66 − 0.056 ∗ ≤ 1.4 1.4,

where A is the site altitude in thousands of feet. Development of Equation 1.4 is documented in
Appendix A4.3.
Drifting on Plains and Other High Exposure Areas
The ASCE 7 recommendation for snow drift loads are derived from data obtained across a wide
range of climates. More extreme drifts are known to occur where prolonged and strong winds are
typical, and studies are currently underway to attempt to quantify this effect. This type of
behavior has been observed on the plains of Colorado, as well as sites above timberline. The
committee does not have enough information to make a quantitative recommendation, but does
make a general recommendation that engineers consider maximum roof drift widths greater than
the maximum roof drift widths specified in ASCE 7 for buildings with roof steps that are located
on exposed sites on the plains. This phenomenon occurs because large snow storms in those
regions are often accompanied by long duration winds (on the order of days) oriented in a
consistent direction. If the wind direction is perpendicular to the roof step, a large snow drift will
form. Designing for snow drift widths beyond what is required in ASCE 7 is ultimately at the
discretion of the engineer and the authority having jurisdiction.

Colorado Design Snow Loads 16

 Sites Above Timberline
This report does not recommend design ground snow loads for sites above timberline (or
approximately 11,500 ft. of altitude2), first because there are few historical snow data records for
locations above timberline, and, second, the relation between ground and roof loads3 is likely
very different at locations above timberline due to the extreme wind exposure. Therefore, flat
roof loads may well be smaller than recommended in ASCE 7, and drift profiles are likely more
extreme (i.e., the length of the drift for a given height of drift will be larger than recommended in
ASCE 7). Snow loads for structures at such sites should be based on knowledge of the local
climate and the judgment of the engineer and the authority having jurisdiction.

1.3. Future Work

The 2015 SEAC design ground snow load recommendations represent the best estimate of
reliability-targeted ground snow loads for the state of Colorado, given the tools and resources
currently available to the SEAC Snow Load Committee. However, there remain ways in which
the SEAC Snow Load Committee believes that these recommendations may be improved in the
future. The mapping process has introduced a significant amount of averaging of locally large
and small loads. Further improvement may be possible by accounting for topographic slope
aspect (i.e. direction) and inclination at mountain sites (DePaolo 2013). While accounting for
local topography is currently infeasible for a study of this scale, it likely will be feasible within
the foreseeable future, as additional data and analysis tools become available.

In addition to improving the recommendations for design snow loads, there are other
improvements that may be needed for snow design procedures more generally. Two areas of
possible improvement are the criteria for calculated snow drift loads on roofs with steps, ridges,
or valleys and the snow load duration factor for wood construction. Current snow drifting criteria
and duration factors are based on data sets that neither come from, nor are consistent with, snow
loading that is observed in portions of the state of Colorado. Therefore, Colorado-specific
analyses of snow drifting and load duration are important topics for future study.

1.4. Acknowledgments
The committee gratefully acknowledges Dick Cunningham for his extensive efforts in getting
this new edition of the SEAC snow loads started, Mike DePaolo for his innovative studies and
new techniques for classifying and determining ground snow loads, Bruce Ellingwood for his
counsel and advice on statistical analysis, and Derek Kozak and Dania Hussain for their work on
reliability assessment of buildings under snow loads. In addition, drafts of this report have been
reviewed by the members of SEAC and related experts in the fields of study; their review
comments are gratefully acknowledged.

2
Timberline is defined as 11,500 ft. altitude in this study, although the precise altitude varies across the state.
3
The ground to roof snow load conversion for the reliability assessments is discussed in Part 2 and in Appendix 2

Colorado Design Snow Loads 17

Part 2. Development of 2015 Design Snow Loads for Colorado
The goal of the newly proposed ground snow load design values is to achieve consistent
structural safety against snow-induced roof failure throughout the state of Colorado. Experience
with snow loads in Colorado indicates that the use of a 50-year mean recurrence interval load as
the specified nominal load on the ground, along with a snow load factor of 1.6 (LRFD) or a
constant safety factor (ASD), will not satisfy the basic safety objective in significant populated
areas of the state and may be unnecessarily costly in other areas. Therefore, the recommended
design ground snow loads are computed so that roof structures designed in accordance with
ASCE 7 will have a safety index (β) of 3 with respect to snow loading. Consequently, the
recommended ground snow loads are not necessarily 50-year or 30-year loads, as with previous
editions, but are instead calibrated to achieve a target safety (β=3). The mean recurrence
intervals of the recommended ground snow loads range between 20 and 300 years.

The target safety index of 3 is for relatively benign4 structural failures in ordinary Risk Category
buildings (Risk Category II). Target safety indices for buildings belonging to risk categories I,
III, and IV, are 2.5, 3.25, and 3.5, respectively (ASCE 2010). Adjustments in snow loads to
achieve the different reliability targets for different Risk Category buildings are achieved through
snow load importance factors that are associated with each Risk Category in ASCE 7. As
recommended in Section 1.2.3, the snow load importance factors for non-Risk Category II
buildings should be adjusted to meet the desired safety objectives. The basis for these adjusted
importance factors is described in Appendix 4.

The procedures for determining the proposed design ground snow loads are outlined as follows:
(1) Data acquisition and processing; (2) Reliability analysis; (3) Mapping of reliability based
ground snow loads.

2.1. Data acquisition and processing

2.1.1. Data Sources
The recommended design ground snow loads in this report are based on historical records of
annual maximum ground snow at 603 snow recording stations throughout the state of Colorado.5
The historical records are obtained from four sources: Snow Course stations (NRCS 2015),
SNOTEL (i.e. SNOwpack TELemetry) stations (NRCS 2015), National Weather Service (NWS)
cooperative observer (CO-OP) stations (WRCC 2015), and first order NWS stations. Appendix
5 tabulates summary statistics for all of the snow stations in this study.

Snow Course Stations

Snow Course stations consist of a series of ten locations along a course (typically about a half
mile long) at which a trained observer measures snow depth (inches) and weight in inches or
water (i.e. snow water equivalent, “SWE”) on a monthly basis (NRCS 2015). The measurements

4
“Relatively benign” means a failure limit state that gives warning (generally, meaning the failure is ductile) and
does not precipitate a widespread failure vertically or horizontally. It is the intent of ASCE 7 that adjustments in
safety for other types of limit states are accomplished through resistance factors in material design standards.
5
These stations are distinctly different from the town/city locations listed in Table 1.1 of the report.

Colorado Design Snow Loads 18

are obtained by plunging an aluminum pipe through the snow pack to the ground and measuring
the depth of snow around the outside of the tube and weight of the snow that is trapped inside it.
The average depth and weight for the ten locations is reported. Snow Course stations are
primarily located in mountainous areas above 8,500 ft. altitude in the state of Colorado. Data
from 177 Snow Course stations are collected from the Natural Resources Conservations Service
(NRCS) website for this study.

SNOTEL Stations
SNOTELs are automated stations that measure the weight of snow (SWE) via a fluid-filled
pillow that senses pressure from the snow on top of it (NRCS 2015). The data are electronically
transmitted to a central repository. SNOTEL stations are often located in remote mountainous
areas where accessibility is limited, and they have replaced a number of Snow Course stations
over the years. Many SNOTEL stations also have sensors to measure snow depth. Data from 117
SNOTEL stations are collected from the NRCS website for this study.

NWS CO-OP Stations

NWS CO-OP stations (referred to hereafter as “NWS” stations) record snow depth on a daily
basis (NWS 2015). However, snow weight is not recorded. For the purposes of this study, the
depth measurements at these stations are converted to snow weights, using the relationships
developed in Appendix 1. Data from 303 NWS stations are collected from the Western Regional
Climate Center (WRCC 2015).

First Order NWS Stations

There are six “First Order” stations in the state of Colorado: Alamosa San Luis Valley Regional
Airport, Colorado Springs Municipal Airport, Denver-Stapleton, Fort Collins, Grand Junction
Walker, and Pueblo Memorial Airport. These are NWS stations that report depth of snow on a
daily basis, but have historically reported daily snow weight (i.e. SWE) as well. For the six First
Order stations, the average number of snow years is 66 and the average number of years with
SWE data is 28.

2.1.2. Initial Data Processing

Data collected from the 603 stations are consolidated to create 463 historical records. The
consolidated records are referred to as “snow sites.” Snow sites are created by combining data
records from stations very near to each other for which snow accumulation patterns are expected
to be the same or very similar. Snow sites with less than 30 years of historical data are eliminated
from the basic analysis, because these records are considered to be insufficient for the statistical
analysis. There are 327 snow sites that have 30 or more years of data. An additional 60 sites with
18 to 29 years of data are used in some areas to inform the mapping process.

Combining Nearby Stations to Make Snow Sites

Data from nearby stations are combined so that longer records of annual maximum snow data are
available. The combinations of stations often result from newer stations that have replaced older
stations in nearly identical locations (e.g., a SNOTEL station that replaces a Snow Course

Colorado Design Snow Loads 19

Station). In some cases, the locations of the different stations that are combined to make a snow
site are different, but they are close enough that no substantial difference in snow accumulation
patterns is expected between them. Similar combinations of stations are already present in the
NWS data that is collected for this study. In particular, it is common for many NWS stations to
have had two to five locations over their lifetime, but only one continuous data record is
reported. The easily available data does not always make clear when a NWS station has changed
location, and at some stations (e.g. Pueblo) the NWS did not combine the record as the station
moved.

The following criteria are implemented to determine if snow stations are in close enough
proximity for their records to be combined: (1) plains stations (east of the Front Range and below
5,000 ft. altitude) within approximately twelve miles and 500 ft. altitude of each other, (2)
stations above 6,000 ft. altitude within approximately two miles and 300 ft. altitude of each
other, and (3) any other stations within approximately five miles and 300 ft. altitude of each
other. These criteria are based on observations of snow recordings at adjacent stations and the
judgment of the SEAC Snow Load Committee.

When a snow site consists of multiple stations that have data for the same year, direct weight
measurements are given priority over snow weights that are converted from depth measurements.
If there are multiple stations with the same data type in a given year (i.e., all having direct weight
measurements or all having depth measurements that are converted to weights), then the
maximum annual snow weight for that year is taken as the maximum from all of the contributing
stations.

2.2. Methodology to Compute Reliability-Targeted Ground Snow Loads

Reliability analyses are performed for each snow site to determine the design ground snow load
that will result in a safety index of β = 3 for roof snow loading on a test building designed
according to ASCE 7. The test structure is a simply supported, uniformly loaded steel roof girder
with lateral support that is controlled by flexure, which meets the definition of “relatively
benign” failure. The safety index β is computed through Monte Carlo simulation (Fishman
2006).

2.2.1. Test Structure Design

Reliability analyses are performed for a test structure6, which is a 30 ft. wide-flange steel roof
girder with 30 ft. of tributary width and 15 psf design dead load (D), including self-weight.
Design roof snow loads (S) are computed per ASCE 7 as 0.7 times the design ground snow load7,
with all other factors of the ASCE 7 method (exposure, thermal, importance, and slope) set equal
to 1.0, implying that the test structure is part of a roof of a heated building situated in a suburban
exposure with a partially exposed roof. The LRFD ultimate design load (U, Equation 2.1) and a

6
The authors conducted sensitivity studies showing that the reliability analysis results are insensitive to the roof
geometry and other design assumptions.
7
The 0.7 factor increases to 1.0 as the ground snow load decreases from 28.6 psf to 20 psf, in accordance with the
minimum snow load provisions of ASCE 7.

Colorado Design Snow Loads 20

strength reduction factor of 0.9 are used to determine the necessary plastic section modulus for
design. Serviceability and non-flexure limit states are not considered in the test structure design.

( = 1.2* + 1.6, 2.1

2.2.2. Reliability Analysis Overview

For each computation of β, 10 million Monte Carlo simulations of maximum annual snow load
are conducted.8 A direct analytical approach to computing the reliability was not appropriate
because of the multiple sources of uncertainty to consider in the reliability calculations.

Each simulation generates a random realization of demand on the roof and a random realization
of the capacity of the roof, considering uncertainties in these quantities as described in more
detail below. These values are compared to determine if the test structure fails in a given
realization. The total number of failures divided by the total number of simulations is the
expected annual rate of failure (λ). The probability of failure for a 50-year time period (Pf,50) is
then computed by the Poisson distribution in Equation 2.2 and converted to a reliability index β
by Equation 2.3:

-.,/0 = 1 − exp4−5056 ≈ 505 2.2

8 = Φ:; 41 − -.,/0 6 2.3,

where Φ:; is the inverse standard normal cumulative distribution function. The β defined by
ASCE 7 is based on a 50-year service life of the structure.

2.2.3. Uncertain Demand Variables

The roof demand is the sum of snow load demand and dead load demand. Snow load demand is
a product of two random variables: the annual maximum ground snow load and the ratio of roof
snow load to ground snow load.

Annual Maximum Ground Snow Load

Each Monte Carlo simulation generates a random annual maximum ground snow load from a
probability distribution of annual maximum ground snow loads for the site of interest. The
lognormal distribution is used to model annual maximum snow loads at every site. At each site,
the process for fitting the distribution emphasizes the upper tail, which is crucial for the
reliability analysis. The central tendency of the distribution is determined directly from data at
the site of interest. However, the historical record is too short to confidently predict the rare
events that cause failure from the data available at any one site. Therefore, the shape of the upper
tail is determined by combining data from 20 or more snow sites that are similar to the site of

8
If the theoretical value of the safety index is β = 3.0, it can be shown by constructing a confidence interval with a
binomial distribution that β computed from 10 million Monte Carlo simulations will be within ± 0.03 of the true
theoretical value approximately 90% of the time.

Colorado Design Snow Loads 21

interest and fitting a shared tail shape. Appendix 3 gives a detailed explanation of the process for
fitting the annual maximum snow load probability distributions at each site.

Ratio of Roof Snow Load to Ground Snow Load

Annual maximum ground snow loads are converted to roof snow loads by a ground to roof
conversion factor (GR), which represents the ratio of roof snow load to ground snow load.9 The
expected value of GR decreases as ground snow load increases, therefore it is a function of the
simulated ground snow load (pg,sim). GR is modeled by a lognormal distribution with median and
logarithmic standard deviation described by Equations 2.4 and 2.5, respectively:

<=> ? = 0.50 ∗ exp@−0.034 ∗ , AB + 0.4 2.4

CDE = min 4.007 ∗ , A + 0.1 , 0.336 2.5

Equations 2.4 and 2.5 are developed in Appendix 2.

Roof Dead Load Demand

The roof dead load demand is modeled by a normal distribution with a mean value of 1.05 times
the design dead load and COV of 0.1. This model is adopted from Ellingwood et al. (1982).

2.2.4. Uncertain Capacity Variables

The capacity of the test structure is a product of the steel yield strength and plastic section
modulus. The yield strength of the steel is modeled as 1.1 times the nominal yield strength of 50
ksi with a logarithmic standard deviation of 0.09, which is within the range of strength values for
rolled steel sections subjected to bending (e.g. as reported by a number of sources, Kennedy and
Gad Aly 1980, Bartlett et al. 2003, Schmidt and Bartlett 2002). The plastic section modulus
follows a normal distribution with a mean of 1.05 times the design plastic section modulus and
COV of 0.05 (Galambos and Ravindra, 1978; Lind, 1977). The 5 percent increase in section
modulus accounts for the average increase in section size that occurs when discrete steel sections
are selected (~3 percent) for design and for variability in compact section properties (~2 percent).

2.2.5. Example Reliability Assessments

Sample results for three snow sites, Copper Mountain, Denver-Stapleton, and Yampa, are shown
in this section (Figure 2.1 - Figure 2.6). These sites are selected for illustration to demonstrate
key differences between mountain sites with high annual snowfall, non-mountain sites with
sporadic snow fall, and intermediate snowfall sites. At high altitude mountain sites like Copper
Mountain10, where the altitude is 10,550 ft., it is common for the reliability-targeted ground snow
load (i.e., the design ground snow to achieve β=3) to be on the order of 10 percent lower than the
50 year ground snow load (Figure 2.1). For lower altitude non-mountain sites, such as Denver-
Stapleton, it is common that the reliability-targeted design ground snow load is on the order of

9
For statistical analysis, we use our best estimate for the ground to roof conversion, which should not be confused
with the ground to roof conversion factor of 0.7 that is used for the design step.
10
Note that the Copper Mountain altitude that is cited here is for the recording station on the slope of the mountain
and not at the population center that is reported in Table 1.1.

Colorado Design Snow Loads 22

75 percent higher than the 50-year ground snow load (Figure 2.3). Yampa is an example of a
snow site where the reliability-targeted ground snow load and the 50-year ground snow load are
approximately equivalent (Figure 2.5).

The reason that mountain sites such as Copper Mountain tend to have reliability-targeted loads
that are lower than their 50-year loads is because the distribution of annual maximum snow loads
have relatively light upper tails (e.g. Figure 2.2); this means that loads significantly larger than
their 50-year load are extremely rare, so the 50-year load times the 1.6 factor is too conservative
for design. For example, the 50-year load for the Copper Mountain site is 110 psf and the
logarithmic standard deviation that describes the tail of the distribution is approximately 0.20;
therefore, an ultimate design load of 1.6 times the 50-year load (176 psf) has a mean recurrence
interval of approximately 146,000 years, which is extremely rare.. A design ground snow load of
98 psf, 11 percent lower than the 50-year load, results in β=3 at the Copper Mountain snow site.
The new design point (1.6 times 98 psf = 157 psf) has a mean recurrence interval of
approximately 13,000 years.

The opposite is true for sites like Denver-Stapleton. High variability of annual maximum snow
loads at these sites results in heavy-tailed ground snow load distributions (e.g. Figure 2.4), which
means that loads significantly higher than the 50-year load are much less rare at these sites. For
example, the 50-year load at Denver-Stapleton is approximately 20 psf and the logarithmic
standard deviation that describes the upper tail shape of the distribution is 0.75. A design point of
1.6 times the 50-year load (32 psf) has a mean recurrence interval of 265 years and has actually
been observed in Denver within the last 100 years. A design point with such a short mean
recurrence interval is inadequate for the desired structural safety, which is why design ground
snow loads at heavy-tailed snow sites need to be significantly larger than their 50-year values. A
design ground snow load of 34 psf, 70 percent higher than the 50-year load, results in β =3 at the
Denver-Stapleton snow site. The new design point (1.6 times 34 psf = 54 psf) has a mean
recurrence interval of approximately 2,700 years. This design ground snow load results in a
design flat roof snow load of 25 psf, which is consistent with local practice.

A third location, Yampa (Figure 2.5 and Figure 2.6), is an example of a site whose annual
maximum ground snow load distribution has a shape that is consistent with the assumptions for
determining the ASCE 7 snow load factor of 1.6 for factoring 50-year snow loads. As seen in
Figure 2.5, the design ground snow load that leads to β=3.0 is 46 psf, which is very near the 50-
year load of 47 psf. The design point (1.6 x 46 = 74 psf) has a mean recurrence interval of
approximately 6,700 years.

Colorado Design Snow Loads 23

Figure 2.1. Reliability analysis results for the Copper Mountain snow site (altitude=10550 ft.). The reliability-
targeted design ground snow load is approximately 98 psf, which is 11 percent lower than the 50-year load of
approximately 110 psf.

Figure 2.2. Histogram of historical maximum annual ground snow loads at the Copper Mountain snow site
(altitude=10550 ft.). The fitted Lognormal Probability Density Function scaled to match the scale of the
histogram is overlain. The parameters of the distribution are: median=73 psf and logarithmic standard
deviation=0.20.

Colorado Design Snow Loads 24

Figure 2.3. Reliability analysis results for the Denver-Stapleton snow site (altitude=5290 ft.). The reliability-
targeted design ground snow load is approximately 34 psf, which is 70 percent higher than the 50-year load of
approximately 20 psf.

Figure 2.4. Histogram of maximum annual historical ground snow loads at the Denver-Stapleton snow site
(altitude=5290 ft.). The fitted Lognormal Probability Density Function scaled to match the scale of the
histogram is overlain. The parameters of the distribution are: median=4.3 psf and logarithmic standard
deviation=0.75.

Colorado Design Snow Loads 25

Figure 2.5. Reliability analysis results for the Yampa snow site (altitude=7860 ft.). The reliability-targeted
design ground snow of 46 psf is very similar to the 50-year load of approximately 47 psf.

Figure 2.6. Histogram of historical maximum annual ground snow loads at the Yampa snow site
(altitude=7860 ft.). The fitted Lognormal Probability Density Function scaled to match the scale of the
histogram is overlain. The parameters of the distribution are: median=25 psf and logarithmic standard
deviation=0.30.

2.3. Mapping Reliability-Targeted Ground Snow Loads

Colorado Design Snow Loads 26

 2.3.1. Overview
Reliability-targeted design ground snow loads for the state of Colorado are communicated via a
map and accompanying set of equations (see Section 1.2.2). The proposed map and equations are
intended to be used in conjunction with ASCE 7. Use of the map and equations requires
knowledge of the location (latitude and longitude) and altitude of a potential building site.

The eastern portion of the state sees little change in design ground snow loads with location.
Therefore, the discussion in this section mostly focuses on the remainder of the state where
altitude and snow loading are both highly variable.

West of the 6500 ft. altitude contour on the eastern slope of the Rockies, a parameter K is
mapped. The parameter K is combined with altitude to compute the design ground snow load,

= ∗ , !"# 2.6,
100
where pg is the design ground snow load at a building location in psf, A is altitude in thousands
of feet, K is a parameter determined from the proposed map, and MIN is 50 psf on the eastern
slope of the Rockies and 25 psf elsewhere. Mapped values of K provide a cleaner, simpler map
than if we mapped pg directly, due to big changes in altitude in some parts of the state. Equation
2.6 applies at altitudes below 11,500 ft. wherever values of K are given on the map. Areas above
11,500 ft. require expert judgment and/or site-specific analysis to determine their ground snow
loads, as described in Section 1.2.3.

2.3.2. Importance of Correlations of Snow Loads with Altitude

Correlations of design ground snow loads with altitude are a vital part of the mapping process,
because they provide a means for estimating snow loading at locations between snow sites where
historical snow data are available. This section illustrates by example that a contour map of
design ground snow loads is susceptible to severe misrepresentations of snow loading at
intermediate locations if it does not account for their altitudes.

Consider the two-dimensional graphic of an example region in Figure 2.7. The reliability-
targeted snow loads at the snow sites with data available (blue triangles) are 30 psf, 80 psf, and
40 psf, from left to right. Suppose that one is interested in determining the design ground snow
loads at the intermediate sites marked by the two red stars. A contour map constructed from the
calculated design ground snow loads would interpolate between the known points, predicting
design ground snow loads of 55 psf and 60 psf at the base of the mountain and the top of the
mountain, respectively. Both of those estimations are inaccurate. The most obvious problem in
this example is that the site at the top of the mountain (9500 ft. altitude) would have a ground
snow load 25 percent lower than the site that is part way up the mountain (8300 ft. altitude).
Therefore, the option of generating a state-wide contour map of design ground snow loads from
the 327 discrete snow sites, without explicitly accounting for altitude, is ruled out.

Now let us reconsider the same two intermediate sites in Figure 2.7, but account for their
altitudes. Using the proposed mapping approach, one can determine that the value of the
parameter K in Equation 2.6 is approximately 14 for the example region. This K value is based

Colorado Design Snow Loads 27

on the neighboring snow sites where snow data are available (blue triangles). Substituting the
altitudes of the two intermediate sites into Equation 2.6 produces design ground snow loads of 32
psf at the base of the mountain and 120 psf at the top. This is expected to be a more accurate
representation of the ground snow loading at the intermediate sites, because it follows a trend
that is observed at sites where data are available, i.e. design ground snow loads at sites in the
same general area are approximately proportional to altitude cubed. Therefore, mapping the
parameter K for Equation 2.6 is a more effective method for determining ground snow loads than
contours of ground snow loads that simply interpolate the load between snow sites.

Snow Site
pg≠60 psf
Intermediate Location
pg≈120 psf
Ground Elevation

pg=80 psf

pg≠55 psf pg=40 psf

pg≈32 psf
pg=30 psf

Figure 2.7. Two dimensional elevation view of three snow sites (blue triangles) and two intermediate sites (red
stars). Design ground snow loads (pg) at the intermediate sites are estimated by interpolation between the
snow sites with and without consideration of their altitudes. Estimates not accounting for altitude are pg≠XX
psf, and estimates that account for altitude are pg≈XX psf.

2.3.3. Correlations of Snow Loads with Altitude

The proposed map takes advantage of correlations between design snow loads and altitude. The
form of Equation 2.6, which relates altitude to design snow loads at locations west of the plains
through the parameter K was determined through a sensitivity study. This sensitivity study
divided the state into a number of sub-regions, showing that variations in reliability-targeted
ground snow loads at sites near to each other are approximately proportional to altitude cubed.
As an example, altitudes and reliability-targeted ground snow loads for sites from two sub-
regions in Southwestern Colorado are shown in Figure 2.8. The best-fit power curves show that
reliability-targeted loads are proportional to altitude to the powers 3.0 and 3.3, respectively. For
simplicity, the mapping equation (Equation 2.6) assumes that design snow loads vary with
altitude cubed (a power of 3.0) throughout the state, so that only one parameter, i.e. K, needs to
be mapped.

Colorado Design Snow Loads 28

Figure 2.8. Reliability-targeted design ground snow loads for two example sub-regions that are located in the
western portion of the state (left) and the southwestern portion of the state (right).

2.3.4. Methods for Mapping the Parameter K

The proposed map provides values of K to be used throughout much of the state. Contours of the
parameter K are mapped in Surfer (Golden Software 2015) using a method called, “inverse
distance to power.” The mapping process is performed by completing the following steps:

1) Compute the theoretical value of the map parameter K at each snow site (327 locations)
by rearranging Equation 2.6 and plugging in the altitude and reliability-targeted design
ground snow load.
2) Establish “Fault Lines,” which are lines of allowed discontinuity in the contour plot,
using expert judgment and local knowledge. These are mostly at mountain ridges that
separate areas with different snow accumulation patterns.
3) Establish a grid of approximately 50,000 points across the state, located at approximately
1.5 mile intervals, which provide the basis for the map.
4) For a grid point of interest, Surfer searches for the nearest 12 neighboring snow sites
without crossing fault lines.
5) A weighted average of the parameter K is computed for the grid point from the identified
12 neighboring snow sites. The contribution of each snow site to the weighted average is
inversely proportional to its distance from the point of interest to the power 2.0.
6) Steps 4 and 5 are repeated for every grid point in the state of Colorado and contours of K
are interpolated from the values at the grid points.

2.3.5. Expert Judgment

Colorado Design Snow Loads 29

The final stage of the mapping process is to refine the map by expert judgment. Primarily, this
involves: (1) manually smoothing the contours of the parameter K where knowledge of the local
climate and geography indicate that it should have little fluctuation, and (2) allowing the values
of K to fluctuate more than the mapping algorithm originally allowed in regions where snow
loads are known to fluctuate significantly across a small distance. The final refinements of the
map have little influence on its predictions of design ground snow loads at the majority of the
snow sites. However, notable changes in the contours due to manual edits do occur in a few
places. In regions near Aspen and Vail, manual edits increase the mapped values by
approximately 15 percent to bring the load that the map predicts up to a value that is comparable
to the reliability assessment for those towns. This occurs because other nearby sites have less
snow, so the original smoothing of the map forced Vail and Aspen snow load values lower than
they should be.

2.3.6. Evaluation of the Proposed Snow Map

The snow map is evaluated by comparing the map predictions of design ground snow loads to
those computed directly by the reliability analysis. Figure 2.9 shows these comparisons, broken
down by general regions of the state (for specific information about the four general regions,
refer to Figure A3.15 in Appendix 3). For each region, design ground snow loads from the map
are plotted against the design ground snow loads that are computed with reliability analysis. Data
points above the 45 degree lines (blue lines) are conservative, because the load from the map is
higher than the load computed from reliability analysis. Data points below the 45 degree lines are
non-conservative, because the load from the map is lower than the computed load. The scatter in
the plots reveals that a significant amount of aleatory uncertainty (noise) is smoothed out by the
mapping process. In other words, sites with unusually large loads tend to have mapped loads
lower than their loads from the reliability analysis, and vice versa for snow sites where no large

Colorado Design Snow Loads 30

snow loads have been observed.

Figure 2.9. Design ground snow loads from the map, plotted against design ground snow loads that are
computed directly from reliability analysis for the four main regions of the state of Colorado. Each plot has a
blue 45 degree line to indicate what a perfect match would be.

2.4. Comparisons of 2015 Recommended Loads with Previous Provisions

2.4.1. Comparison of Reliability-Targeted Design Ground Snow Loads to 50-year
Ground Snow Loads
The differences between the reliability-targeted ground snow loads and the 50-year ground snow
loads for the 327 snow sites that are used in this study are expressed as a ratio in Figure 2.10.
The reliability-targeted loads are approximately equal to the 50-year loads at sites located at
8000-8500 ft. altitude. At lower altitude sites, where distributions of annual maximum snow
loads are heavily skewed and tend to have heavy upper tails, the reliability-targeted design
ground snow loads are larger than the 50-year loads. The typical ratio of reliability-targeted loads
to 50-year loads increases from approximately 1.0 to 2.0 as altitude decreases from 8000 ft. to

Colorado Design Snow Loads 31

3500 ft.11 At mountainous sites above 8500 ft. altitude, the reliability-targeted ground snow
loads are approximately 90 percent of the 50-year loads.

Figure 2.10. Ratio of the computed reliability-targeted design ground snow load (for β=3) to the 50-year
ground snow load at the 327 snow sites in Colorado.

2.4.2. Comparison with Prior SEAC Recommendations

SEAC has recommended design snow loads twice before: SEAC (1971) and SEAC (2007). A
comparison of the 297 tabulated locations are provided in Appendix 6. Additionally,
comparisons of mapped and tabulated loads are shown for 17 example locations in Table 2.1.
The basis of the 1971 recommendations was a 30-year roof load. Since the 1971 version gives
roof loads directly, they are divided by a ground-to-roof conversion factor of 0.7 to be
comparable to design ground snow loads that are used with modern editions of ASCE 7. The
basis of the 2007 edition was 50-year ground snow loads.

In addition to a having a different basis for determining design snow loading, the two older
editions of the SEAC snow load recommendations utilize different methods of mapping those
loads. The 2007 snow load map shows contours of design ground snow loads, but it does not do
any spatial smoothing of those loads before generating the contours, nor does it fully account for
altitude. The 1971 version maps a parameter that must be combined with a site’s altitude to
compute the design snow load. This approach is similar to the proposed mapping method that is

11
This indicates that a large number of snow-induced failures should have been observed on the plains over the
course of the last century. That is not the case, however, because it has been common practice to design for roof
snow loads of 20 psf – 30 psf on the plains, despite the 50-year load being much lower.

Colorado Design Snow Loads 32

used in the 2015 recommendations, except that the parameter is determined by region rather than
from a contour map.

Lastly, more data are available now than were available at the time that previous editions of the
SEAC snow map were produced. However, we note that differences due to increases in available
data are less significant than the differences that are due to changes in the basis for computing
the design snow load values and the methods by which they are mapped.

General conclusions about the differences in recommended design snow loads between the three
editions of the snow map are made by comparing example design ground snow load values in
Table 2.1. The 2015 map tends to provide loads that are slightly lower than the 1971 map at most
locations. It also recommends loads that are lower than the 2007 map at high altitude locations,
but the loads are higher than the 2007 map at low altitude locations.

In addition to producing reliability-targeted design snow loads instead of hazard-targeted design

snow loads, a significant advantage of the 2015 snow load map over the 2007 map is its
capability to predict snow loading at locations where recorded snow data are not available, as
discussed in Section 2.3.2. Consider, for example, the cases of Carbondale and Colorado Springs
in Table 2.1. Carbondale and Colorado Springs are both similar to the hypothetical example site
at the foot of the mountain in Figure 2.7. Carbondale is situated between the Glenwood Springs
snow site (similar in altitude) and the Aspen snow site (much higher altitude). Colorado Springs
is situated between the Colorado Springs Municipal Airport snow site (similar altitude) and the
Ruxton Park snow site (much higher altitude). The 2007 snow load map assigned snow loads at
Carbondale and Colorado Springs that are between the snow loads at their neighboring sites,
even though they should actually be nearly the same as the neighboring snow site that has similar
altitude. As a result, the 2007 map overestimated the Carbondale and Colorado Springs snow
loads by about 40 psf, because the snow loads are interpolated between neighboring snow sites
without fully accounting for their altitudes. In the case of Colorado Springs, the 2007 map shows
that the load increases from 20 psf at the airport to 70 psf at downtown over a distance of only
six miles with less than 300 ft. of altitude gain. The 2015 edition of the SEAC snow map
significantly reduces this source of mapping error.

Table 2.1 Design ground snow loads (pg) at 17 example locations in Colorado

1971
1971 Map Tabulated
Lat. Lon. Alt. (Converted (Converted 2007 2007 2015 2015
City/Town (deg.) (deg.) (ft.) to pg) to pg) Map Tabulated Map Tabulated
Aspen 39.19 -106.82 7890 99 107 105 105 75 75
Boulder 40.01 -105.27 5330 45 43 28 25 38 40
Carbondale 39.40 -107.21 6170 50 57 90 90 49 50
Colorado
Springs 38.83 -104.82 6010 46 43 70 20 45 45
(downtown)
Colorado
Springs
38.81 -104.68 6150 48 NA 20 20 46 NA
Municipal
Airport

Colorado Design Snow Loads 33

 1971
1971 Map Tabulated
Lat. Lon. Alt. (Converted (Converted 2007 2007 2015 2015
City/Town (deg.) (deg.) (ft.) to pg) to pg) Map Tabulated Map Tabulated
Cripple
38.75 -105.18 9490 69 57 35 35 72 70
Creek
Denver 39.74 -104.98 5280 44 43 20 30 38 35
Estes Park 40.43 -105.52 7520 52 57 50 45 59 65
Fort Collins 40.59 -105.08 5000 37 43 20 20 35 35
Glenwood
39.55 -107.32 5760 55 57 30 30 43 40
Springs
Lamar 38.09 -102.62 3620 29 29 25 30 30 30
Rangely 40.09 -108.80 5230 30 29 20 20 32 35
Silverton 37.81 -107.66 9310 125 129 150 140 112 105
Snowmass
39.21 -106.94 8210 110 NA 165 165 87 90
Village
Steamboat
40.48 -106.83 6730 107 107 100 100 75 85
Springs
Vail 39.64 -106.37 8190 109 107 170 145 84 90
Vail Mid-
39.61 -106.37 10300 197 NA 210 210 170 175
Mountain
*Where the tabulated load for the 2007 edition differs from the map load, the tabulated load is listed second.

2.5. Conclusions
For the first time, a map of reliability-consistent design ground snow loads has been developed
for the state of Colorado. The development of this map was in response to a recognition that the
current LRFD design load factor for snow loads (i.e. 1.6), when applied to a ground snow load
that has a 50-year mean recurrence interval, does not lead to consistent safety against snow-
induced collapse across the state of Colorado. The committee believes that inconsistent
reliability is not a Colorado-specific issue, and this study is an example of one solution to that
problem.

The new reliability-targeted design ground snow loads target a safety index of β=3.0, which is
the stated safety objective of the ASCE 7 for relatively benign structural failures in ordinary Risk
Category buildings (Risk Category II). The new loads are on the order of 50 to 100 percent larger
than the 50-year loads in the plains and approximately 10 percent lower than the 50-year loads at
high elevation locations. Although the increases of design ground snow loads for the plains seem
large, they result in roof snow loads that are consistent with current and historic design practice
in that portion of the state, which is to design roofs for 20 to 30 psf of snow load.

2.6. Works Cited

ASCE (2010). Minimum Design Loads for Buildings and Other Structures, ASCE Standard ASCE/SEI 7-
10, American Society of Civil Engineers, Reston, VA.

Bartlett, F.M, Dexter, R.J., Graeser, M.D., Jelinek, J.J., Schmidt, B.J. & Galambos, T.V. (2003). “Updating
Standard Shape Material Properties Database for Design and Reliability” Eng. J., First Qrt.

DePaolo, M. (2013). A Proposal for a Unified Process to Improve Probabilistic Ground Snow Loads in the

Colorado Design Snow Loads 34

 United States using SNODAS Modeled Weather Station Data. M.S. Thesis, University of Colorado
Boulder.

Ellingwood, B., MacGregor, J.G., Galambos, T.V., & Cornell, C.A. (1982). “Probability based load criteria:
load factors and load combinations.” J. Struct. Div., 108(5), 978-997.

Fishman, GS (2006). A First Course in Monte Carlo. Duxbury: Belmont, CA.

Galambos, T.V. & Ravindra, M.K. (1978). “Properties of Steel for use in LRFD.” J. Struct. Div., 104(9),
1459-1468.

Golden Software (2015). Surfer 12. Golden Software, Inc. Golden, CO.

Kennedy, D.J.L., & Aly, M.G. (1980). Limit states design of steel structures-performance
factors. Canadian J. Civil Eng, 7(1), 45-77.

Lind, N.C. (1977). “Rationalizations of Sections Properties Tables.” J. Struct. Div., Vol. 103, No. 3, March
1977, pp. 649-662.

NRCS (2015). Natural Resources Conservation Center: National Water and Climate Center.
http://www.wcc.nrcs.usda.gov/snow/. (Last accessed March 17, 2015).

Luco, N., Ellingwood, B. R., Hamburger, R. O., Hooper, J. D., Kimball, J. K., and Kircher, C. (2007).
“Risk-targeted versus current seismic design maps for the conterminous United States.” SEAOC 2007
Convention Proc., Structural Engineers Association of California, Sacramento, CA

NWS (2015). National Weather Service. http://www.weather.gov. (Last accessed March, 2015)

Schmidt, B. J., & Bartlett, F. M. (2002). “Review of resistance factor for steel: resistance distributions and
resistance factor calibration.” Canadian J. Civil Eng., 29(1), 109-118.

SEAC (1971). Snow Load Design Data for Colorado. Prepared by SEAC.

SEAC, (2007). Colorado Ground Snow Loads. Prepared by the SEAC Snow Load Committee.

WRCC (2015). Western Regional Climate Center. http://www.wrcc.dri.edu/. (Last Accessed March, 2015)

Colorado Design Snow Loads 35

Appendix 1. Development of Conversions between Snow Depth and
Snow Weight
A1.1. Introduction
Direct measurements of snow weight, reported in the form of snow water equivalent, are used for
determining annual maximum snow weights whenever possible. However, the majority of NWS
stations do not report snow weight measurements, so weight must be estimated from depth. This
appendix describes the process for obtaining the relationships between annual maximum snow
depth and annual maximum snow weight that are used to estimate annual maximum ground
snow loads at NWS stations for the state of Colorado.

A1.2. Compacted Snow and Settled Snow

For the purpose of developing depth-to-weight relationships for ground snow in Colorado, snow
is classified as “compacted” or “settled” (SEAC 2007). Compacted snow generally occurs at
mountain sites and is a result of multi-storm accumulation (i.e. the snow does not melt off
completely between snow storms). Compacted snow tends to be relatively dense, due to multiple
layers of snow from a series of snowstorms condensing the snow layers beneath them over time.
Sites with altitude greater than 8500 ft. are considered to be compacted snow sites in this study.
The 8500 ft. criterion for compacted snow is based on the judgment of the SEAC Snow Load
Committee and is grounded in the observation that most Snow Course stations (i.e. stations
intended to gauge winter snow pack for estimating spring run-off and therefore located where
season-long accumulation is expected to occur) are located at altitudes of 8500 ft. and higher.

Settled snow sites are at lower altitudes and are characterized by ground snow that tends to melt
off between snowstorms. The annual maximum snow loads at settled snow sites are often the
result of only one storm and snow at these sites is generally less dense because there is less time
for it to consolidate. This study considers settled snow sites in the state of Colorado to be those
located east of the Rocky Mountains with altitude less than 6500 ft. and those located in or west
of the Rocky Mountains with altitude less than 5500 ft. The criteria for settled snow sites are
based on the judgment and experience of the SEAC Snow Load Committee. The upper limit
altitude for settled snow sites is lower in the Rocky Mountains and western portion of Colorado
than in the eastern part of the state, because snow in mountain valleys tends to persist longer
without melting than snow in the eastern portion of the state at similar altitudes.

Separate depth-to-weight relationships are employed for compacted and settled snow sites,
because compacted snow is generally denser than settled snow. There is also a large portion of
the state of Colorado that does not meet the criteria for compacted or settled snow; snow sites in
these areas are termed “intermediate” snow sites. This study considers intermediate snow sites as
a hybrid combination of settled and compacted snow. Conversions of snow depth to snow weight
for the three snow types (compacted, settled, and intermediate) are presented in the sections that
follow.

A1.3. Compacted Snow Sites

To examine the relationship between snow depth and weight at compacted snow sites, annual
maximum snow depth and annual maximum snow weight data from all Snow Course sites are
plotted on Figure A1.1; only years for which depth and weight data are both available are used.
Colorado Design Snow Loads 36
Figure A1.1 Annual maximum snow weight vs. annual maximum snow depth for compacted snow sites in the
state of Colorado.

Two possible depth-to-weight relationships are shown on Figure A1.1: (1) a power curve
developed by Tobiason and Greatorex (1997), and (2) a power curve fitted to the Colorado
specific data. The Tobiasson and Greatorex (1997) relationship and the power curve fitted to the
Colorado snow course data are described by Equations A1.1 and A1.2, respectively.

J = 0.279 ∗ L;. M A1.1

J = 0.584 ∗ L;.O/ A1.2

where J is snow weight or load in psf and L is snow depth in inches.

Examination of Figure A1.1 shows that the power curve that is fitted specifically to the Colorado
compacted snow data predicts snow weight better than the Tobiasson and Greatorex relationship,
which underestimates density at compacted snow sites. This is reasonable because their
relationship was not developed for high altitude sites where compacted snow conditions are
expected. Therefore, annual maximum depth data are converted to annual maximum weights by
Equation A1.2 at compacted snow sites where only snow depth data are available.

A1.4. Settled Snow Sites

An analysis of settled snow sites is shown in Figure A1.2. There are fewer stations at settled
snow sites that report both depth and weight data, so the data set for fitting a Colorado-specific
depth-to-weight relationship for settled snow is small in comparison to the data set available for
fitting a compacted snow relationship (125 data points vs. 6,524 data points, respectively).

Colorado Design Snow Loads 37

Visual inspection of Figure A1.2 reveals that a power curve relationship fitted to Colorado data
is very similar to the Tobiasson and Greatorex (1997) relationship up to a snow depth of about
12 inches. Beyond 12 inches, there is very little data available to calibrate a Colorado-specific
relationship. We note that the Tobiasson and Greatorex (1997) equation was fitted to data from
first order NWS stations across the United States, many of which are be expected to be settled
snow sites. Therefore, since the Tobiasson and Greatorex relationship fits the Colorado settled
snow data well at depths below 12 inches (where significant data are available), it is not
unreasonable to assume that it also predicts the depth-to-weight relationship at depths greater
than 12 inches as well12. Therefore, the Tobiasson and Greatorex (1997) equation (Equation
A1.1) is selected for use at settled snow sites.

Figure A1.2 Annual maximum snow weight vs. annual maximum snow depth for settled snow sites in the
state of Colorado.

A1.5. Intermediate Snow Sites

Intermediate snow sites are those that are neither classified as compacted or settled. There is
insufficient data from stations at altitudes between 6500 ft. and 8500 ft. in the state of Colorado
that report both depth and weight to fit a model to intermediate site data directly.

As an alternative, the depth-to-weight conversion at intermediate snow sites is computed by

linearly interpolating (based on altitude) between the compacted and settled depth-to-weight
conversions. Given the absence of data in the state of Colorado to either reject or verify an

12
A fundamental difference, however, is that the Tobiasson and Greatorex relationship was fitted to 50-year maxima
of snow depth and weight, rather than annual maxima. This difference does not appear to bias the predictions for
Colorado settled snow data.

Colorado Design Snow Loads 38

approach for determining snow weights at intermediate snow sites, the SEAC Snow Load
Committee considers the interpolation approach to be a rational solution for those locations.

A1.6. Treatment of Uncertainty/Variability in the Depth-to-Weight Conversions

The depth-to-weight conversions are uncertain, as evidenced by the scatter of the recorded data
points about the theoretical relationships in Figure A1.1 and Figure A1.2. Therefore, statistical
analyses were performed to determine if and how to deal with the additional uncertainty that
comes from having depth rather than weight data at some sites. These analyses led to a
conclusion that unbiased estimations of the annual maximum ground snow load distribution
shapes, and ultimately the safety index beta, are possible when annual maximum snow depths are
converted directly to weight without consideration of the additional variability of the depth-to-
weight conversions.

The analyses conducted to investigate the variability in the depth-to-weight conversions are
presented in Figure A1.3. Figure A1.3 compares statistics of annual maximum weight data that
are computed from two kinds of weight data: (1) weights computed indirectly from depth
measurements by depth-to-weight conversions (i.e. “converted” weight data) and (2) weight data
that are measured. For this comparison, only the 133 snow sites having both depth and weight
recordings are used. For each site, the ratio of four statistics that describe the annual maximum
weights are compared: mean, coefficient of variation, 50-year load (weight), and the design snow
load to produce a safety index of 3. On average, the ratio of the statistic computed from
converted weight data to that computed from recorded weight data is approximately 1.0,
although it varies from site to site as shown in Figure A1.3. First, we observe that our estimates
of those statistics are unbiased on average. Since our goal is to produce our best estimates of the
design loads that will result in a safety index of 3, annual maximum depth measurements are
converted directly to annual maximum weights, without incorporating the uncertainty of the
conversions, since this is shown to be an unbiased approach. However, our confidence in the
annual maximum weight statistics that are determined from converted weight data is lower than
if actual snow weights had been measured. This additional source of uncertainty is not
considered in the analysis.

Colorado Design Snow Loads 39

Figure A1.3 Boxplots, showing the 5th, 25th, 50th, 75th, and 95th percentiles of the ratio of statistics computed
from converted weights to statistics computed from measured weights, for 133 sites.

A1.7. Works Cited

Tobiasson, W., & Greatorex, A. (1997). “Database and methodology for conducting site specific snow load
case studies for the United States.” 3rd International Conference on Snow Engineering (pp. 249-
256).

SEAC (2007). Colorado Ground Snow Loads. Prepared by the SEAC Snow Load Committee.

Colorado Design Snow Loads 40

Appendix 2. Development of Ground to Roof Weight Conversion for
Monte Carlo Simulation
A2.1. Introduction
This appendix describes the development of a probabilistic model to convert ground snow
weights to roof snow weights (hereafter referred to as “GR”). This is not the ground to roof
conversion for design (i.e. 0.7 in ASCE 7-10); it is a ground to roof conversion for computing the
reliability of roof structures subjected to given ground snow loads. Therefore, the objective is to
develop a mathematical model that computes the expected roof snow load and its variability,
given a ground snow load, as accurately and precisely as possible. The resulting roof snow load
represents the estimate of the snow load that is on the roof at the time of the annual maximum
ground snow load.

Probabilistic models for determining roof snow loads based on ground snow loads have been
developed in the past, notably Ellingwood and O’Rourke (1985). Based on the data available at
the time, they determined that the ratio of roof snow load to ground snow load (i.e. GR) is
approximately lognormally distributed with a median of 0.47 and logarithmic standard deviation
of 0.42. Additional data has been collected since Ellingwood and O’Rourke published their
results in 1985, so this study uses a new data set to make a more refined GR model.

A2.2. Roof and Ground Snow Loads Data Set

The data set for constructing the GR model (see Figure A2.1) was collected by Høibø (1988,
1989) and reported by Thiis and O’Rourke (2015). The data set consists of approximately 870
simultaneous measurements of balanced roof snow load and ground snow load. The
measurements were taken over a span of 20 years (1966 to 1986), primarily for buildings with
low-slope (<45°), unheated, rough (not slippery) roofs in southern Norway. This constitutes a
larger data set than was previously available, including a wide variation in ground snow loads at
which the measurements were made. The data is shown in Figure A2.1. For completeness, the
we note that the Thiis and O’Rourke (2015) data represents the ratio of roof snow load to ground
snow load occurring at an instant in time, rather than the ratio of annual maximum roof snow
load to annual maximum ground snow load, which may not occur simultaneously. Data for
comparing the annual maxima is not currently available, so the available data are used with the
expectation that differences between the two ratios are relatively small.

As shown in Figure A2.1, there is significant variability in GR, but there is also an apparent
downward trend such that the GR is smaller for higher roof loads. This occurs because large
ground snow loads are often the result of snow accumulating over time, and roof snow weight
tends to decrease more quickly with time than does ground snow weight, due to factors such as
wind removal, melting, and sublimation.

Colorado Design Snow Loads 41

Figure A2.1 Ratios of measured roof load to measured ground snow load. Data collected by Høibø (1988,
1989) and obtained from Thiis and O’Rourke (2015).

A2.3. Fitting the Probabilistic Model

For this report, we fit a new model for GR to the data presented in Figure A2.1. Following
Ellingwood and O’Rourke (1985), the GR is assumed to be lognormally distributed. However,
the proposed model also accounts for the downward trend in GR as ground snow weight
increases (see Figure A2.1).

For use in Monte Carlo simulation analysis, it is ideal that the GR model have an analytical form
to aid fast simulation of large sets of random variables. However, the nature of the data does not
lend itself well to fitting an analytical function directly to it; specifically, it is not homoscedastic
and it is not linearized through a simple transformation (e.g. a log transformation). Consequently,
the GR model is fit to the data in two steps: a nonparametric model is fit to the data in step one,
and that model is used to inform an analytical model that is developed in step 2.

A2.3.1. Step 1: Non-parametric Characterization of the Ground-to-Roof Snow

Weight Conversion
The median and logarithmic standard deviation of GR are quantified at 25 different ground snow
weights by local polynomial regression (Fan and Gijbels 1996). The results of the local
polynomial regression are shown in Figure A2.2. As the ground snow weight increases, GR
decreases.

The local polynomial regression is a moving weighted regression of the natural log of GR vs.
ground snow weight. For each of the 25 selected ground snow weights, a weighted linear
regression is performed using the closest 20 percent of the data points (distances between data
points are measured by the difference in ground snow weight) to determine the median and

Colorado Design Snow Loads 42

logarithmic standard deviation of GR at each ground snow weight of interest. To investigate the
sensitivity of the number of data points used in the local polynomial regression, we repeated the
local polynomial regression using the closest 50 percent of the data points rather than 20 percent.
There are only very minor changes in the fitted curve, increasing the estimate of logarithmic
standard deviation by about 0.01 and producing a smoothing effect on GR at snow weights less
than 40 psf. These calculations are used to determine the moving median and log standard
deviation in Figure A2.2.

A2.3.2. Step 2: Fitting the Analytical Ground-to-Roof Snow Weight Conversion

Model
An analytical model of GR is fitted to match the results of the local polynomial regression
analysis. The GR model relates the ground snow load to the median GR and its logarithmic
standard deviation, as shown by Equations A2.1 and A2.2.

<=> ? = 0.50 ∗ exp@−0.034 ∗ B + 0.4 A2.1

CDE = min 4.007 ∗ + 0.1 , 0.336 A2.2

where - is the ground snow weight in psf and CDE is the logarithmic standard deviation of GR.
The analytical GR model, plus and minus one logarithmic standard deviation, is shown with solid
and dashed black lines in Figure A2.2, and the model for the logarithmic standard deviation of
GR is also shown in Figure A2.3.

Figure A2.2 Local polynomial (i.e. moving median) and analytical approximations of GR, with the ratio of
roof snow weight to ground snow weight for the recorded data underlain. Raw data was collected by Høibø
(1988, 1989) and obtained from Thiis and O’Rourke (2015).

Colorado Design Snow Loads 43

Figure A2.3 Logarithmic standard deviation of GR computed by local polynomial regression and by the
analytical model.

A2.4. Discussion and Limitations

The analytical model of GR closely agrees, on average, with the moving median GR (which is
estimated with local polynomial regression) for ground snow weights greater than 25 psf.
Ground snow weights greater than 25 psf are the primary focus for fitting the GR model, because
the proposed minimum design ground snow load in Colorado is 25 psf. Since the corresponding
LRFD design point for a 25 psf ground snow load is 1.6 x 25 = 40 psf, failures are not are
expected to occur when the weight of snow on the ground is less than 25 psf13. Deviations of the
moving median from the analytical function at ground snow weights greater than 25 psf are due
to local fluctuations of the data that the committee believes are random and not systematic; the
proposed analytical function smooths those fluctuations. There are, however, three limitations of
the GR model that may impact the reliability analysis results.

The first limitation of the GR model is that there is insufficient data beyond ground snow weights
of approximately 80 psf. Due to a lack of knowledge about GR beyond 80 psf, the function that
is selected to model GR has an asymptote of 0.4, which is nearly reached at a ground snow load
of 80 psf. The SEAC Snow Load Committee hypothesizes that the median GR should decrease
below 0.4 at ground snow weights greater than those in the data set. However, without GR data
to confirm the hypothesis, we are compelled to constrain the GR model to a median value of 0.4

13
A sensitivity study showed that ground snow loads less than 25 psf would account for approximately 0%-1% of
roof failures in buildings designed for ground snow loads of 25 psf, assuming there are no design or construction
defects.

Colorado Design Snow Loads 44

or above. This assumption is conservative, since it leads to higher GRs than are probably
realistic for large ground snow loads.

The second limitation is that the data used to fit the GR model was obtained from unheated roofs
with slopes up to 45° having a variety of exposures (Thiis and O’Rourke 2015), whereas the
reliability analyses in this study that utilize the GR model consider only flat roofs with normal
thermal conditions and exposure. Since the surfaces of the roofs represented by the data set were
primarily rough (non-slippery), it is unlikely that roof slope significantly affects the GR data.
However, it is highly likely that the recorded GRs are biased high (i.e. conservative), since the
data is mostly coming from unheated roofs, where snow melting occurs more slowly. In addition,
we note that the variety of exposure types and thermal conditions increases the scatter of the
data. As a result, the logarithmic standard deviation of the GR model is most certainly
overestimated, which adds to the conservatism of the model. Corrections could be made to the
GR model to account for roof thermal conditions and exposure, but the necessary information to
do so is not currently available.

The third limitation is that, if the GR model is used to simulate random values of GR in a Monte
Carlo reliability analysis, it is possible to get GR greater than 1.0, particularly at lower ground
snow loads. Physically, GR > 1 in the reliability assessment implies that the annual maximum
roof snow weight is greater than the annual maximum ground snow weight. The committee
acknowledges that there are circumstances under which the roof snow weight may be larger than
the ground snow weight (such as a roof that is colder than the ground and accumulates snow
more quickly), but it is unlikely that the difference between them is large. Therefore, for
simulation purposes, GR values are capped at 1.25; this upper bound is based on the judgment
and experience of the SEAC Snow Load Committee, and the committee expects that it is
conservative.

A2.5. Conclusions
An analytical model to predict the median and logarithmic standard deviation of the ground to
roof snow weight conversion factor (GR) is fit to a Norwegian data set that consists of
simultaneous measurements of roof snow weight and ground snow weight. The use of this new
data set allows us to quantify GR as a function of the ground snow load. This change reduces the
median GR at large loads (>80 psf) to approximately 0.4, rather than the 0.47 used in older
models. In addition, this change reduces the error term (i.e. the logarithmic standard deviation)
from 0.42 to 0.33. Nonetheless, the predictions of median GR and its logarithmic standard
deviation are conservatively high due to limitations of the model and of the data set to which it is
fit, particularly at large snow loads. Without additional data, there is not a statistically defensible
way to either quantify or eliminate these biases. Despite this conservatism, the use of the new
GR model reduces the estimates of roof loads compared to the previously used Ellingwood
model, thus improving the reliability assessed at all Colorado sites.

A2.6. Works Cited

Ellingwood, B., & O'Rourke, M. (1985). Probabilistic models of snow loads on structures. Structural
safety, 2(4), 291-299.

Fan, J., & Gijbels, I. (1996). Local polynomial modeling and its applications: Monographs on statistics and
applied probability 66 (Vol. 66). CRC Press.

Colorado Design Snow Loads 45

Thiis, T., & O’Rourke, M. (2015). “A model for snow loading on gable roofs.” Journal of Structural
Engineering, In Press.

Colorado Design Snow Loads 46

Appendix 3. Determining Probability Distributions for Snow Sites
A3.1. Introduction
Probability density functions are used to represent the distributions of annual maximum ground
snow loads at snow sites. Previous studies (e.g. Ellingwood and Redfield 1983, SEAC 2007)
have used probability distributions to determine characteristic loads for a given return period
interval, e.g. a snow weight with a mean recurrence of 50 years. In this study, the role of the
annual maximum snow weight probability distributions is distinctly different than that of
previous studies in that the distributions are intended to be used for reliability analyses, rather
than to predict a single characteristic snow load. Therefore, the distributions of annual maximum
snow loads in this study should be representative of the true annual maximum snow load
distribution over the entire region of the distribution that impacts the structural reliability
analysis, not just the 50-year load.

Probability density functions that represent annual maximum ground snow loads at a given site
are computed in two stages. First, a distribution is fit specifically to the annual maximum snow
weight data at each individual site (“site-specific” distribution fitting). Next, the upper tail of the
distribution is adjusted to account for information from snow data at similar sites. This appendix
describes the site-specific tail-fitting approach that is used, as well as the method for adjusting
the distribution upper tails to account for observations at similar snow sites.

A3.2. Site-Specific Distribution Fitting

A3.2.1. Background
Distributions of annual maximum ground snow loads in the United States have been
predominantly modeled by the Lognormal distribution, which was recommended by Ellingwood
and Redfield (1983), who studied annual maximum snow loads in the Northeast quadrant of the
United States. However, when fitted to snow data in many locations on the plains and lower
foothills of Colorado, the lognormal distribution fitted to the entire historical record is
insufficient for modeling rare loads, because its tail dies off too quickly. As an example, the 50-
year load at Denver-Stapleton using 121 years of data from 1893 to 2013 is 20.4 psf when a
Type II Extreme Value (Hereafter “Type II”) distribution is fitted to the data, versus 17.1 psf
when a Lognormal distribution is fitted to the data. In this case, the Type II is a better fit based
on a number of goodness-of-fit statistical tests. For this reason, the SEAC Snow Load
Committee, in developing its 2007 snow load recommendations for Colorado (SEAC 2007),
explored a variety of distributions for modeling annual maximum ground snow loads: Normal, 3-
parameter Lognormal, 3-parameter Gamma, 3-parameter Loggamma, Frechet Type II Extreme
Value, and Gumbel Type 1 Extreme Value.

Colorado Design Snow Loads 47

 A3.2.2. Tail-Fitting Approach
This study takes a different approach to modeling probability distributions of annual maximum
ground snow loads. Rather than consider a variety of possible distributions, the lognormal
distribution is used, but it is fitted only to the upper tail of the data, as in Ellingwood (1981). This
approach, hereafter referred to as “tail-fitting,” ensures that the site-specific distribution is fitted
to the portion of the data set that is important for determining rare loads.

Tail fitting is performed by the following steps:

1. Rank order the annual maximum ground snow data.

2. Plot the rank-ordered data on probability paper.
3. Fit a least-squares linear regression line to the top 33 percent of the data.

An example of the tail-fitting technique applied to the same Denver-Stapleton data is shown in
Figure A3.1. Figure A3.1 shows that the tail-fit distribution captures the tail behavior well,
whereas the upper tail of the recorded data is thicker than what is predicted by a lognormal
distribution fitted to the entire data set. For comparison, the 50-year load according to the tail-fit
Lognormal distribution is 20.5 psf, versus 17.1 psf for the Lognormal fitted to all of the data and
20.4 psf for the Type II distribution (which is the best fit distribution if the data are fit to the
entire data set). We note also that the tail-fit distribution often does not match the observed data
for low snow loads, but low snow loads are not important for the reliability assessment, because
they do not cause failures.

Figure A3.1 Probability plot of a tail-fit lognormal distribution for the Denver-Stapleton data alongside a
lognormal distribution fitted to the entire data set. The thickened part of the solid line represents the range
over which the tail-fitted distribution was fitted (i.e. upper 33 percent of the data points).

Colorado Design Snow Loads 48

 A3.2.3. Limitations of the Site-Specific Tail-Fitting Approach
For most sites in Colorado, the lognormal tail-fitting technique works well for predicting extreme
loads with recurrence intervals up to about 50 years. However, extrapolating the distribution to
predict rare loads whose recurrence intervals are much greater than the length of the data time-
series (i.e. greater than approximately 50 years for most snow sites) should be done cautiously,
because the extrapolations will only be correct if the true distribution of the data is Lognormal.
As an illustration, four tail-fits are performed on the Denver-Stapleton data using the Normal,
Lognormal, Gamma, and Loggamma distributions, in addition to a Type II distribution fit to the
entire data set. These distributions are overlaid on a histogram of the original data in Figure
A3.2, as well as Lognormal probability plots in Figure A3.3. The Loggamma and Type II
distributions are the best-fitting distributions according to several goodness-of-fit statistical tests.
However, with the exception of the Normal distribution (which bears no resemblance to the
Denver-Stapleton data whatsoever), each of the distributions are similar over the portion of the
data set to which they are fitted (i.e. the top 33 percent). The 50, 100, 500, and 1000 year loads
for the four tail-fitted distributions are reported in Table A3.1. Table A3.1 shows that the tail-fit
distributions agree fairly well for time-intervals within the span of the data set (121 years in the
case of Denver-Stapleton), but deviate substantially beyond it; this trend can be observed
visually in Figure A3.3.

Figure A3.2 Histogram of the Denver-Stapleton annual maximum snow loads with four different tail-fit
distributions overlaid, in addition to a Type II distribution fitted to the entire data set.

Colorado Design Snow Loads 49

Figure A3.3 Denver-Stapleton annual maximum snow weights plotted on (a) a Lognormal probability plot
and (b) a linearized lognormal probability plot. The results of tail-fit distributions, in addition to a Type II
distribution that is fitted to the entire data set, are overlaid.

Colorado Design Snow Loads 50

 Table A3.1 Ground snow loads predicted by four tail-fitted distributions and the Type II
distribution for Denver-Stapleton
Mean Tail-fitted Distribution
Recurrence Type II
Normal Gamma Lognormal Loggamma
Interval
50 yrs 21.5 psf 21.6 psf 20.5 psf 20.8 psf 20.4 psf
100 yrs 24.6 psf 26.0 psf 25.4 psf 26.8 psf 26.5 psf
500 yrs 30.9 psf 36.3 psf 39.4 psf 46.9 psf 47.6 psf
1000 yrs 33.3 psf 40.8 psf 46.6 psf 59.0 psf 60.8 psf

Therefore, it is concluded that Lognormal tail-fit distributions are satisfactory for predicting rare
loads with up to approximately 50-year recurrence intervals, but they may be unsatisfactory for
predicting extremely rare loads (e.g. 500 or 1000 year loads) unless the data are actually
lognormally distributed in the far tail. In this study, Lognormal tail-fit distributions are used to
predict 20-year loads at each site, which is needed in the later analysis.

If the true underlying distribution of the data were known, then a simple solution for predicting
extreme loads would be to use the true distribution, rather than a lognormal distribution.
However, determining the true underlying distribution of annual maximum snow loads at a site is
not as simple as selecting the distribution that fits the data best. There are two reasons for this
assertion. The first is that randomness of the data can result in the selection of a best-fit
distribution that is not the true underlying distribution. For example, a sensitivity study showed
that random 50-year data sets generated from a Lognormal distribution with logarithmic standard
deviation of 0.6 are identified as being Type II distributions (i.e. best fit by a Type II) more often
than they are identified as Lognormal. This misidentification occurs partly because the Type II
distribution has three parameters and is therefore more adaptable to intricacies of a particular
dataset, even when those intricacies are random and not actually real properties of the true
underlying distribution. In this example, fitting a Type II distribution to data whose true parent
distribution is not Type II is relatively inconsequential for predicting loads with mean recurrence
intervals near or within the span of the data, but it results in large disparities at extremely large
loads, similar to the disparities shown among the various distributions in Figure A3.3 and Table
A3.1.

The second reason that determining snow load distributions based only site-specific analyses is
not practical is because data records at individual sites are not long enough to verify whether a
best-fit distribution for the 30 to100 years of recorded data is still applicable for loads as rare as
500-1000 years. It may be that the extreme upper tails are described by distributions that are
different than those that describe the bulk of the data below the extreme upper tail, or it may be
that they do not follow any particular probability density function at all. If each snow site is
analyzed in isolation, these problems are difficult to overcome. However, further insight about
the true shapes of the distribution tails over the range that is critical for the reliability assessment
is achieved by accounting for observations at several similar snow sites collectively, as described
in the following sections.

Colorado Design Snow Loads 51

 A3.3. Distribution Tail Shapes Informed by Multiple Snow Sites
Additional information about the upper tail shapes is obtained by combining data from similar
snow sites. The essence of the procedure is to determine the magnitude of the snow loads in the
distribution tail with a site-specific analysis, but determine the shape of the tail by considering
the combined data from a number of stations whose distribution tails are expected to have similar
shapes. Grouping, or “clustering” of snow sites in Colorado with similar properties to obtain
information about their distribution shapes is suggested by DePaolo (2013).

To explain, we start with a simple case, wherein several snow sites are assumed to have the same
underlying distribution, in terms of the mean as well as the shape of the upper tail of the
distribution. This assumption may be valid for the eastern plains of Colorado, for example,
where there are not big variations in climate from site to site. For this case, the procedure is as
simple as combining the data from multiple stations to create one large data set. A collection of
20 snow sites having 50 years of annual maximum snow data per site, for example, could be
combined to make a single station with 20x50=1000 data points. In this example, a distribution
fit to the enlarged data set can allow a better prediction of rare loads having up to a 1000-year
mean recurrence interval without extrapolation. If all of the contributing snow sites are expected
to have the same distribution shape and magnitude, then the enlarged data set provides a better
estimation of that distribution.

Combining data from multiple sites to obtain and enlarged data record assumes that the snow
data from the different sites are statistically independent. However, sites in a common region are
often affected by the same storm and, as a result, the annual maxima at different sites are likely
to be correlated. Some of this correlation is eliminated by combining nearby snow sites into
single stations before fitting the distributions. However the correlation is not entirely eliminated,
so its potential impact on the resulting fitted distributions was determined by a sensitivity study.
To test the impact of correlation between annual maximum snow loads, random sets of historical
snow data were sampled from a known distribution, with and without correlation between them.
The correlated data sets had correlation coefficients between them of 0.6, which is based on the
correlation coefficient between the Denver and Boulder historical snow records. The distribution
fitting methods that are explained in this appendix were applied to the simulated data. The results
showed that correlations between snow sites reduces the confidence in the fitted distribution
parameters, because the impacts of correlations among data are similar to the impact of reducing
the size of the data set. However, the presence of correlations did not bias the estimates of the
distribution parameters significantly (less than 5% bias was introduced by the correlation).

For the case where the magnitudes of the annual maximum snow weight distributions are
expected to vary among snow sites, the process of combining the data to make a single enlarged
data set is described by the following steps, which are illustrated in Figure A3.4 - Figure A3.7:

1. Select a group of snow sites whose distribution shapes are expected to be similar to that
of the site of interest (see Section A3.3.2).
2. For each site, determine the 20-year load using a site-specific tail-fit lognormal
distribution (refer to Section A3.2.2).
3. Scale each site’s historical record so that the 20-year load is equal to the average 20-year
load for all of the sites in the group.
4. Combine all of the scaled data into an enlarged data set.

Colorado Design Snow Loads 52

 5. Fit a lognormal distribution to the top 10 percent of the enlarged data set, using the tail-
fitting method described in Section A3.2.2. The smaller (10 percent) fraction of the tail,
as compared to the 33 percent used in the site-specific approach, is permissible because
there are far more data points in the enlarged data set.
6. Apply a correction to the logarithmic standard deviation of the fitted distribution to
account for systematic biases that are introduced by the scaling process (see Section
A3.3.1).
7. For the site of interest, re-scale the distribution from Steps 5 and 6 to the 20-year snow
load at that site from Step 2.

The distributions resulting from this procedure match the 20-year load at each site, but have
upper tails whose shapes are informed by a number of sites whose upper tail behaviors are
expected to be similar. Therefore, they maintain a site-specific scale, or magnitude, but a
regionally-informed tail shape. This process is referred to as “clustered station distribution
fitting” for the remainder of this appendix, because clusters of sites are used for determining
distribution tail shapes.

This procedure is different than fitting a distribution to the data at one site without considering
adjacent and similar sites in another significant way. Considering one site at a time, it is not
unusual to discard the largest recorded value as an outlier based on statistical tests. Combined
normalized data from several similar sites results in such data points not being discarded,
because they represent the truly rare event for the group of sites.

Figure A3.4 Illustration of Steps 1-2 of the cluster station distribution fitting procedure, for four hypothetical
sites (noting that a minimum of 20 sites are used for the actual procedure). At this stage, tail-fit distributions
have been developed at each site and their 20-year loads have been determined. Historical records for each
hypothetical site are generated from lognormal distributions having σ=0.6, but different median values.

Colorado Design Snow Loads 53

Figure A3.5 Illustration of Step 3 of the cluster station distribution fitting procedure. Data from each site are
scaled to have a common 20-year load.

Figure A3.6 Illustration of Steps 4-6 of the cluster station distribution fitting procedure. The scaled data are
combined to make a single data set. The plot shows the data in their new rank-ordered plotting positions. A

Colorado Design Snow Loads 54

lognormal distribution is fitted to the top 10 percent of the enlarged data set and then corrected to account
for bias that is systematically introduced.

Figure A3.7 Illustration of Step 7 of the cluster station distribution fitting procedure for the case where the
site of interest is Site 4. The cluster-fit distribution is scaled so that its 20-year load matches the 20-year load
that is determined for Site 4 by a site-specific tail-fit distribution.

A3.3.1. Accounting for Systematic Biases in the Distributions that are Fitted to the
Clustered Data Sets
Sources of Bias in the Distributions that are Fitted to the Clustered Data Sets
In scaling the data from several snow sites to create a clustered data set, a systematic bias is
introduced. The bias is a result of data sets with large rare loads tending to be scaled down,
because their calculated 20-year loads tend to be higher than those of sites that have not
experienced large rare loads; this leads to a systematic reduction of large rare loads in the
clustered data sets. This section explains, by example, the origin of that bias and is followed by
an explanation of how it is corrected.

Let us consider the data shown in Figure A3.8. The data represent two hypothetical sites, each
having the same parent distribution, i.e. following the same lognormal distribution with the same
mean and standard deviations, but with unique sets of random data representing their historical
records. In their 50 years of record, Site 1 has not experienced any significantly rare loads, but
Site 2 has. The largest loads at Sites 1 and 2 are 13.3 psf and 21 psf, respectively, which
correspond to mean recurrence intervals of 29 years and 199 years, according to the theoretical
parent distribution. As a result, the distributions that are fit to the two site’s data predict 20-year
loads of 11.3 psf and 14.0 psf, in comparison to the theoretical 20-year load of 12 psf.

Colorado Design Snow Loads 55

Figure A3.8 Two randomly generated data sets, each coming from the same theoretical parent distribution
(black line, Lognormal with Mu=1.5 and Sigma=0.6). A site-specific tail-fit Lognormal distribution is fit to
each site; the parameters are Mu=1.62, Sigma=0.49 and Mu=1.62, Sigma=0.62 for sites 1 and 2 respectively.
The thickened portion of the lines represent the range over which the tail-fitted distributions are fitted.

The data from Sites 1 and 2 are combined to make a clustered data set, according to the clustered
station distribution fitting procedure described in Section A3.3. Each data set is scaled so that its
20-year load is equal to the average of the two (12.7 psf); the necessary scale factors for Sites 1
and 2 are 1.12 and 0.9, respectively. The top 10 percent of the clustered, scaled data set is shown
with the solid symbols in Figure A3.9. Accompanying it on the same figure is the same subset of
the data, but with no scaling.

Comparing the two distributions in Figure A3.9 shows that scaling the data before combining it
tends to reduce the logarithmic standard deviation (i.e. the shape parameter) of the distribution
that that is fitted to it. That is because rare loads (e.g. the largest load in Site 2 of the example)
tend to raise the 20-year load that is computed at a given site, thus reducing the scale factor
applied to sites with those data, and diminishing their impact on the distribution tail. In this
example, the unscaled combined data set has a logarithmic standard deviation of 0.6, which is
equal to that of the parent distribution. However, the scaled combined data set has logarithmic
standard deviation of 0.52, which is an underestimation of the true logarithmic standard
deviation. For groups of sites in which the shape and scale of the distribution are expected to be
the same, this problem could be circumvented by combining their data without performing any
scaling. However, with the exception of possibly the eastern plains, the groups of snow sites that
are formed are expected to vary in scale, so scaling must occur before combining them to make
an enlarged station.

Colorado Design Snow Loads 56

Figure A3.9 Lognormal probability plot of the top 10 percent of data points for the enlarged example data
set, with and without scaling of the Site 1 and Site 2 data.

Correcting the Systematic Bias Introduced by the Scaling Method

The shapes (i.e. the logarithmic standard deviation, σ) that are obtained by the clustered station
distribution fitting approach are corrected by Equations A3.1 and A3.2.

CPQRR S = 1.16 ∗ CP R T A3.1

CPQRR S = 0.97 ∗ CP R T + 0.03 A3.2

Equation A3.1 is applicable for groups of snow sites where the parent distribution is expected to
be Lognormal. Equation A3.2 is applicable for groups of snow sites where the parent distribution
is expected to be Normal or Gamma.

Selection of either Equation A3.1 or A3.1 requires knowledge of the parent distribution family.
The likely parent distribution for a given group of stations is strongly related to the shape of a
lognormal distribution that is tail-fitted to their combined (scaled) data. Groups of snow sites
with larger logarithmic standard deviation terms are often fit better by more heavy tailed
distributions (i.e. Lognormal and sometimes even Loggamma). Based on sensitivity studies
conducted by the SEAC Snow Load Committee, groups of sites with uncorrected logarithmic
standard deviations greater than 0.6 are assumed to have parent distributions that are Lognormal,
so they are corrected with Equation A3.1. Those with logarithmic standard deviations less than
0.3 are assumed to have parent distributions that are Normal or Gamma, so they are corrected
with Equation A3.2. If the uncorrected logarithmic standard deviation is between 0.3 and 0.6,
then the underlying parent distribution is assumed to be thinner-tailed than lognormal, but
thicker-tailed than Gamma and Normal; in such cases the correction is computed by a weighted

Colorado Design Snow Loads 57

average of Equations A3.1 and A3.2, where the weights are proportional to the uncorrected
logarithmic standard deviation, relative to the threshold values of 0.3 and 0.6.

Development and Validation of the Distribution Shape Corrections

Equations A3.1 and A3.2 are calibrated for the range distributions observed at Colorado snow
sites, and may not be accurate for other regions. A series of sensitivity studies showed that these
corrections are unbiased for the purpose of estimating safety indices in the range of 2.0-3.5 at
sites with parent distributions that are Normal, Gamma, or Lognormal, and having coefficients of
variation ranging from 0.2 to 1.4. Some of the details of the validation and sensitivity studies are
provided here.

The sufficiency of the Equations A3.1 and A3.2 for correcting distribution shapes (i.e.
logarithmic standard deviation, σ) for the clustered station distribution fitting method are tested,
starting with Equation A3.1. Recall that Equation A3.1 is intended to correct the distribution
shape that is determined from a clustered data set when the parent distributions of the individual
stations are assumed to be Lognormal. Therefore, it is tested by applying the correction to data
that is randomly generated from a Lognormal distribution. One-thousand sets of data are
generated from a lognormal distribution, each set containing 50 data points to represent a 50-year
historical record. Then, the clustered station distribution fitting approach is applied to the
simulated data to compute a distribution shape. The shape of the fitted distribution is compared
to the shape of the true underlying (parent) distribution by comparing their logarithmic standard
deviations. The results of several of these tests are shown in Figure A3.10, demonstrating that
Equation A3.1 is effective for correcting bias in the distribution shape that is introduced by the
clustered station distribution fitting method.

Colorado Design Snow Loads 58

Figure A3.10 Logarithmic standard deviation (σ) obtained by scaling and grouping 1000 sets of data and
performing a lognormal tail-fit on the top 10 percent of the combined data set. Each set is composed of 50
data points generated from a lognormal distribution. The process is repeated for theoretical distributions
having σ ranging from 0.2 to 1.05, which corresponds to COV of 0.4 to 1.4. The correction (Equation A1.1)
removes the bias in σ that is introduced by the cluster station distribution fitting method.

Testing Equation A3.2 for correcting the clustered station distribution fitting approach when it is
used to model Gamma or Normal distributed data is less straightforward, because there is no
direct comparison of parameters. Therefore, the correction for Normal and Gamma distributed
data is tested by comparing reliability analysis results from distributions fitted to randomly
generated historical records to reliability analysis results for the theoretical distributions from
which the random historical records were generated. As with the previous tests, 1,000 data sets,
each containing 50 data points, are generated from a theoretical distribution; then, a lognormal
distribution is fitted to the clustered data set by the clustered station distribution fitting method.
Reliability analyses performed with the theoretical distributions and with the fitted distributions
are compared in Figure A3.11 through Figure A3.14. These comparisons show that the
correction equation (Equation A3.2) corrects the shape of the distribution tail well enough to
compute reliability-consistent design ground snow loads (i.e. the design ground snow load that
would result in a reliability index of 3.0) that are nearly indistinguishable from those that are
computed from the true theoretical parent distribution. These reliability results demonstrate that
the corrected distributions approximate well the theoretical distributions over the portion of the
distribution upper tails that is critical for the reliability assessment.

Figure A3.11 Reliability analysis results for a corrected Lognormal distribution that is fitted to data that are
generated from a Normal distribution with mean 100 psf and COV 0.3. Annual maximum ground snow load
distributions of this shape and scale are common at altitudes above 8,000 ft. in Colorado.

Colorado Design Snow Loads 59

Figure A3.12 Reliability analysis results for a corrected Lognormal distribution that is fit to data that are
generated from a Gamma distribution with parameters k=15 and theta=8 (i.e. mean=120 psf and COV=0.26).
Annual maximum ground snow load distributions of this shape and scale are common at altitudes above
8,000 ft. in Colorado.

Figure A3.13 Reliability analysis results for a corrected Lognormal distribution that is fit to data that are
generated from a Gamma distribution with parameters k=6 and theta=7 (i.e. mean=42 psf and COV=0.4).
Annual maximum ground snow load distributions of this shape and scale are common at altitudes between
6,500 ft. and 8,500 ft. in Colorado.

Colorado Design Snow Loads 60

Figure A3.14 Reliability analysis results for a corrected Lognormal distribution that is fit to data that are
generated from a Gamma distribution with parameters k=1 and theta=5 (i.e. mean=5 psf and COV=1.0).
Annual maximum ground snow load distributions of this shape and scale are common at altitudes below
6,500 ft., however the expected parent distributions at such sites are often Lognormal, not Gamma.

A3.3.2. Approach for Clustering Sites with Similar Distribution Shapes

To carry out the proposed distribution fitting method, it is necessary to determine groups or
clusters of snow stations whose upper tails are expected to be similarly shaped prior to applying
the cluster station distribution fitting method. The assembly of snow sites into clusters is
performed in two stages: (1) stations are divided into coarse regions for which different snow
distribution characteristics are expected, based on climatic and macro-topographical features, and
(2) for each site, a minimum of 20 additional sites from the same topographic region with similar
altitudes are clustered with it.

For the purpose of examining annual maximum snow load distribution properties on a regional
basis, the COV is taken as a proxy for the distribution shape. Distributions with heavy upper tails
tend to have larger COV, as compared to distributions with light upper tails. Another proxy for
distribution shape, skewness, is not discussed, but additional studies by the committee show that
trends in skewness and COV are very similar, and that skewness and COV are strongly
correlated for Colorado snow sites (correlation coefficient = 0.8).

Division of Snow Sites into Topographic Regions

Initially, snow stations are divided into six regions, shown in Figure A3.15: (1) the eastern slope
(i.e. Front Range and eastward), (2) the western slope of the Rocky Mountains, (3) the northern
Rocky Mountains, (4) the east-central Rocky Mountains, (5) the southwest Rocky Mountains,

Colorado Design Snow Loads 61

and (6) the San Luis Valley. The initial division into these topographic regions is based on a
combination of expert judgment and k-means cluster analysis. The concept of clustering snow
sites into groups to characterize their distributions of annual maximum snow loading is not new
(e.g. DePaolo 2013).

Figure A3.15 Map of sites by topographic region assignment. The base map has blue, green, and brown
altitude contours at 5,500 t., 8,500 ft., and 10,500 ft., respectively. Eastern slope sites are red triangles,
western slope sites are blue Squares, northern Rockies sites are black circles, east-central Rockies sites are
green diamonds, southwest Rockies sites are green squares, and San Luis Valley sites are green circles. The
latter three, represented by green markers, were eventually combined into one topographic region.

Analysis of the historical annual maximum snow loads in these regions shows that the proxy for
distribution shape, COV, is correlated with altitude and topographic region. For example, Figure
A3.16 shows the trends in COV with altitude for sites within the western slope and eastern slope
topographic regions. COV decreases with altitude for both regions, but the trend is somewhat
different for each of them. There is little distinction in the trends of COV with altitude among
the east central Rocky Mountains, southwest Rocky Mountains, and San Luis Valley sites (i.e. all
of the snow sites plotted in green in Figure A3.15) so these are combined into one topographic
region for analysis. The remaining three topographic regions are considered separately, so a total
of four topographic regions are used.

Colorado Design Snow Loads 62

Figure A3.16 Comparison of COV and altitude for annual maximum snow weight data in two topographic
regions: Eastern Slope (red triangles) and Western Slope (blue squares).

Cluster Analysis
Clusters of snow sites for performing the clustered station distribution fitting method at each site
are formed by selecting all snow sites that are: (a) within the same topographic region as the site
of interest and (b) within 1000 ft. altitude of the site of interest. If a cluster contains less than 20
snow sites, the altitude criterion is relaxed until a minimum of 20 snow sites are included. This
approach implies that each site is grouped with its 20 (or more) most similar sites on the basis of
region and altitude. This cluster is used in determining the shape of the tail for the site of interest,
as described in Section A3.2.2.

A3.4. Discussion and Conclusions

This Appendix describes the method for determining probability distributions for annual
maximum ground snow loads, “cluster station distribution fitting”. The method emphasizes the
portion of the distribution that is critical for reliability analysis, the extreme upper tail. The
resulting distributions are constructed to match the 20-year load at each snow site that is
determined from a site-specific analysis, but have an extreme upper-tail that is informed by a
broader group of snow stations whose distribution shapes are expected to be similar to the site of
interest. Clusters of stations whose distribution shapes are expected to be similar are determined
by a combination of topographic region and altitude.

The method computes 20-year loads at each site from a Lognormal distribution that is fit to the
upper 33 percent of the site data. The shape of the distribution at each site is determined by
combining data from a minimum of 20 stations to make an enlarged data set and fitting a
lognormal distribution to the top 10 percent of its data. This appendix shows that estimating the

Colorado Design Snow Loads 63

upper tail distribution shape this way and applying a calibrated correction factor leads to
unbiased predictions of the reliability index β. Annual maximum snow load distributions at all of
the snow sites are determined by the cluster station distribution method.

A3.5. Works Cited

DePaolo, M. (2013). A Proposal for a Unified Process to Improve Probabilistic Ground Snow Loads in the
United States using SNODAS Modeled Weather Station Data. M.S. Thesis, University of Colorado
Boulder.

Ellingwood, B. (1981). “Wind and snow load statistics for probabilistic design.” Journal of the Structural
Division, 107(7), 1345-1350.

Ellingwood, B., & Redfield, R. (1983). “Ground snow loads for structural design.” Journal of Structural
Engineering, 109(4), 950-964.

SEAC, 2007. Colorado Ground Snow Loads. Prepared by the SEAC Snow Load Committee.

Colorado Design Snow Loads 64

Appendix 4. Recommendations for Serviceability Loads and
Importance Factors
A4.1. Introduction
This Appendix addresses serviceability loads and importance factors for snow load design in the
state of Colorado. The newly proposed design ground snow loads, for the first time, target
consistent reliability against structural failure rather than a consistent hazard, resulting in design
ground snow loads with mean recurrence intervals on the order of 20 years to 300 years.
Serviceability checks for snow loads are generally performed for loads with a 50-year mean
recurrence interval (MRI) hazard.

In addition, the large range of mean recurrence intervals of the design ground snow loads raises
the following question: will snow load importance factors (Is) for Risk Category III and IV lead
to uniform increases in structural reliability across all sites in the state of Colorado?

A4.2. Serviceability Loads

Serviceability checks for snow loads on roofs have traditionally been performed with the 50-year
snow load. Error! Reference source not found. shows that the ratio of 50-year loads to
reliability-targeted loads (referred to as rservice) varies from about 0.5 to 1.3 for Colorado snow
sites and is correlated with altitude. The empirical relationship for rservice in Error! Reference
source not found. is defined by Equation A4.1.

U RV S = 0.55 ≤ .13 ∗ − 0.06 ≤ 1.15 A4.1,

where A is the altitude of a site in thousands of feet.

Given the approximations involved in the statistical analysis and especially the mapping, the
SEAC Snow Load Committee believes that the reliability targeted loads are sufficient for
serviceability checks in most structures. For sensitive structures at high altitudes, the rservice
could be used to increase the reliability-targeted load to a 50-year load. If rservice values less than
1.0 are used at lower altitudes, the SEAC Snow Load Committee recommends that the ground
snow load for serviceability not be taken less than 25 psf.

Colorado Design Snow Loads 65

Figure A4.1. Relationship between the ratios of 50-year ground snow loads to reliability-based loads (rservice)
for each site plotted against altitude for Colorado snow sites.

A4.3. Snow Load Importance Factors

This section shows that uniform importance factors of 1.2 for Risk Category IV buildings and
1.1 for Risk Category III buildings do not achieve the target safety indices for these higher Risk
Category buildings, so an alternative model is proposed. Figure A4.2 shows the required Is to
achieve safety indices of β=3.5 (i.e. the intended reliability for importance category IV
buildings) at the Colorado snow sites. The results indicate that the ASCE 7 Is of 1.2 for Risk
Category IV buildings is non-conservative at low altitudes in the state of Colorado and
conservative at high altitudes. A higher Is is needed at low altitude sites because their annual
maximum snow loads are characterized by heavy-tailed distributions, which results in a larger
increase in load being necessary to increase β from 3.0 to the desired safety level (i.e. 3.5 for risk

Colorado Design Snow Loads 66

Category IV and 3.25 for Risk Category III).

Figure A4.2. Importance factors to achieve a safety index of b=3.5 at Colorado snow sites.

The relationship between altitude and Is for Risk Category IV buildings is approximated by the
empirical equation,

! = 1.15 ≤ 1.66 − 0.056 ∗ ≤ 1.4 A4.2,

where A is the altitude of a site in thousands of feet.

The performance of the empirical equation for computing Is for Risk Category IV buildings
(Equation A4.2) is compared to the performance of Risk Category IV buildings designed for a
constant snow importance factor of 1.2, as in ASCE 7. Calculations of the safety index β at each
snow site for the design load factored by Is are displayed in Figure A4.3 and Figure A4.4. By
visual inspection, a constant Is of 1.2 is conservatively biased for sites with altitudes above 8000
ft. and non-conservatively biased for sites with altitudes below 7500 ft. in the state of Colorado
(Figure A4.3). However, when Is is computed with Equation A4.2, the biases are removed
(Figure A4.4).

Therefore, the SEAC Snow Load Committee recommends that Is be determined by Equation
A4.2 for Risk Category IV buildings in Colorado. For Risk Category III buildings, Is should be
the average of 1.0 and Equation A4.2.

Colorado Design Snow Loads 67

Figure A4.3. Calculations of the safety index β for each snow site, using Is=1.2.

Figure A4.4. Calculations of the safety index β for each snow site, using Is from Equation A4.2.

Colorado Design Snow Loads 68

Appendix 5. Tabulated Data Summary
A5.1. Introduction
Raw station data and assembled snow site data are summarized in this appendix. The summary
statistics in this appendix are for annual maximum ground snow loads. Reliability-based ground
snow loads for each assembled snow site are also reported. We note, however, that the
reliability-targeted ground snow loads are different than those reported in Table 1.1 primarily
because the altitudes of the assembled stations are different that the altitudes of the population
centers reported in Table 1.1 and spatial averaging (smoothing) of the design ground snow loads
is not yet applied. Snow load values in this appendix should not be used for design.

The raw data come from four types of snow recording stations: first-order National Weather
Service (NWS) stations, CO-OP NWS stations, snow courses, and SNOTEL (short for
SNOwpack TELemetry) stations. Descriptions of each station type are available in Section 2.1.1.
In the raw data summary, first order NWS stations are denoted “First Order,” NWS Co-op
stations are denoted “NWS,” snow courses are denoted “SNOW,” and SNOTEL stations are
denoted “SNTL.”

The assembled snow site data are the result of combining snow stations that are either at the
same location or very near each other. For example, many snow courses have been replaced by
SNOTEL stations at the same location; these are combined to make one snow site. The
procedure for combining snow stations to assemble the snow sites is described in Section 2.1.2.
When tabulating the assembled site data, only the station name with the longest data record is
reported (in most cases, the stations from which a snow site is assembled have the same name).
Recorded snow weights are reported as snow water equivalents (SWE), which is the height in
inches of an equivalent (weight) column of water. Each inch of SWE is approximately 5.2 psf.

A5.2. Raw Data Summary

Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
ALAMOSA SAN LUIS
First
VALLEY REGIONAL 2 37.44 -105.86 7534 53 53 18 39 3.1
Order
AIRPORT
COLORADO SPRINGS First
3 38.81 -104.69 6182 65 65 20 16 2.5
MUNICIPAL AIRPORT Order
First
DENVER-STAPLETON 1 39.76 -104.87 5286 121 121 33 43 4.2
Order
First
FORT COLLINS 4 40.62 -105.13 5004 65 65 23 3 3.0
Order
GRAND JUNCTION First
5 39.13 -108.54 4858 51 51 16 34 3.4
WALKER Order
PUEBLO MEMORIAL First
6 38.29 -104.50 4720 41 41 24 31 1.4
AIRPORT Order
AGUILAR 50102 NWS 37.40 -104.65 6400 32 32 29 0 NA
AKRON 1 N 50114 NWS 40.17 -103.22 4663 48 48 28 0 NA
AKRON 4 E 50109 NWS 40.15 -103.14 4540 35 35 18 0 NA
ALAMOSA 50125 NWS 37.47 -105.88 7536 81 81 18 0 NA

Colorado Design Snow Loads 69

Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
ALLENSPARK 1 NW 50183 NWS 40.20 -105.55 8540 11 11 38 0 NA
ALLENSPARK LODGE 50183 NWS 40.20 -105.53 8450 32 32 38 0 NA
ALTENBERN 50214 NWS 39.50 -108.38 5678 65 65 54 0 NA
AMES 50228 NWS 37.87 -107.88 8700 38 38 78 0 NA
AMY 50242 NWS 38.88 -103.65 5243 26 26 24 0 NA
ANTERO RSVR 50263 NWS 38.99 -105.89 8920 52 52 25 0 NA
ARAPAHOE 50304 NWS 38.85 -102.18 4020 22 22 17 0 NA
AROYA 6 NE 50343 NWS 38.92 -103.08 4793 24 24 6 0 NA
ARRIBA 50348 NWS 39.29 -103.25 5242 17 17 18 0 NA
ASPEN 50370 NWS 39.20 -106.83 8045 77 77 60 0 NA
AYER RCH 50437 NWS 39.02 -104.60 7234 26 26 24 0 NA
AYER RCH 50437 NWS 39.02 -104.60 7234 22 22 24 0 NA
BAILEY 50454 NWS 39.40 -105.48 7730 60 60 45 0 NA
BLANCA 50776 NWS 37.42 -105.52 7815 59 59 13 0 NA
BLOOM 50784 NWS 37.68 -103.95 4484 23 23 36 0 NA
BLUE MESA LAKE 50797 NWS 38.47 -107.17 7568 45 45 40 0 NA
BOND 50810 NWS 39.88 -106.68 6706 16 16 17 0 NA
BONHAM RSVR 50825 NWS 39.10 -107.90 9852 26 26 98 0 NA
BONNY LAKE 50834 NWS 39.63 -102.18 3701 58 58 24 0 NA
BOULDER 50848 NWS 39.99 -105.27 5484 63 63 27 0 NA
BRANDON 50895 NWS 38.45 -102.45 3931 39 39 10 0 NA
BRANSON 50898 NWS 37.02 -103.88 6280 32 32 33 0 NA
BRECKENRIDGE 50909 NWS 39.49 -106.04 9580 48 48 56 0 NA
BRIGGSDALE 50945 NWS 40.64 -104.33 4834 40 40 19 0 NA
BRIGHTON 50950 NWS 39.99 -104.82 4980 40 40 18 0 NA
BROWNS PARK
51017 NWS 40.80 -108.92 5354 30 30 16 0 NA
REFUGE
BUCKHORN MTN 1 E 51060 NWS 40.62 -105.29 7400 25 25 37 0 NA
BUENA VISTA 51071 NWS 38.83 -106.13 7963 113 113 22 0 NA
BURLINGTON 51121 NWS 39.30 -102.26 4197 115 115 27 0 NA
BUTLER RCH 51157 NWS 38.03 -104.47 4852 25 25 21 0 NA
BYERS 5 ENE 51179 NWS 39.74 -104.13 5100 82 82 24 0 NA
CABIN CREEK 51186 NWS 39.66 -105.71 10020 43 43 65 0 NA
CAMPO 7 S 51268 NWS 37.02 -102.56 4118 54 54 16 0 NA
CANON CITY 51294 NWS 38.45 -105.24 5355 111 111 25 0 NA
CASCADE 51384 NWS 37.67 -107.80 8855 9 9 82 0 NA
CASTLE ROCK 51401 NWS 39.41 -104.91 6186 46 46 44 0 NA
CEDAREDGE 51440 NWS 38.90 -107.93 6213 104 104 35 0 NA
CENTER 4 SSW 51458 NWS 37.70 -106.14 7676 58 58 18 0 NA
CHEESMAN 51528 NWS 39.22 -105.28 6880 111 111 44 0 NA

Colorado Design Snow Loads 70

Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
CHERAW 1 N 51539 NWS 38.11 -103.51 4147 19 19 14 0 NA
CHERRY CREEK DAM 51547 NWS 39.65 -104.85 5647 50 50 28 0 NA
CHEYENNE WELLS 51564 NWS 38.82 -102.36 4250 67 67 15 0 NA
CIMARRON 51609 NWS 38.44 -107.56 7010 46 46 46 0 NA
CLIMAX 51660 NWS 39.37 -106.19 11294 63 63 81 0 NA
COAL CREEK CANYON 51681 NWS 39.90 -105.38 8950 19 19 56 0 NA
COCHETOPA CREEK 51713 NWS 38.45 -106.76 8002 65 65 38 0 NA
COLLBRAN 51741 NWS 39.24 -107.97 6043 44 44 33 0 NA
COLORADO NM 51772 NWS 39.10 -108.73 5779 66 66 30 0 NA
CORTEZ 51886 NWS 37.34 -108.59 6167 76 76 21 0 NA
CRAIG 51928 NWS 40.52 -107.55 6255 75 75 21 0 NA
CRAIG 4 SW 51932 NWS 40.45 -107.59 6496 36 36 22 0 NA
CRESTED BUTTE 51959 NWS 38.87 -106.98 8865 99 99 120 0 NA
CRESTONE 1 SE 51964 NWS 37.98 -105.68 8115 31 31 26 0 NA
CRIPPLE CREEK 51973 NWS 38.75 -105.18 9541 6 6 14 0 NA
CUMBRES 52048 NWS 37.02 -106.45 10026 18 18 136 0 NA
DEL NORTE 52184 NWS 37.67 -106.35 7884 58 58 19 0 NA
DELHI 52178 NWS 37.63 -104.02 5092 25 25 20 0 NA
DELTA 52192 NWS 38.75 -108.07 5031 38 38 9 0 NA
DILLON 1 E 52281 NWS 39.63 -106.04 9065 68 68 43 0 NA
DINOSAUR NATL
52286 NWS 40.24 -108.97 5972 46 46 23 0 NA
MONUMNT
DOHERTY RCH 52312 NWS 37.38 -103.88 5135 32 32 18 0 NA
DOLORES 52326 NWS 37.47 -108.50 6953 56 56 43 0 NA
DURANGO 52432 NWS 37.29 -107.87 6665 63 63 39 0 NA
DURANGO WATER
52441 NWS 37.29 -107.86 6750 21 21 39 0 NA
RESOURCE
EADS 2 S 52446 NWS 38.45 -102.78 4217 72 72 14 0 NA
EAGLE FAA AP 52454 NWS 39.65 -106.92 6497 52 52 32 0 NA
EASTONVILLE 1 NNW 52494 NWS 39.08 -104.57 7245 57 57 36 0 NA
EDGEWATER 52557 NWS 39.75 -105.08 5453 50 50 18 0 NA
ELBERT 52593 NWS 39.22 -104.55 6766 11 11 20 0 NA
ELECTRA LAKE 52624 NWS 37.55 -107.80 8406 14 14 57 0 NA
ESTES PARK 52759 NWS 40.38 -105.52 7524 93 93 42 0 NA
EVERGREEN 52790 NWS 39.64 -105.32 6985 41 41 42 0 NA
EVERSOLL RCH 52803 NWS 37.03 -102.07 3583 23 23 12 0 NA
FAIRPLAY 52814 NWS 39.23 -106.00 10007 14 14 23 0 NA
FLAGLER 2 NW 52932 NWS 39.32 -103.08 4980 64 64 19 0 NA
FLEMING 1 W 52944 NWS 40.68 -102.83 4290 76 76 18 0 NA
FLORISSANT FOSSIL
52965 NWS 38.91 -105.29 8379 35 35 24 0 NA
BED

Colorado Design Snow Loads 71

Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
FORDER 8 S 52997 NWS 38.55 -103.68 4749 30 30 14 0 NA
FOUNTAIN 53063 NWS 38.68 -104.70 5565 46 46 25 0 NA
FOUNTAIN 6 NNE 53068 NWS 38.78 -104.62 5965 12 12 25 0 NA
FOWLER 1 SE 53079 NWS 38.12 -104.01 4330 69 69 17 0 NA
FRASER 53113 NWS 39.94 -105.82 8560 86 86 59 0 NA
FRUITA 1 W 53146 NWS 39.17 -108.75 4480 65 65 18 0 NA
FT CARSON 53002 NWS 38.68 -104.77 5869 22 22 16 0 NA
FT COLLINS 53005 NWS 40.61 -105.13 5004 65 65 23 0 NA
FT LEWIS 53016 NWS 37.23 -108.05 7640 63 63 49 0 NA
FT LUPTON 2 SE 53027 NWS 4NA7 -104.78 5023 64 64 30 0 NA
FT MORGAN 53038 NWS 40.26 -103.81 4340 57 57 14 0 NA
GARDNER 53222 NWS 37.77 -105.18 6974 32 32 30 0 NA
GATEWAY 1 SE 53246 NWS 38.67 -108.97 4550 64 64 15 0 NA
GENOA 53258 NWS 39.28 -103.50 5606 69 69 14 0 NA
GEORGETOWN 53261 NWS 39.72 -105.70 8520 50 50 46 0 NA
GLENDEVEY 53340 NWS 40.80 -105.88 8274 9 9 46 0 NA
GLENWOOD SPGS #2 53359 NWS 39.52 -107.32 5895 62 62 30 0 NA
GORE PASS RCH 53423 NWS 40.15 -106.47 7605 6 6 18 0 NA
GRAND JUNCTION 6
53489 NWS 39.04 -108.47 4760 49 49 14 0 NA
ESE
GRAND JUNCTION
53488 NWS 39.13 -108.54 4858 114 114 18 0 NA
WALKER
GRAND LAKE 1 NW 53496 NWS 40.27 -105.83 8720 65 65 56 0 NA
GRAND LAKE 6 SSW 53500 NWS 40.18 -105.87 8288 47 47 38 0 NA
GRAND VALLEY 53508 NWS 39.45 -108.05 5092 16 16 26 0 NA
GRANT 53530 NWS 39.46 -105.68 8675 50 50 36 0 NA
GREAT SAND DUNES
53541 NWS 37.73 -105.51 8183 59 59 24 0 NA
NM
GREELEY 53546 NWS 40.42 -104.68 4652 19 19 18 0 NA
GREELEY UNC 53553 NWS 40.40 -104.70 4715 46 46 20 0 NA
GREEN MTN DAM 53592 NWS 39.88 -106.33 7743 64 64 38 0 NA
GROSS RSVR 53629 NWS 39.94 -105.35 6910 35 35 43 0 NA
GUFFEY 10 SE 53656 NWS 38.68 -105.38 8595 55 55 26 0 NA
GUNNISON 1 N 53662 NWS 38.55 -106.92 7680 58 58 38 0 NA
HAMILTON 53738 NWS 40.37 -107.62 6234 44 44 39 0 NA
HARMON RCH 53783 NWS 37.48 -102.68 4544 11 11 10 0 NA
HARTSEL 53811 NWS 39.03 -105.80 8875 16 16 22 0 NA
HASWELL 53828 NWS 38.45 -103.16 4525 62 62 22 0 NA
HAWTHORNE 53850 NWS 39.93 -105.28 5925 27 27 29 0 NA
HAYDEN 53867 NWS 40.49 -107.25 6467 66 66 43 0 NA
HERMIT 7 ESE 53951 NWS 37.77 -107.11 9048 58 58 46 0 NA

Colorado Design Snow Loads 72

Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
HIGBEE 2 SW 53982 NWS 37.75 -103.47 4252 8 8 6 0 NA
HOHNHOLZ RCH 54054 NWS 40.97 -106.00 7760 28 28 26 0 NA
HOLLY 54076 NWS 38.05 -102.12 3390 61 61 22 0 NA
HOLYOKE 54082 NWS 40.55 -102.34 3780 64 64 20 0 NA
HOT SULPHUR SPGS 2
54129 NWS 4NA5 -106.13 7605 21 21 48 0 NA
SW
HOURGLASS RSVR 54135 NWS 40.58 -105.63 9520 25 25 62 0 NA
HUGO 54172 NWS 39.13 -103.47 5030 24 24 11 0 NA
IDAHO SPRINGS 54234 NWS 39.75 -105.52 7566 19 19 36 0 NA
IDALIA 4 NNE 54242 NWS 39.75 -102.27 3969 52 52 13 0 NA
IGNACIO 1 N 54250 NWS 37.13 -107.63 6437 42 42 32 0 NA
INDEPENDENCE PASS 5
54270 NWS 39.08 -106.62 10558 19 19 96 0 NA
SW
INTER CANYON 54293 NWS 39.57 -105.22 7040 29 29 57 0 NA
JOES 2 SE 54380 NWS 39.63 -102.65 4251 29 29 18 0 NA
JOHN MARTIN DAM 54388 NWS 38.06 -102.93 3814 66 66 22 0 NA
JONES PASS 2 E 54397 NWS 39.77 -105.85 10328 12 12 63 0 NA
JULESBURG 54413 NWS 40.99 -102.27 3469 90 90 25 0 NA
KARVAL 54444 NWS 38.74 -103.54 5075 58 58 18 0 NA
KASSLER 54452 NWS 39.49 -105.10 5586 65 65 38 0 NA
KAUFFMAN 4 SSE 54460 NWS 40.85 -103.90 5253 34 34 14 0 NA
KIM 10 SSE 54546 NWS 37.12 -103.30 5300 25 25 30 0 NA
KIM 15 NNE 54538 NWS 37.45 -103.32 5190 25 25 43 0 NA
KIOWA 4 SW 54585 NWS 39.30 -104.52 6555 10 10 19 0 NA
KIOWA 5 SE 54584 NWS 39.28 -104.43 6355 7 7 12 0 NA
KIT CARSON 6 SE 54603 NWS 38.70 -102.73 4202 65 65 17 0 NA
KREMMLING 54664 NWS 4NA6 -106.38 7406 43 43 36 0 NA
LA JUNTA 20 S 54726 NWS 37.75 -103.48 4210 31 31 31 0 NA
LA JUNTA 4 NNE 54720 NWS 38.05 -103.52 4194 52 52 21 0 NA
LA VETA 54865 NWS 37.50 -105.00 7034 8 8 27 0 NA
LAKE CITY 54734 NWS 38.04 -107.32 8764 57 57 46 0 NA
LAKE GEORGE 8 SW 54742 NWS 38.91 -105.47 8550 53 53 32 0 NA
LAKE MORAINE 54750 NWS 38.82 -104.98 10273 15 15 57 0 NA
LAKEWOOD 54762 NWS 39.75 -105.12 5640 51 51 38 0 NA
LAMAR 54770 NWS 38.09 -102.63 3627 118 118 36 0 NA
LAS ANIMAS 54834 NWS 38.06 -103.22 3890 67 67 30 0 NA
LEADVILLE 54884 NWS 39.25 -106.30 10131 32 32 54 0 NA
LEADVILLE 2 SW 54885 NWS 39.23 -106.32 9938 30 30 30 0 NA
LEMON DAM 54934 NWS 37.38 -107.66 8363 31 31 62 0 NA
LEROY 5 WSW 54945 NWS 40.51 -103.00 4470 65 65 24 0 NA

Colorado Design Snow Loads 73

Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
LIME 3 SE 55001 NWS 38.12 -104.58 4905 17 17 19 0 NA
LIMON 55017 NWS 39.27 -103.68 5368 23 23 8 0 NA
LIMON 10 SSW 55015 NWS 39.15 -103.77 5564 52 52 31 0 NA
LIMON WSMO 55018 NWS 39.18 -103.70 5562 23 23 18 0 NA
LINDON 4 S 55025 NWS 39.68 -103.42 4890 25 25 14 0 NA
LITTLE HILLS 55048 NWS 4NA0 -108.20 6139 43 43 30 0 NA
LITTLETON 55056 NWS 39.62 -105.02 5340 14 14 33 0 NA
LONGMONT 2 ESE 55116 NWS 40.17 -105.07 4953 55 55 16 0 NA
LOVELAND NCWCD 55236 NWS 40.40 -105.11 5040 21 21 20 0 NA
MANASSA 55322 NWS 37.17 -105.94 7690 58 58 13 0 NA
MANCOS 55327 NWS 37.34 -108.30 6960 50 50 28 0 NA
MARVINE 55408 NWS 4NA2 -107.54 7283 25 25 61 0 NA
MARVINE RCH 55414 NWS 4NA3 -107.46 7800 10 10 62 0 NA
MASSADONA 3 E 55422 NWS 40.28 -108.60 6185 22 22 24 0 NA
MAYBELL 55446 NWS 40.52 -108.09 5944 46 46 29 0 NA
MEEKER 55484 NWS 4NA3 -107.90 6236 43 43 21 0 NA
MEEKER #2 55487 NWS 4NA3 -107.92 6351 20 20 30 0 NA
MEREDITH 55507 NWS 39.37 -106.75 7826 42 42 57 0 NA
MESA LAKES RESORT 55520 NWS 39.05 -108.08 9806 8 8 98 0 NA
MESA VERDE NP 55531 NWS 37.20 -108.49 7087 65 65 39 0 NA
MONTE VISTA 55706 NWS 37.57 -106.14 7662 63 63 18 0 NA
MONTROSE #2 55722 NWS 38.49 -107.88 5789 60 60 18 0 NA
MONTROSE 1 55717 NWS 38.48 -107.88 5785 25 25 20 0 NA
MONUMENT 55734 NWS 39.10 -104.87 7079 14 14 26 0 NA
MONUMENT 2 WSW 55730 NWS 39.08 -104.92 7346 15 15 30 0 NA
MT EVANS RSCH STN 55797 NWS 39.65 -105.60 10630 16 16 104 0 NA
NEDERLAND 2 NNE 55878 NWS 39.98 -105.50 8240 18 18 32 0 NA
NEW RAYMER 55922 NWS 40.60 -103.85 4793 38 38 19 0 NA
NEW RAYMER 21 N 55934 NWS 40.93 -103.87 5180 26 26 20 0 NA
NORTH LAKE 55990 NWS 37.22 -105.05 8806 27 27 24 0 NA
NORTHDALE 55970 NWS 37.82 -109.03 6655 52 52 29 0 NA
NORTHGLENN 55984 NWS 39.90 -105.01 5407 29 29 22 0 NA
NORWOOD 56012 NWS 38.13 -108.28 7019 51 51 28 0 NA
NUNN 56023 NWS 40.71 -104.78 5196 17 17 12 0 NA
OLATHE 56081 NWS 38.62 -107.98 5364 7 7 14 0 NA
ORDWAY 2 ENE 56131 NWS 38.22 -103.72 4315 62 62 18 0 NA
ORDWAY 21 N 56136 NWS 38.53 -103.71 4767 33 33 15 0 NA
OTIS 11 NE 56192 NWS 40.27 -102.84 4229 38 38 12 0 NA
OURAY 56203 NWS 38.02 -107.67 7828 58 58 49 0 NA

Colorado Design Snow Loads 74

Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
OVID 56225 NWS 40.96 -102.38 3533 11 11 10 0 NA
PAGOSA SPRINGS 56258 NWS 37.27 -107.02 7181 49 49 46 0 NA
PALISADE 56266 NWS 39.11 -108.35 4751 62 62 12 0 NA
PALISADE LAKES 6 SSE 56271 NWS 37.43 -107.15 8094 20 20 58 0 NA
PALMER LAKE 56280 NWS 39.12 -104.92 7273 10 10 36 0 NA
PAONIA 56306 NWS 38.86 -107.58 5851 9 9 11 0 NA
PAONIA 1 SW 56306 NWS 38.85 -107.62 5576 56 56 26 0 NA
PARACHUTE 56311 NWS 39.45 -108.05 5090 8 8 20 0 NA
PARADOX 1 E 56315 NWS 38.37 -108.95 5282 26 26 20 0 NA
PARADOX 1 W 56318 NWS 38.38 -108.98 5530 18 18 23 0 NA
PARKER 6 E 56326 NWS 39.52 -104.65 6304 50 50 36 0 NA
PARSHALL 10 SSE 56342 NWS 39.92 -106.12 8274 8 8 75 0 NA
PENROSE 56410 NWS 38.45 -105.07 5413 22 22 5 0 NA
PERRY PARK 56430 NWS 39.26 -104.97 6326 7 7 0 0 NA
PITKIN 56513 NWS 38.60 -106.53 9199 27 27 70 0 NA
PLACERVILLE 56520 NWS 38.02 -108.05 7383 59 59 45 0 NA
PLATORO 56559 NWS 37.35 -106.53 9834 6 6 72 0 NA
PUEBLO 6 SSW 56767 NWS 38.18 -104.65 4915 12 12 10 0 NA
PUEBLO ARMY DEPOT 56763 NWS 38.31 -104.35 4682 17 17 16 0 NA
PUEBLO CITY RSVR 56743 NWS 38.28 -104.65 4692 20 20 12 0 NA
PUEBLO FIRE STN #2 56748 NWS 38.27 -104.60 4705 6 6 4 0 NA
PUEBLO RSVR 56765 NWS 38.26 -104.72 4855 38 38 15 0 NA
PUEBLO WB AP 56738 NWS 38.23 -104.63 4806 7 7 11 0 NA
PUEBLO WSO AP 56740 NWS 38.28 -104.50 4671 59 59 12 0 NA
PYRAMID 56797 NWS 40.23 -107.09 8030 23 23 50 0 NA
RALSTON RSVR 56816 NWS 39.83 -105.24 5900 35 35 38 0 NA
RAND 56820 NWS 40.43 -106.17 8630 10 10 26 0 NA
RANGELY 1 E 56832 NWS 4NA9 -108.77 5285 63 63 27 0 NA
RED FEATHER LAKES 2
56925 NWS 40.79 -105.56 8207 34 34 26 0 NA
SE
RED WING 1 WSW 56977 NWS 37.72 -105.32 7900 15 15 0 0 NA
REDSTONE 4 W 56970 NWS 39.20 -107.30 8066 12 12 59 0 NA
RICO 57017 NWS 37.69 -108.03 8821 53 53 89 0 NA
RIDGWAY 57020 NWS 38.15 -107.76 7034 31 31 25 0 NA
RIFLE 57031 NWS 39.53 -107.79 5337 79 79 40 0 NA
RIO GRANDE RSVR 57050 NWS 37.73 -107.27 9686 23 23 59 0 NA
ROCKY FORD 2 SE 57167 NWS 38.04 -103.69 4170 65 65 19 0 NA
RUSH 57287 NWS 38.83 -104.08 6020 54 54 23 0 NA
RUSTIC 9 WSW 57296 NWS 40.71 -105.71 7700 20 20 36 0 NA
RUXTON PARK 57309 NWS 38.84 -104.97 9050 46 46 49 0 NA

Colorado Design Snow Loads 75

Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
RYE 57315 NWS 37.92 -104.93 6796 32 32 36 0 NA
SAGUACHE 57337 NWS 38.08 -106.14 7701 59 59 23 0 NA
SALIDA 57370 NWS 38.53 -106.02 7160 48 48 42 0 NA
SALIDA 3 W 57371 NWS 38.53 -106.05 7488 11 11 30 0 NA
SAN LUIS 1 E 57430 NWS 37.18 -105.43 8060 24 24 8 0 NA
SARGENTS 57460 NWS 38.40 -106.42 8460 40 40 37 0 NA
SARGENTS 6W 57461 NWS 38.40 -106.50 8136 10 10 55 0 NA
SEDALIA 4 SSE 57510 NWS 39.40 -104.95 5975 49 49 28 0 NA
SEDGWICK 57513 NWS 40.94 -102.52 3584 19 19 14 0 NA
SEDGWICK 5 S 57515 NWS 40.86 -102.52 3990 55 55 24 0 NA
SHAW 2 E 57557 NWS 39.55 -103.35 5181 13 13 18 0 NA
SHEEP MTN 57572 NWS 37.72 -105.24 7754 26 26 29 0 NA
SHOSHONE 57618 NWS 39.57 -107.23 5930 60 60 39 0 NA
SILVERTON 57656 NWS 37.81 -107.66 9285 92 92 70 0 NA
SOUTH PLATTE 57816 NWS 39.41 -105.18 6156 10 10 7 0 NA
SPICER 57848 NWS 40.46 -106.46 8368 52 52 49 0 NA
SPRINGFIELD 57862 NWS 37.40 -102.61 4408 29 29 13 0 NA
SPRINGFIELD 7 WSW 57866 NWS 37.37 -102.75 4622 46 46 21 0 NA
SPRINGFIELD 8 S 57867 NWS 37.28 -102.62 4505 15 15 8 0 NA
SQUAW MTN 57881 NWS 39.68 -105.50 11509 13 13 49 0 NA
STATE TURKEY EXP FAR 57928 NWS 37.22 -107.27 6663 11 11 34 0 NA
STEAMBOAT SPRINGS 57936 NWS 40.49 -106.82 6865 94 94 61 0 NA
STERLING 57950 NWS 40.63 -103.21 3974 58 58 18 0 NA
STONINGTON 57992 NWS 37.30 -102.19 3816 48 48 12 0 NA
STRATTON 58008 NWS 39.30 -102.59 4401 55 55 28 0 NA
STRONTIA SPRINGS
58022 NWS 39.43 -105.12 5840 29 29 39 0 NA
DAM
SUGARLOAF RSVR 58064 NWS 39.25 -106.37 9871 60 60 66 0 NA
TACOMA 58154 NWS 37.52 -107.78 7300 38 38 54 0 NA
TACONY 10 SE 58157 NWS 38.38 -104.07 4960 56 56 11 0 NA
TAYLOR PARK 58184 NWS 38.82 -106.61 9179 48 48 72 0 NA
TELLURIDE 58204 NWS 37.95 -107.82 8775 75 75 64 0 NA
TIMPAS 13 SW 58290 NWS 37.67 -103.92 4831 15 15 16 0 NA
TRINIDAD 58429 NWS 37.18 -104.49 6030 61 61 31 0 NA
TRINIDAD FAA AP 58434 NWS 37.25 -104.33 5742 53 53 16 0 NA
TRINIDAD LAKE 58436 NWS 37.15 -104.56 6310 23 23 20 0 NA
TRINIDAD RIVER 58431 NWS 37.17 -104.51 6002 16 16 0 0 NA
TROUT LAKE 58454 NWS 37.83 -107.88 9707 16 16 79 0 NA
TROY 1 SE 58468 NWS 37.13 -103.30 5514 37 37 18 0 NA

Colorado Design Snow Loads 76

Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
TWIN LAKES
58501 NWS 39.09 -106.35 9205 59 59 38 0 NA
RESERVOIR
TWO BUTTES 58510 NWS 37.56 -102.39 4090 20 20 11 0 NA
URAVAN 58560 NWS 38.38 -108.74 5021 50 50 10 0 NA
UTLEYVILLE 58574 NWS 37.27 -103.03 5003 8 8 6 0 NA
VAIL 58575 NWS 39.64 -106.35 8304 27 27 67 0 NA
VALLECITO DAM 58582 NWS 37.38 -107.58 7758 61 61 53 0 NA
VICTOR 58649 NWS 38.72 -105.15 9708 7 7 24 0 NA
VONA 58722 NWS 39.30 -102.73 4505 34 34 18 0 NA
WAGON WHEEL GAP 3
58742 NWS 37.80 -106.83 8507 22 22 32 0 NA
N
WALDEN 58756 NWS 40.74 -106.28 8056 64 64 33 0 NA
WALSENBURG 58781 NWS 37.63 -104.79 6188 64 64 30 0 NA
WALSH 1 W 58793 NWS 37.38 -102.30 3978 46 46 28 0 NA
WATERDALE 58839 NWS 40.43 -105.21 5230 65 65 24 0 NA
WESTCLIFFE 58931 NWS 38.13 -105.47 7860 65 65 34 0 NA
WETMORE 2 S 58986 NWS 38.22 -105.10 6585 19 19 28 0 NA
WETMORE 9 S 58990 NWS 38.13 -105.08 7365 6 6 24 0 NA
WHEAT RIDGE 58994 NWS 39.78 -105.12 5394 33 33 26 0 NA
WIGGINS 7 SW 59025 NWS 40.15 -104.18 4715 11 11 11 0 NA
WILLIAMS FORK DAM 59096 NWS 4NA4 -106.20 7618 31 31 34 0 NA
WINDSOR 59147 NWS 40.47 -104.90 4743 34 34 14 0 NA
WINTER PARK 59175 NWS 39.87 -105.76 9108 65 65 75 0 NA
WOLF CREEK PASS 1 E 59181 NWS 37.48 -106.78 10640 35 35 251 0 NA
WOLF CREEK PASS 4 W 59183 NWS 37.48 -106.87 9436 21 21 196 0 NA
WOODROW 6 NNE 59213 NWS 4NA8 -103.57 4374 20 20 12 0 NA
WOOTTON RCH 59216 NWS 37.00 -104.48 7582 23 23 30 0 NA
WRAY 59243 NWS 4NA6 -102.22 3582 90 90 16 0 NA
YAMPA 59265 NWS 40.16 -106.91 7857 64 64 35 0 NA
YELLOW JACKET 2 W 59275 NWS 37.52 -108.75 6860 40 40 36 0 NA
YUMA 59295 NWS 40.12 -102.72 4140 58 58 30 0 NA
YUMA 10 NW 59297 NWS 40.21 -102.81 4110 24 24 16 0 NA
ALEXANDER LAKE 07K03 SNOW 39.03 -107.97 10160 53 53 109 53 38.9
ANTERO 05L05 SNOW 38.92 -105.97 9300 45 45 39 45 7.1
ANTERO RESERVOIR 05L06 SNOW 39.00 -105.88 9000 29 29 26 29 5.3
APISHAPA 05M07 SNOW 37.33 -105.07 10000 27 27 62 27 17.4
ARROW 05K06 SNOW 39.92 -105.77 9680 52 52 67 52 26.1
BALTIMORE 05K23 SNOW 39.90 -105.58 8800 52 52 55 52 16.2
BEAR RIVER 07J03 SNOW 4NA7 -107.02 9100 28 28 54 28 17.2
BENNETT CREEK 05J33 SNOW 40.67 -105.62 9200 47 47 48 47 13.4

Colorado Design Snow Loads 77

Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
BERTHOUD FALLS 05K13 SNOW 39.78 -105.78 10500 62 62 72 62 23.2
BERTHOUD PASS 05K03 SNOW 39.83 -105.75 9700 77 77 85 77 30.3
BERTHOUD SUMMIT 05K14 SNOW 39.80 -105.78 11300 41 41 92 41 34.8
BIG MEADOWS 06M25 SNOW 37.53 -106.80 9260 44 44 92 44 26.1
BIG SOUTH 05J03 SNOW 40.62 -105.82 8600 75 75 31 75 7.9
BIGELOW DIVIDE 05L03 SNOW 38.10 -105.13 9350 29 29 58 29 14.8
BISON RESERVOIR 05L08 SNOW 38.77 -105.10 10220 15 15 31 15 8.6
BLUE RIVER 06K21 SNOW 39.38 -106.05 10500 56 56 54 56 16.2
BOULDER FALLS 05J25 SNOW 4NA2 -105.57 10000 60 60 74 60 22.9
BOURBON 05M05 SNOW 37.18 -105.10 9600 35 35 54 35 13.0
BROWN CABIN 05M04 SNOW 37.55 -105.40 9600 46 46 54 46 15.7
BURRO MOUNTAIN 07K02 SNOW 39.88 -107.60 9400 72 72 82 72 29.5
BUTTE 06L11 SNOW 38.90 -106.95 10160 25 25 85 25 25.3
BUTTERHILL 06J37 SNOW 40.93 -106.98 7880 34 34 62 34 20.2
CAMERON PASS 05J01 SNOW 40.52 -105.57 10285 76 76 107 76 40.9
CASCADE 07M05 SNOW 37.65 -107.80 8840 51 51 83 51 27.5
CHAMBERS LAKE 05J02 SNOW 40.62 -105.83 9000 76 76 51 76 15.4
COCHETOPA PASS 06L06 SNOW 38.17 -106.60 10000 64 64 38 64 10.5
COLUMBINE DITCH 06K47 SNOW 39.38 -106.25 11576 5 5 84 5 25.4
COLUMBINE LODGE 06J03 SNOW 40.40 -106.60 9160 63 63 99 63 34.9
COLUMBINE PASS 08L02 SNOW 38.42 -108.38 9140 7 7 90 7 28.6
COMO 05K25 SNOW 39.35 -105.92 10370 46 46 45 46 13.5
COPELAND LAKE 05J18 SNOW 40.20 -105.57 8600 43 43 40 43 11.8
CORRAL CREEK 06J17 SNOW 40.15 -106.15 9700 17 17 58 17 18.0
CRESTED BUTTE 07L01 SNOW 38.88 -107.00 8920 74 74 79 74 26.7
CROSHO 07J04 SNOW 40.17 -107.05 9100 2 2 51 2 17.6
CUCHARAS CREEK 05M12 SNOW 37.33 -105.08 9700 38 38 68 38 16.5
CULEBRA 05M03 SNOW 37.22 -105.20 10500 53 53 65 53 24.1
CUMBRES PASS 06M07 SNOW 37.02 -106.45 10020 54 54 104 54 41.0
CUMBRES TRESTLE 06M22 SNOW 37.02 -106.45 10020 19 19 126 19 47.5
DEADMAN HILL 05J06 SNOW 40.80 -105.77 10220 64 64 73 64 26.6
DEER RIDGE 05J17 SNOW 40.40 -105.63 9000 64 64 35 64 10.4
DRY LAKE 06J01 SNOW 40.53 -106.78 8400 51 51 88 51 31.3
ELEVEN MILE 05L07 SNOW 38.95 -105.53 8590 29 29 13 29 3.1
ELK RIVER #2 06J15 SNOW 40.85 -106.97 8700 51 51 79 51 27.1
ELKHORN 06J36 SNOW 40.98 -106.92 8475 11 11 92 11 41.3
EMPIRE 05K10 SNOW 39.77 -105.78 9600 51 51 44 51 14.1
EMPIRE #2 05K29 SNOW 39.77 -105.78 9680 40 40 44 40 12.1
EWING UPPER 06K48 SNOW 39.37 -106.28 11270 5 5 68 5 20.8

Colorado Design Snow Loads 78

Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
FOUR MILE PARK 06K07 SNOW 39.07 -106.43 9700 61 61 48 61 11.8
FREMONT PASS 06K08 SNOW 39.38 -106.20 11400 59 59 76 59 27.7
GENEVA PARK 05K11 SNOW 39.52 -105.72 9600 64 64 35 64 8.0
GLEN MAR RANCH 06K20 SNOW 39.82 -106.05 8750 53 53 55 53 15.7
GORE PASS 06J11 SNOW 4NA8 -106.55 9400 58 58 54 58 16.0
GRANBY 05J16 SNOW 40.15 -106.00 8600 64 64 45 64 14.8
GRAND LAKE 05J19 SNOW 40.27 -105.83 8600 44 44 54 44 16.1
GRAYBACK 06M21 SNOW 37.47 -106.53 11600 40 40 81 40 26.5
GRIZZLY PEAK 05K09 SNOW 39.65 -105.87 11100 48 48 83 48 29.5
GROUNDHOG 08M03 SNOW 37.80 -108.27 8940 35 35 61 35 22.4
HAGERMAN TUNNEL 06K42 SNOW 39.25 -106.50 11150 15 15 92 15 37.0
HAHN'S PEAK 06J14 SNOW 40.80 -106.93 8200 23 23 66 23 22.3
HERMIT LAKE 05L04 SNOW 38.10 -105.63 10400 18 18 50 18 18.0
HIDDEN VALLEY 05J13 SNOW 40.40 -105.65 9480 72 72 64 72 17.8
HIWAY 06M19 SNOW 37.47 -106.78 10750 34 34 120 34 48.1
HOOSIER PASS 06K01 SNOW 39.37 -106.07 11400 53 53 66 53 20.4
HORSESHOE
06K35 SNOW 39.20 -106.13 11220 46 46 61 46 17.7
MOUNTAIN
HOURGLASS LAKE 05J11 SNOW 40.58 -105.63 9360 73 73 52 73 18.1
HUERFANO 05M15 SNOW 37.65 -105.47 10080 19 19 55 19 15.8
IDARADO 07M27 SNOW 37.93 -107.67 9800 12 12 67 12 22.8
INDEPENDENCE PASS 06K04 SNOW 39.07 -106.62 10600 76 76 83 76 27.4
IRONTON PARK 07M06 SNOW 37.97 -107.67 9600 75 75 79 75 28.1
IVANHOE 06K10 SNOW 39.28 -106.55 10400 52 52 86 52 27.4
JEFFERSON CREEK 05K08 SNOW 39.43 -105.87 10280 53 53 53 53 15.5
JOE WRIGHT 05J37 SNOW 40.53 -105.88 10120 23 23 102 23 39.9
JONES PASS 05K21 SNOW 39.77 -105.90 10400 56 56 78 56 26.8
KEYSTONE 07L04 SNOW 38.87 -107.03 9960 49 49 97 49 35.7
KILN 06K30 SNOW 39.32 -106.62 9600 24 24 59 24 17.9
LA MANGA 06M11 SNOW 37.08 -106.38 10120 28 28 114 28 48.3
LA PLATA 08M04 SNOW 37.42 -108.05 9340 39 39 108 39 41.9
LA VETA PASS 05M01 SNOW 37.60 -105.20 9440 74 74 57 74 18.3
LAKE CITY 07M08 SNOW 37.98 -107.25 10160 64 64 48 64 14.7
LAKE HUMPHREY 06M15 SNOW 37.67 -106.87 9000 49 49 71 49 17.8
LAKE IRENE 05J10 SNOW 40.42 -105.82 10700 58 58 90 58 41.0
LAPLAND 05K07 SNOW 39.89 -105.89 9300 73 73 62 73 18.8
LEFT HAND 05J44 SNOW 4NA8 -105.53 9900 3 3 46 3 14.8
LEMON RESERVOIR 07M23 SNOW 37.45 -107.67 8700 44 44 67 44 22.9
LEMON RESERVOIR #2 07M24 SNOW 37.45 -107.65 10000 11 11 80 11 28.1
LIFT 06K27 SNOW 39.15 -106.82 11250 32 32 103 32 33.0

Colorado Design Snow Loads 79

Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
LIZARD HEAD 07M03 SNOW 37.80 -107.93 10200 49 49 83 49 30.4
LIZARD HEAD PASS 07M29 SNOW 37.80 -107.93 10200 11 11 79 11 24.4
LONE CONE 08M07 SNOW 37.90 -108.20 9600 28 28 75 28 25.0
LONG DRAW
05J27 SNOW 40.51 -105.77 9980 29 29 71 29 22.7
RESERVOIR
LONGS PEAK 05J22 SNOW 40.27 -105.58 10500 62 62 72 62 20.9
LOSTMAN 06K49 SNOW 39.12 -106.62 10626 5 5 78 5 23.7
LOVE LAKE 07M10 SNOW 37.67 -107.02 10000 48 48 73 48 21.7
LOVELAND PASS 05K05 SNOW 39.68 -105.90 10800 55 55 71 55 26.9
LYNX PASS 06J06 SNOW 4NA8 -106.67 8880 54 54 65 54 19.2
MANCOS T-DOWN 08M02 SNOW 37.43 -108.17 10000 32 32 101 32 37.6
MC CLURE PASS 07K09 SNOW 39.13 -107.28 9500 40 40 74 40 27.0
MC INTYRE 05J15 SNOW 40.78 -105.93 9100 11 11 50 11 16.3
MC KENZIE GULCH 06K28 SNOW 39.50 -106.75 8500 51 51 44 51 9.7
MESA LAKES 08K04 SNOW 39.05 -108.08 10000 63 63 91 63 32.8
MIDDLE CREEK 07M21 SNOW 37.62 -107.03 11250 12 12 111 12 42.2
MIDDLE FORK
06K12 SNOW 39.78 -106.02 9000 77 77 55 77 17.6
CAMPGROUND
MILNER PASS 05J24 SNOW 40.40 -105.83 9750 57 57 68 57 21.8
MINERAL CREEK 07M14 SNOW 37.85 -107.73 10040 39 39 79 39 24.9
MOLAS LAKE 07M12 SNOW 37.75 -107.68 10500 49 49 82 49 31.2
MONARCH OFFSHOOT 06L09 SNOW 38.52 -106.33 10500 68 27 72 68 26.8
MONARCH PASS 06L04 SNOW 38.52 -106.33 10500 49 49 80 49 30.9
MOSQUITO CREEK 06K34 SNOW 39.28 -106.13 10980 46 46 58 46 16.2
NAST LAKE 06K06 SNOW 39.30 -106.60 8700 71 71 50 71 13.2
NORTH INLET GRAND
05J09 SNOW 40.28 -105.77 9000 74 74 50 74 17.9
LAKE
NORTH LOST TRAIL 07K01 SNOW 39.07 -107.15 9200 55 55 71 55 27.1
NORTH MOUNTAIN 08M09 SNOW 37.93 -108.40 9360 19 19 80 19 29.6
NORTHGATE 06J07 SNOW 40.93 -106.28 8550 43 43 40 43 10.5
OPHIR LOOP 07M18 SNOW 37.92 -107.83 11320 15 15 76 15 24.6
PANDO 06K19 SNOW 39.47 -106.33 9500 38 38 51 38 15.7
PARK CONE 06L02 SNOW 38.82 -106.58 9600 73 73 67 73 22.4
PARK RESERVOIR 07K06 SNOW 39.05 -107.88 9960 51 51 119 51 46.9
PARK VIEW 06J02 SNOW 40.37 -106.10 9160 77 77 54 77 14.6
PASS CREEK 06M18 SNOW 37.55 -106.77 9300 32 32 76 32 27.2
PHANTOM VALLEY 05J04 SNOW 40.40 -105.85 9030 51 51 61 51 19.0
PINE CREEK 05J31 SNOW 40.77 -105.50 7900 39 39 29 39 6.8
PINOS MILL 06M24 SNOW 37.05 -106.42 10000 44 44 146 44 41.0
PLATORO 06M09 SNOW 37.35 -106.55 9880 61 61 90 61 34.6

Colorado Design Snow Loads 80

Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
POOL TABLE
06M14 SNOW 37.80 -106.80 9840 64 64 57 64 15.6
MOUNTAIN
PORCUPINE 07M20 SNOW 37.85 -107.17 10280 62 62 73 62 22.9
PORPHYRY CREEK 06L03 SNOW 38.48 -106.33 10760 59 59 85 59 30.4
RABBIT EARS 06J09 SNOW 40.37 -106.73 9400 49 48 110 49 41.9
RANCH CREEK 05K18 SNOW 39.93 -105.73 9400 56 56 62 56 22.3
RED FEATHER 05J20 SNOW 40.82 -105.65 9000 57 57 43 57 14.7
RED MOUNTAIN PASS 07M15 SNOW 37.88 -107.70 11020 39 39 125 39 49.0
RICO 08M05 SNOW 37.67 -108.03 8700 54 54 60 54 20.4
RIO BLANCO 07J01 SNOW 4NA3 -107.28 8500 71 71 64 71 24.1
RIVER SPRINGS 06M05 SNOW 37.07 -106.27 9300 53 53 55 53 19.8
ROACH 06J12 SNOW 40.87 -106.05 9700 46 46 87 46 31.1
SAINT ELMO 06L05 SNOW 38.70 -106.37 10400 47 46 70 47 22.0
SANTA MARIA 07M17 SNOW 37.82 -107.12 9600 74 74 46 74 13.4
SAWTOOTH 05J45 SNOW 40.13 -105.58 9740 11 11 90 11 3NA
SHRINE PASS 06K09 SNOW 39.53 -106.22 10700 71 71 79 71 27.9
SILVER LAKES 06M04 SNOW 37.38 -106.40 9500 75 75 49 75 15.0
SNAKE RIVER 05K16 SNOW 39.63 -105.90 10000 62 62 56 62 16.0
SOUTH COLONY 05M13 SNOW 37.97 -105.55 11140 21 21 98 21 38.8
SPRUCE CREEK 05L10 SNOW 38.22 -105.68 10940 11 11 68 11 17.6
SPUD MOUNTAIN 07M11 SNOW 37.70 -107.78 10660 42 42 121 42 49.9
SUMMIT RANCH 06K14 SNOW 39.72 -106.17 9400 34 34 52 34 15.3
SUMMITILLE (DISC.) 06M06 SNOW 37.43 -106.60 11500 36 36 100 36 37.5
SUNDANCE 05K22 SNOW 39.57 -105.73 11100 33 33 55 33 15.3
TELLURIDE 07M02 SNOW 37.93 -107.80 8800 74 74 52 74 13.8
TENNESSEE PASS 06K02 SNOW 39.35 -106.33 10200 55 55 62 55 19.1
TENNESSEE PASS #2 06K25 SNOW 39.35 -106.35 10280 26 26 60 26 18.2
TOWER 06J29 SNOW 40.53 -106.68 10500 24 24 175 24 79.2
TRICKLE DIVIDE 07K05 SNOW 39.13 -107.90 10000 47 47 126 47 49.2
TRINCHERA 05M08 SNOW 37.35 -105.23 10860 33 33 62 33 15.6
TROUT CREEK PASS 06L12 SNOW 38.92 -106.05 9720 39 39 36 39 9.3
TROUT LAKE 07M09 SNOW 37.83 -107.88 9780 35 35 73 35 24.2
TROUT LAKE #2 07M28 SNOW 37.83 -107.88 9780 30 30 69 30 21.2
TWIN LAKES TUNNEL 06K03 SNOW 39.08 -106.53 10450 75 75 60 75 18.8
TWO MILE 05J26 SNOW 40.38 -105.67 10500 41 41 80 41 25.8
UNIVERSITY CAMP 05J08 SNOW 4NA3 -105.57 10300 61 61 97 61 38.8
UPPER RIO GRANDE 07M16 SNOW 37.72 -107.27 9400 62 62 73 62 21.6
UPPER SAN JUAN 06M03 SNOW 37.48 -106.83 10200 61 61 142 61 58.6
UTE CREEK 05M17 SNOW 37.62 -105.37 10650 17 17 62 17 19.6
UTE PASS 06K41 SNOW 39.82 -106.10 9550 13 13 62 13 19.3

Colorado Design Snow Loads 81

Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
VALLECITO 07M31 SNOW 37.48 -107.50 10880 12 12 84 12 28.2
VASQUEZ 05K19 SNOW 39.85 -105.82 9600 56 56 72 56 24.3
WARD 05J21 SNOW 4NA7 -105.52 9500 62 62 50 62 12.8
WESTCLIFFE 05L02 SNOW 38.12 -105.58 9400 47 47 53 47 13.2
WESTON 06K26 SNOW 39.07 -106.02 9300 30 30 48 30 8.6
WHISKEY CREEK 05M14 SNOW 37.22 -105.12 10220 8 8 51 8 15.4
WILD BASIN 05J05 SNOW 40.20 -105.60 9600 76 76 69 76 24.5
WILLOW CREEK PASS 06J05 SNOW 40.35 -106.10 9540 72 72 69 72 20.6
WILLOW PARK 05J40 SNOW 40.43 -105.73 10700 9 9 84 9 31.0
WINFIELD MIDDLE 06K50 SNOW 38.98 -106.45 10340 4 4 42 4 12.1
WOLF CREEK PASS 06M01 SNOW 37.47 -106.78 10320 50 50 131 50 55.3
WOLF CREEK SUMMIT 06M17 SNOW 37.48 -106.80 11000 39 39 137 39 58.1
WURTZ LOWER 06K51 SNOW 39.39 -106.36 10690 5 5 70 5 22.6
WURTZ MIDDLE 06K52 SNOW 39.41 -106.37 10436 5 5 80 5 26.8
YAMPA VIEW 06J10 SNOW 40.37 -106.77 8200 62 61 71 62 24.5
APISHAPA 05M07S SNTL 37.33 -105.07 10000 55 0 NA 55 15.7
ARAPAHO RIDGE 06J08S SNTL 40.35 -106.38 10960 13 0 NA 13 29.4
ARROW PILLOW 05K06S SNTL 39.92 -105.75 9680 62 0 NA 62 24.8
BEAR LAKE 05J39S SNTL 40.31 -105.64 9500 37 0 NA 37 27.3
BEAR RIVER 07J03S SNTL 4NA6 -107.01 9080 11 0 NA 11 14.2
BEARTOWN 07M32S SNTL 37.71 -107.51 11600 33 0 NA 33 37.2
BEAVER CK VILLAGE 06K45S SNTL 39.60 -106.51 8500 12 0 NA 12 14.9
BERTHOUD SUMMIT 05K14S SNTL 39.80 -105.78 11300 55 0 NA 55 29.2
BISON LAKE 07K12S SNTL 39.76 -107.36 10880 30 0 NA 30 43.3
BLACK MESA 08M12S SNTL 37.79 -108.18 11580 3 0 NA 3 20.1
BLACK MOUNTAIN 05J28S SNTL 40.89 -105.66 8920 5 0 NA 5 12.6
BRUMLEY 06K40S SNTL 39.09 -106.54 10600 55 0 NA 55 16.8
BUCKSKIN JOE 06K16S SNTL 39.30 -106.11 11150 16 0 NA 16 12.3
BUFFALO PARK 06J18S SNTL 40.23 -106.60 9240 20 0 NA 20 21.4
BURRO MOUNTAIN 07K02S SNTL 39.88 -107.60 9400 80 0 NA 80 33.1
BUTTE 06L11S SNTL 38.89 -106.95 10160 55 0 NA 55 24.7
CASCADE 07M05S SNTL 37.65 -107.81 8880 77 0 NA 77 30.7
CASCADE #2 07M35S SNTL 37.66 -107.80 8920 25 0 NA 25 25.3
CATHEDRAL BLUFFS
08K02S SNTL 39.88 -108.60 8500 8 0 NA 8 26.9
PILLOW
CHAPMAN TUNNEL 06K46S SNTL 39.26 -106.63 10110 8 0 NA 8 21.4
COCHETOPA PASS 06L06S SNTL 38.16 -106.60 10020 11 0 NA 11 5.8
COLUMBINE 06J03S SNTL 40.39 -106.60 9160 80 0 NA 80 37.9
COLUMBINE PASS 08L02S SNTL 38.42 -108.38 9400 29 0 NA 29 33.5
COLUMBUS BASIN 08M10S SNTL 37.44 -108.02 10785 21 0 NA 21 42.3

Colorado Design Snow Loads 82

Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
COPELAND LAKE 05J18S SNTL 40.21 -105.57 8600 35 0 NA 35 9.0
COPPER MOUNTAIN 06K24S SNTL 39.49 -106.17 10550 38 0 NA 38 20.9
CROSHO 07J04S SNTL 40.17 -107.06 9100 29 0 NA 29 19.3
CULEBRA #2 05M03S SNTL 37.21 -105.20 10500 55 0 NA 55 19.3
CUMBRES TRESTLE 06M22S SNTL 37.02 -106.45 10040 55 0 NA 55 50.9
DEADMAN HILL 05J06S SNTL 40.81 -105.77 10220 78 0 NA 78 25.2
DRY LAKE 06J01S SNTL 40.53 -106.78 8400 77 0 NA 77 35.3
ECHO LAKE 05K27S SNTL 39.66 -105.59 10600 17 0 NA 17 12.5
EL DIENTE PEAK 08M06S SNTL 37.79 -108.02 10000 29 0 NA 29 25.2
ELK RIVER 06J15S SNTL 40.85 -106.97 8700 77 0 NA 77 30.4
ELLIOT RIDGE 06K29S SNTL 39.86 -106.42 10520 6 0 NA 6 23.1
FOOL CREEK 05K30S SNTL 39.87 -105.87 11150 4 0 NA 4 25.0
FREMONT PASS 06K08S SNTL 39.38 -106.20 11400 77 0 NA 77 27.5
GLEN COVE 05L11S SNTL 38.88 -105.07 11460 11 0 NA 11 9.5
GRAYBACK 06M21S SNTL 37.47 -106.54 11620 11 0 NA 11 25.6
GRIZZLY PEAK 05K09S SNTL 39.65 -105.87 11100 55 0 NA 55 28.8
HAGERMAN TUNNEL
06K42S SNTL 39.25 -106.50 11150 6 0 NA 6 42.9
PILLOW
HAYDEN PASS 05L12S SNTL 38.29 -105.85 10720 8 0 NA 8 21.0
HIGH LONESOME 05J45S SNTL 4NA4 -105.75 10620 2 0 NA 2 24.8
HOOSIER PASS 06K01S SNTL 39.36 -106.06 11400 77 0 NA 77 22.0
HOURGLASS LAKE 05J11S SNTL 40.58 -105.63 9380 7 0 NA 7 14.6
IDARADO 07M27S SNTL 37.93 -107.68 9800 37 0 NA 37 23.0
INDEPENDENCE PASS 06K04S SNTL 39.08 -106.61 10600 80 0 NA 80 26.9
IVANHOE 06K10S SNTL 39.29 -106.55 10400 70 0 NA 70 21.7
JACKWHACKER GULCH 05K26S SNTL 39.57 -105.80 10960 17 0 NA 17 15.7
JOE WRIGHT 05J37S SNTL 40.53 -105.89 10120 49 0 NA 49 33.3
JONES PASS 05K21S SNTL 39.76 -105.91 10400 16 0 NA 16 20.5
KILN 06K30S SNTL 39.32 -106.61 9600 49 0 NA 49 19.5
LAKE ELDORA 05J41S SNTL 39.94 -105.59 9700 37 0 NA 37 23.9
LAKE IRENE 05J10S SNTL 40.41 -105.82 10700 78 0 NA 78 41.1
LILY POND 06M23S SNTL 37.38 -106.55 11000 64 0 NA 64 36.6
LIZARD HEAD PASS 07M29S SNTL 37.80 -107.92 10200 55 0 NA 55 28.1
LONE CONE 08M07S SNTL 37.89 -108.20 9600 55 0 NA 55 27.7
LONG DRAW RESV 05J27S SNTL 40.51 -105.77 9980 6 0 NA 6 23.0
LOST DOG 06J38S SNTL 40.82 -106.75 9320 17 0 NA 17 35.8
LOVELAND BASIN 05K05S SNTL 39.67 -105.90 11400 23 0 NA 23 28.7
LYNX PASS 06J06S SNTL 4NA8 -106.67 8880 80 0 NA 80 20.8
MANCOS 08M02S SNTL 37.43 -108.17 10000 36 0 NA 36 34.5
MC CLURE PASS 07K09S SNTL 39.13 -107.29 9500 55 0 NA 55 28.5

Colorado Design Snow Loads 83

Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
MCCOY PARK 06K44S SNTL 39.60 -106.54 9480 13 0 NA 13 17.4
MEDANO PASS 05M16S SNTL 37.85 -105.44 9649 20 0 NA 20 10.3
MESA LAKES 08K04S SNTL 39.06 -108.06 10000 29 0 NA 29 28.9
MICHIGAN CREEK 05K28S SNTL 39.44 -105.91 10600 17 0 NA 17 15.3
MIDDLE CREEK 07M21S SNTL 37.62 -107.03 11250 37 0 NA 37 38.3
MIDDLE FORK CAMP 06K12S SNTL 39.80 -106.03 8940 14 0 NA 14 14.5
MINERAL CREEK 07M14S SNTL 37.85 -107.73 10040 55 0 NA 55 28.9
MOLAS LAKE 07M12S SNTL 37.75 -107.69 10500 65 0 NA 65 46.5
MOON PASS 06M26S SNTL 37.97 -106.56 11140 7 0 NA 7 7.9
NAST LAKE 06K06S SNTL 39.30 -106.61 8700 77 0 NA 77 14.2
NAVAL OILSHALE
07K10S SNTL 39.60 -107.95 8800 10 0 NA 10 32.7
PILLOW
NEVER SUMMER 06J27S SNTL 40.40 -105.96 10280 13 0 NA 13 33.3
NIWOT 05J42S SNTL 4NA4 -105.54 9910 77 0 NA 77 24.8
NORTH LOST TRAIL 07K01S SNTL 39.08 -107.14 9200 80 0 NA 80 35.6
OVERLAND RES. 07K14S SNTL 39.09 -107.63 9840 26 0 NA 26 21.1
PARK CONE 06L02S SNTL 38.82 -106.59 9600 79 0 NA 79 22.8
PARK RESERVOIR 07K06S SNTL 39.05 -107.87 9960 55 0 NA 55 46.9
PHANTOM VALLEY 05J04S SNTL 40.40 -105.85 9030 80 0 NA 80 17.7
PORPHYRY CREEK 06L03S SNTL 38.49 -106.34 10760 76 0 NA 76 27.7
RABBIT EARS 06J09S SNTL 40.37 -106.74 9400 62 0 NA 62 44.4
RAWAH 06J20S SNTL 40.71 -106.01 9020 13 0 NA 13 16.7
RED MOUNTAIN PASS 07M33S SNTL 37.89 -107.71 11200 55 0 NA 55 43.5
RIPPLE CREEK 07J05S SNTL 40.11 -107.29 10340 29 0 NA 29 38.6
ROACH 06J12S SNTL 40.88 -106.05 9700 76 0 NA 76 26.9
ROUGH AND TUMBLE 06K43S SNTL 39.03 -106.08 10360 17 0 NA 17 10.4
SAINT ELMO 06L05S SNTL 38.70 -106.37 10540 8 0 NA 8 14.3
SARGENTS MESA 06L13S SNTL 38.29 -106.37 11530 7 0 NA 7 11.1
SCHOFIELD PASS 07K11S SNTL 39.02 -107.05 10700 30 0 NA 30 48.8
SCOTCH CREEK 08M08S SNTL 37.65 -108.01 9100 29 0 NA 29 21.2
SHARKSTOOTH 08M04S SNTL 37.50 -108.11 10720 11 0 NA 11 29.5
SLUMGULLION 07M30S SNTL 37.99 -107.20 11440 37 0 NA 37 18.6
SOUTH COLONY 05M13S SNTL 37.97 -105.54 10800 37 0 NA 37 30.8
SPUD MOUNTAIN 07M11S SNTL 37.70 -107.78 10660 64 0 NA 64 61.6
STILLWATER CREEK 05J12S SNTL 40.23 -105.92 8720 67 0 NA 67 12.8
STUMP LAKES 07M34S SNTL 37.48 -107.63 11200 29 0 NA 29 34.7
SUMMIT RANCH 06K14S SNTL 39.72 -106.16 9400 35 0 NA 35 17.5
TOWER 06J29S SNTL 40.54 -106.68 10500 50 0 NA 50 71.1
TRAPPER LAKE 07K13S SNTL 4NA0 -107.24 9700 30 0 NA 30 31.9
TRINCHERA 05M08S SNTL 37.35 -105.23 10860 49 0 NA 49 15.2

Colorado Design Snow Loads 84

 Table A5.1 Raw data summary
Num Yrs Max Yrs Max
Lat Alt Snow with Depth with SWE
Station Name ID Type (deg) Lon (deg) (ft.) Yrs Depth (in) SWE (in)
UNIVERSITY CAMP 05J08S SNTL 4NA3 -105.58 10300 78 0 NA 78 36.4
UPPER RIO GRANDE 07M16S SNTL 37.72 -107.26 9400 55 0 NA 55 17.4
UPPER SAN JUAN 06M03S SNTL 37.49 -106.84 10200 78 0 NA 78 60.1
UPPER TAYLOR 06L14S SNTL 38.99 -106.75 10640 6 0 NA 6 21.2
UTE CREEK 05M17S SNTL 37.61 -105.37 10650 15 0 NA 15 20.3
VAIL MOUNTAIN 06K39S SNTL 39.62 -106.38 10300 38 0 NA 38 30.6
VALLECITO 07M31S SNTL 37.49 -107.51 10880 35 0 NA 35 32.3
WAGER GULCH 07M37S SNTL 37.88 -107.36 11100 4 0 NA 4 9.9
WEMINUCHE CREEK 07M36S SNTL 37.52 -107.32 10740 5 0 NA 5 15.2
WHISKEY CK 05M14S SNTL 37.21 -105.12 10220 55 0 NA 55 18.9
WILD BASIN 05J05S SNTL 40.20 -105.60 9560 10 0 NA 10 21.3
WILLOW CREEK PASS 06J05S SNTL 40.35 -106.09 9540 78 0 NA 78 2NA
WILLOW PARK 05J40S SNTL 40.43 -105.73 10700 38 0 NA 38 29.6
WOLF CREEK SUMMIT 06M17S SNTL 37.48 -106.80 11000 55 0 NA 55 64.0
ZIRKEL 06J19S SNTL 40.79 -106.60 9340 12 0 NA 12 42.2

A5.3. Assembled Snow Site Data Summary

Statistical data that is computed for each snow site is tabulated in this section. Snow sites with
fewer than 18 years of historical data are not used in this study. Snow sites with 18-29 years are
not used directly for generating the snow map, but information at some of these sites is used
indirectly to inform various decisions relating to the snow map. Many of the snow sites have the
same or similar names as city/town locations that are tabulated in other sections of this report
(e.g. Table 1.1). However, their tabulated snow loads may differ somewhat. That is for two
reasons: (1) the altitude at the city/town is different than the altitude of the snow site location, (2)
the design snow load is increased or decreased from the computed value based on information at
other nearby locations (i.e. fluctuations in snow loads are smoothed by the snow map and
reflected in Table 1.1). The snow load values in this appendix should not be used for design.

Table A5.2 Assembled snow site data

Mean
Annual Max Reliability-
Max Ground 50-year Based
Ground Snow Ground Ground
Long Num Snow Load Snow Snow
Site Name Lat (deg) (deg) Alt (ft) Yrs Load (psf) (psf) Load (psf) Load
AGUILAR 37.4 -104.65 6400 32 11.2 27 32 58
Akron 1 N 40.16 -103.19 4611 65 5.7 26 19 36
Alamosa 37.47 -105.88 7536 81 4.9 19 16 19
ALAMOSA SAN LUIS
VALLEY REGIONAL 37.44 -105.86 7534 53 4.1 16 16 18
AIRPORT
ALEXANDER LAKE 39.03 -107.97 10160 53 132.5 202 200 180

Colorado Design Snow Loads 85

 Table A5.2 Assembled snow site data
Mean
Annual Max Reliability-
Max Ground 50-year Based
Ground Snow Ground Ground
Long Num Snow Load Snow Snow
Site Name Lat (deg) (deg) Alt (ft) Yrs Load (psf) (psf) Load (psf) Load
Allenspark Lodge 40.2 -105.53 8450 32 20.7 55 55 57
ALTENBERN 39.5 -108.38 5678 65 11.9 65 56 69
Ames 37.87 -107.88 8700 38 58.3 135 136 110
Amy 38.88 -103.65 5243 26 3.5 21 23 32
Antero 38.92 -105.97 9300 45 14.2 37 40 39
Antero RSVR 39 -105.89 8949 53 7.7 31 29 28
Apishapa 37.33 -105.07 10000 55 42.6 90 95 79
Arapaho Ridge 40.35 -106.38 10960 13 106.6 153 NA NA
AROYA 6 NE 38.92 -103.08 4793 24 1.2 3 5 8
ARROW PILLOW 39.92 -105.75 9588 75 74.2 136 124 108
ASPEN 39.2 -106.83 8045 77 46.4 94 89 80
AYER RCH 39.02 -104.6 7234 26 8.3 25 47 59
BAILEY 39.4 -105.48 7730 60 17.1 63 64 81
Baltimore 39.9 -105.58 8800 52 39.4 84 74 69
Bear Lake 40.31 -105.64 9500 37 89.6 142 153 138
Bear River 40.07 -107.02 9094 31 58 89 87 81
Beartown 37.71 -107.51 11600 33 117 193 207 178
Beaver Ck Village 39.6 -106.51 8500 12 58.4 77 NA NA
Bennett Creek 40.67 -105.62 9200 47 37.7 70 74 69
Berthoud Falls 39.78 -105.78 10500 62 73.6 121 120 106
Berthoud Pass 39.83 -105.75 9700 77 90 158 139 123
Berthoud Summit 39.8 -105.78 11300 65 106 181 170 148
Big Meadows 37.53 -106.8 9260 44 75.9 136 140 140
Big South 40.62 -105.82 8600 75 17.8 41 41 41
BIGELOW DIVIDE 38.1 -105.13 9350 29 43.2 77 83 76
Bison Lake 39.76 -107.36 10880 30 130.6 225 218 189
BISON RESERVOIR 38.77 -105.1 10220 15 23.1 45 NA NA
Black Mesa 37.79 -108.18 11580 3 91.5 105 NA NA
Black Mountain 40.89 -105.66 8920 5 50.8 66 NA NA
Blanca 37.42 -105.52 7815 59 5.4 13 14 18
BLOOM 37.68 -103.95 4484 22 7.1 36 65 79
BLUE MESA LAKE 38.47 -107.17 7568 45 18.6 54 56 50
Blue River 39.38 -106.05 10500 56 46.1 84 81 71
Bond 39.88 -106.68 6706 16 5.6 16 NA NA
BONHAM RSVR 39.1 -107.9 9852 26 126.5 180 188 172
BONNY LAKE 39.69 -102.22 3828 62 4.6 21 16 33
BOULDER 39.99 -105.27 5484 63 8.3 25 22 43
Boulder Falls 40.02 -105.57 10000 60 72.1 119 122 106

Colorado Design Snow Loads 86

 Table A5.2 Assembled snow site data
Mean
Annual Max Reliability-
Max Ground 50-year Based
Ground Snow Ground Ground
Long Num Snow Load Snow Snow
Site Name Lat (deg) (deg) Alt (ft) Yrs Load (psf) (psf) Load (psf) Load
BOURBON 37.18 -105.1 9600 35 42.4 68 67 63
Brandon 38.45 -102.45 3931 39 1.5 6 7 15
BRANSON 37.02 -103.88 6280 32 7.7 32 34 57
BRECKENRIDGE 39.49 -106.04 9580 48 39.6 89 97 80
BRIGGSDALE 40.64 -104.33 4834 40 2.6 15 13 21
BRIGHTON 39.99 -104.82 4980 40 4.9 14 18 31
Brown Cabin 37.55 -105.4 9600 46 38.9 82 83 76
BROWNS PARK REFUGE 40.8 -108.92 5354 30 3.5 12 15 26
BUCKHORN MTN 1 E 40.62 -105.29 7400 25 20.3 45 51 76
Buena Vista 38.83 -106.13 7963 113 8.6 26 25 30
Buffalo Park 40.23 -106.6 9240 20 68.5 111 123 105
Burlington 39.3 -102.26 4197 115 3.8 25 17 30
Burro Mountain 39.88 -107.6 9400 80 100.5 172 154 139
Butler RCH 38.07 -104.51 4873 28 4.6 18 19 33
Butte 38.89 -106.95 10160 55 79.9 132 139 120
Butterhill 40.93 -106.98 7880 34 69.4 105 111 105
Byers 5 ENE 39.74 -104.13 5100 82 6.1 21 17 34
CABIN CREEK 39.66 -105.71 10020 43 27.1 108 85 64
Cameron Pass 40.52 -105.57 10285 76 147.5 213 209 192
CAMPO 7 S 37.02 -102.56 4118 54 3 12 15 25
CANON CITY 38.45 -105.24 5355 111 4.7 22 17 30
Cascade 37.65 -107.8 8872 77 70.9 160 149 125
CATHEDRAL BLUFFS
39.88 -108.6 8500 8 105 140 NA NA
PILLOW
Cedaredge 38.9 -107.93 6213 104 6.6 39 29 34
CENTER 4 SSW 37.7 -106.14 7676 58 4.6 20 16 18
Chambers Lake 40.62 -105.83 9000 76 44.2 80 80 75
Chapman Tunnel 39.26 -106.63 10110 8 68.6 111 NA NA
CHEESMAN 39.22 -105.28 6880 111 8.8 51 37 60
CHERRY CREEK DAM 39.65 -104.85 5647 50 6.8 26 22 38
CHEYENNE WELLS 38.83 -102.32 4193 67 3.6 13 14 26
CIMARRON 38.44 -107.56 7010 46 12.1 61 77 62
COAL CREEK CANYON 39.9 -105.38 8950 19 40.6 89 108 93
COCHETOPA CREEK 38.45 -106.76 8002 65 16.5 52 50 55
Cochetopa Pass 38.17 -106.6 10003 67 30.5 55 55 54
COLLBRAN 39.24 -107.97 6043 44 9.3 35 38 46
Colorado NM 39.1 -108.73 5779 66 5.2 30 22 30
COLORADO SPRINGS
38.81 -104.68 6148 65 4.6 22 21 36
MUNICIPAL AIRPORT

Colorado Design Snow Loads 87

 Table A5.2 Assembled snow site data
Mean
Annual Max Reliability-
Max Ground 50-year Based
Ground Snow Ground Ground
Long Num Snow Load Snow Snow
Site Name Lat (deg) (deg) Alt (ft) Yrs Load (psf) (psf) Load (psf) Load
Columbine 40.39 -106.6 9160 80 131.3 197 199 179
Columbine Ditch 39.38 -106.25 11576 5 88.8 132 NA NA
Columbine Pass 38.42 -108.38 9349 32 92.5 174 177 150
Columbus Basin 37.44 -108.02 10785 21 124.4 220 258 209
Como 39.35 -105.92 10370 46 33.7 70 62 58
Copeland Lake 40.2 -105.57 8593 67 28.4 61 57 57
Copper Mountain 39.49 -106.17 10550 38 73.7 109 110 99
Corral Creek 40.15 -106.15 9700 17 65.7 94 NA NA
Cortez 37.34 -108.59 6167 76 5.5 19 19 26
CRAIG 40.52 -107.55 6255 75 4.6 20 18 23
CRAIG 4 SW 40.45 -107.59 6496 36 10.8 22 23 33
CRESTED BUTTE 38.88 -106.99 8889 104 78.2 232 164 137
CRESTONE 1 SE 37.98 -105.68 8115 31 13.4 33 37 43
CRIPPLE CREEK 38.75 -105.18 9541 6 7.3 16 NA NA
Crosho 40.17 -107.06 9100 30 63.5 100 108 96
Cucharas Creek 37.33 -105.08 9700 38 49.1 86 94 80
Culebra #2 37.21 -105.2 10500 76 63.1 125 117 103
Cumbres Trestle 37.02 -106.45 10028 97 140.7 271 251 217
Deadman Hill 40.81 -105.77 10220 78 93.4 138 135 123
Deer Ridge 40.4 -105.63 9000 64 28.3 54 55 53
DEL NORTE 37.67 -106.35 7884 58 7.1 22 21 26
DELHI 37.65 -103.98 4994 40 4.8 16 16 28
Delta 38.75 -108.07 5031 38 2.1 6 7 11
DENVER-STAPLETON 39.76 -104.87 5286 121 5.8 32 20 35
Dillon 1 E 39.63 -106.04 9065 68 21 64 60 50
DINOSAUR NATL
40.24 -108.97 5972 46 7.5 21 25 33
MONUMNT
DOHERTY RCH 37.38 -103.88 5135 32 4.2 14 15 28
DOLORES 37.47 -108.5 6953 56 18 55 56 58
Dry Lake 40.53 -106.78 8400 77 117.2 184 179 164
Durango 37.29 -107.87 6686 63 15.3 47 61 64
EADS 2 S 38.45 -102.78 4217 72 2.8 10 11 22
EAGLE FAA AP 39.65 -106.92 6497 52 9.8 36 40 45
EASTONVILLE 1 NNW 39.08 -104.57 7245 57 13.3 42 36 57
Echo Lake 39.66 -105.59 10615 32 64.8 194 277 187
El Diente Peak 37.79 -108.02 10000 29 73.6 131 141 118
ELBERT 39.22 -104.55 6766 11 8.3 18 NA NA
ELECTRA LAKE 37.55 -107.8 8406 14 49.8 91 NA NA

Colorado Design Snow Loads 88

 Table A5.2 Assembled snow site data
Mean
Annual Max Reliability-
Max Ground 50-year Based
Ground Snow Ground Ground
Long Num Snow Load Snow Snow
Site Name Lat (deg) (deg) Alt (ft) Yrs Load (psf) (psf) Load (psf) Load
Eleven Mile 38.95 -105.53 8590 29 3.6 16 20 20
Elk River 40.85 -106.97 8700 77 102.3 158 155 140
ELKHORN 40.98 -106.92 8475 11 135.8 215 NA NA
Elliot Ridge 39.86 -106.42 10520 6 86.1 120 NA NA
EMPIRE 39.77 -105.78 9635 64 43.7 73 69 65
ESTES PARK 40.38 -105.52 7524 93 12.7 57 45 68
EVERGREEN 39.64 -105.32 6985 41 11.7 49 44 72
EVERSOLL RCH 37.03 -102.07 3583 23 2.9 8 10 22
Ewing Upper 39.37 -106.28 11270 5 73.4 108 NA NA
FAIRPLAY 39.23 -106 10007 14 14.1 29 NA NA
FLAGLER 2 NW 39.31 -103.12 5035 73 4.4 15 16 29
FLEMING 1 W 40.68 -102.83 4290 76 5.2 14 16 32
FLORISSANT FOSSIL BED 38.91 -105.29 8379 35 10.3 30 29 34
Fool Creek 39.87 -105.87 11150 4 89.8 130 NA NA
FORT COLLINS 40.61 -105.13 5004 65 6 20 19 34
FOUNTAIN 38.68 -104.7 5565 46 4.2 22 22 33
FOUR MILE PARK 39.07 -106.43 9700 61 29.1 61 61 58
FOWLER 1 SE 38.12 -104.01 4330 69 3.2 13 12 24
FRASER 39.94 -105.82 8560 86 40.8 95 88 77
Fremont Pass 39.38 -106.2 11367 77 96.2 144 140 128
Fruita 1 W 39.17 -108.75 4480 65 2.6 14 11 17
FT CARSON 38.68 -104.77 5869 22 3.8 12 13 23
FT LEWIS 37.23 -108.05 7640 63 28.9 70 76 67
Ft Lupton 2 SE 40.07 -104.78 5023 64 5.6 28 22 35
FT MORGAN 40.26 -103.81 4340 57 2.9 10 11 22
Gardner 37.77 -105.18 6974 32 8.7 31 40 40
GATEWAY 1 SE 38.67 -108.97 4550 64 2.2 11 11 16
Geneva Park 39.52 -105.72 9600 64 21 42 42 42
GENOA 39.28 -103.5 5606 68 3.6 10 10 21
GEORGETOWN 39.72 -105.7 8520 50 13.6 70 45 41
Glen Cove 38.88 -105.07 11460 11 24.1 49 NA NA
GLEN MAR RANCH 39.82 -106.05 8750 53 46.7 82 77 70
Glendevey 40.8 -105.88 8274 9 37.5 69 NA NA
Glenwood SPGS #2 39.52 -107.32 5895 62 9.4 30 33 43
Gore Pass 40.08 -106.55 9400 58 55 83 87 79
Gore Pass RCH 40.15 -106.47 7605 6 8.1 19 NA NA
Granby 40.15 -106 8600 64 39.5 77 69 63
Grand Junction 6 ESE 39.04 -108.47 4760 49 1.7 10 11 13

Colorado Design Snow Loads 89

 Table A5.2 Assembled snow site data
Mean
Annual Max Reliability-
Max Ground 50-year Based
Ground Snow Ground Ground
Long Num Snow Load Snow Snow
Site Name Lat (deg) (deg) Alt (ft) Yrs Load (psf) (psf) Load (psf) Load
Grand Junction Walker 39.13 -108.54 4858 114 2.7 18 14 19
Grand Lake 1 NW 40.27 -105.83 8671 65 48.1 85 86 77
Grand Lake 6 SSW 40.18 -105.87 8288 47 24.6 54 62 59
Grand Valley 39.45 -108.05 5091 24 6.1 23 32 42
Grant 39.46 -105.68 8675 50 21.6 51 56 58
Grayback 37.47 -106.53 11604 43 81.8 138 150 131
Great Sand Dunes NM 37.73 -105.51 8183 59 8.5 30 29 32
Greeley UNC 40.43 -104.77 4712 65 4.9 16 17 32
Green MTN Dam 39.88 -106.33 7743 64 21.3 51 53 48
Grizzly Peak 39.65 -105.87 11100 74 100 153 158 139
Gross RSVR 39.94 -105.35 6910 35 18.7 50 47 93
Groundhog 37.8 -108.27 8940 35 64.1 116 118 103
Guffey 10 SE 38.68 -105.38 8595 55 11.4 34 32 36
Gunnison 1 N 38.55 -106.92 7680 58 12.4 51 44 38
Hagerman Tunnel 39.25 -106.5 11150 17 141.2 223 NA NA
HAHN'S PEAK 40.8 -106.93 8200 23 74.9 116 122 111
Hamilton 40.37 -107.62 6234 44 8 45 75 64
Hartsel 39.03 -105.8 8875 16 11 28 NA NA
Haswell 38.45 -103.16 4525 62 3 19 14 24
Hawthorne 39.93 -105.28 5925 27 11.4 27 31 53
Hayden 40.49 -107.25 6467 66 22.2 52 50 64
Hayden Pass 38.29 -105.85 10720 8 74 109 NA NA
Hermit 7 ESE 37.77 -107.11 9048 58 18.2 70 78 64
HERMIT LAKE 38.1 -105.63 10400 18 51.2 94 104 90
Hidden Valley 40.4 -105.65 9480 72 55.6 93 98 91
High Lonesome 40.04 -105.75 10620 2 98.3 129 NA NA
Hohnholz RCH 40.97 -106 7760 28 14.9 32 39 56
Holly 38.05 -102.12 3390 61 3.5 19 16 29
Holyoke 40.55 -102.34 3780 64 4.4 16 15 30
Hoosier Pass 39.36 -106.06 11400 79 75.4 114 116 103
Horseshoe Mountain 39.2 -106.13 11220 46 57.2 92 95 85
HOT SULPHUR SPGS 2
40.05 -106.13 7605 21 14 68 73 53
SW
Hourglass Lake 40.58 -105.63 9399 76 44 94 82 76
HUERFANO 37.65 -105.47 10080 19 58.5 82 89 89
Hugo 39.13 -103.47 5030 24 3.1 7 9 19
Idaho Springs 39.75 -105.52 7566 19 7.3 47 56 63
Idarado 37.93 -107.68 9800 37 73.3 120 120 106

Colorado Design Snow Loads 90

 Table A5.2 Assembled snow site data
Mean
Annual Max Reliability-
Max Ground 50-year Based
Ground Snow Ground Ground
Long Num Snow Load Snow Snow
Site Name Lat (deg) (deg) Alt (ft) Yrs Load (psf) (psf) Load (psf) Load
Ignacio 1 N 37.13 -107.63 6437 42 6.2 35 42 42
Independence Pass 39.08 -106.62 10595 80 93.5 142 141 125
Inter Canyon 39.57 -105.22 7040 29 26.7 74 95 148
Ironton Park 37.97 -107.67 9600 75 74.3 146 127 111
Ivanhoe 39.29 -106.55 10400 70 94.2 142 146 130
Jackwhacker Gulch 39.57 -105.8 10960 17 56.1 82 NA NA
JEFFERSON CREEK 39.43 -105.87 10280 53 50.8 81 84 79
Joe Wright 40.53 -105.89 10120 49 128.5 207 212 189
Joes 2 SE 39.63 -102.65 4251 29 5.2 14 18 34
John Martin Dam 38.06 -102.93 3814 66 3.3 19 18 32
Jones Pass 39.77 -105.9 10400 59 91.5 139 137 122
Jones Pass 2 E 39.77 -105.85 10328 12 70 104 NA NA
Julesburg 40.98 -102.28 3476 90 4 22 14 29
Karval 38.74 -103.54 5075 58 2.5 14 11 20
Kassler 39.47 -105.1 5665 65 10.8 41 40 62
Kauffman 4 SSE 40.85 -103.9 5253 34 2.4 10 10 18
Keystone 38.87 -107.03 9960 49 104.6 186 190 163
Kiln 39.32 -106.61 9600 49 64.9 101 99 90
Kim 15 NNE 37.45 -103.32 5190 25 8.8 46 51 69
Kiowa 4 SW 39.3 -104.52 6555 10 7.8 16 NA NA
Kiowa 5 SE 39.28 -104.43 6355 7 3.1 8 NA NA
Kit Carson 6 SE 38.7 -102.73 4202 65 2.2 13 10 18
Kremmling 40.06 -106.38 7406 43 12.3 46 46 38
La Junta 20 S 37.75 -103.48 4219 39 5.4 30 28 45
LA MANGA 37.08 -106.38 10120 28 114.4 251 240 212
La Plata 37.42 -108.05 9340 39 92.4 218 203 170
La Veta 37.5 -105 7034 8 13 28 NA NA
La Veta Pass 37.6 -105.2 9440 74 50 95 92 85
Lake City 37.98 -107.25 10160 64 39.9 76 71 68
Lake City 38.04 -107.32 8764 57 22.6 70 66 62
Lake Eldora 39.94 -105.59 9700 37 62.7 124 116 98
Lake George 8 SW 38.91 -105.47 8550 53 10.8 44 33 35
LAKE HUMPHREY 37.67 -106.87 9000 49 37.2 93 87 82
Lake Irene 40.41 -105.82 10700 78 140.3 214 221 198
Lake Moraine 38.82 -104.98 10273 15 29.3 91 NA NA
Lakewood 39.76 -105.11 5510 101 7.2 39 24 42
Lamar 38.09 -102.63 3627 118 4.5 36 20 36
Lapland 39.89 -105.89 9300 73 54.1 98 92 83

Colorado Design Snow Loads 91

 Table A5.2 Assembled snow site data
Mean
Annual Max Reliability-
Max Ground 50-year Based
Ground Snow Ground Ground
Long Num Snow Load Snow Snow
Site Name Lat (deg) (deg) Alt (ft) Yrs Load (psf) (psf) Load (psf) Load
Las Animas 38.06 -103.22 3890 67 3.4 28 18 32
Leadville 39.24 -106.31 10037 62 28.1 85 75 65
Left Hand 40.08 -105.53 9900 3 60.8 77 NA NA
Lemon Dam 37.38 -107.66 8363 31 50.9 100 118 100
Lemon Reservoir 37.45 -107.67 8700 44 50 119 116 96
LEMON RESERVOIR #2 37.45 -107.65 10000 11 68.2 146 NA NA
Leroy 5 WSW 40.51 -103 4470 65 5.3 21 15 30
LIFT 39.15 -106.82 11250 32 106.7 172 177 153
Lily Pond 37.38 -106.55 11000 64 82.8 190 176 150
Limon 39.27 -103.68 5368 23 2.4 5 6 13
Limon 10 SSW 39.16 -103.75 5563 75 4 30 17 29
Lindon 4 S 39.64 -103.39 4990 30 3.8 14 14 25
Little Hills 40 -108.2 6139 43 9.6 31 30 39
Littleton 39.62 -105.02 5340 14 11.2 32 NA NA
Lizard Head Pass 37.8 -107.93 10200 75 88.7 158 154 133
Lone Cone 37.89 -108.2 9600 55 89.3 144 143 129
Long Draw Reservoir 40.51 -105.77 9980 31 77.5 120 132 115
Longmont 2 ESE 40.17 -105.07 4953 55 3.6 12 14 24
Longs Peak 40.27 -105.58 10500 62 64.5 109 110 96
Lost Dog 40.82 -106.75 9320 17 120.4 186 NA NA
Lostman 39.12 -106.62 10626 5 85.9 123 NA NA
Love Lake 37.67 -107.02 10000 48 52.8 113 104 96
Loveland Basin 39.67 -105.9 11400 23 97.3 149 158 140
Loveland NCWCD 40.4 -105.11 5040 21 5.1 16 19 32
LOVELAND PASS 39.68 -105.9 10800 55 87.2 140 133 120
Lynx Pass 40.08 -106.67 8880 80 67.8 108 104 94
Manassa 37.17 -105.94 7690 58 3.8 13 14 16
Mancos 37.34 -108.3 6960 50 8.5 32 29 33
Mancos 37.43 -108.17 10000 36 94.9 196 187 159
Marvine 40.02 -107.54 7283 25 44.5 89 93 86
MARVINE RCH 40.03 -107.46 7800 10 50.4 96 NA NA
Massadona 3 E 40.28 -108.6 6185 22 12.4 23 25 38
Maybell 40.52 -108.09 5944 46 9.9 29 27 37
Mc Clure Pass 39.13 -107.29 9500 66 94 148 156 140
Mc Intyre 40.78 -105.93 9100 11 66.2 85 NA NA
Mc Kenzie Gulch 39.5 -106.75 8500 51 34.2 50 54 53
Mccoy Park 39.6 -106.54 9480 13 70.2 90 NA NA
Medano Pass 37.85 -105.44 9649 20 33.2 54 66 63

Colorado Design Snow Loads 92

 Table A5.2 Assembled snow site data
Mean
Annual Max Reliability-
Max Ground 50-year Based
Ground Snow Ground Ground
Long Num Snow Load Snow Snow
Site Name Lat (deg) (deg) Alt (ft) Yrs Load (psf) (psf) Load (psf) Load
Meeker 40.03 -107.91 6272 63 10.3 32 28 37
Meredith 39.37 -106.75 7826 42 38.8 86 85 77
Mesa Lakes 39.05 -108.07 9984 79 98.5 171 165 143
Mesa Verde NP 37.2 -108.49 7087 65 18.7 49 57 59
Michigan Creek 39.44 -105.91 10600 17 53.7 80 NA NA
Middle Creek 37.62 -107.03 11250 37 105.4 219 210 179
Middle Fork
39.78 -106.02 8991 80 54.9 92 86 78
Campground
Milner Pass 40.4 -105.83 9750 57 68.9 113 110 98
Mineral Creek 37.85 -107.73 10040 65 80.7 150 139 120
Molas Lake 37.75 -107.69 10500 65 110.4 242 204 174
Monarch Offshoot 38.52 -106.33 10500 68 92 161 152 137
Monte Vista 37.57 -106.14 7662 63 6.4 20 19 24
Montrose #2 38.49 -107.88 5788 60 3.1 17 13 20
Monument 39.1 -104.87 7079 14 7.8 27 NA NA
Monument 2 WSW 39.08 -104.92 7346 15 13.4 34 NA NA
Moon Pass 37.97 -106.56 11140 7 30.5 41 NA NA
Mosquito Creek 39.29 -106.12 11024 49 51 84 87 79
Nast Lake 39.3 -106.61 8700 77 44.2 74 75 68
NAVAL OILSHALE
39.6 -107.95 8800 10 114.1 170 NA NA
PILLOW
Nederland 2 NNE 39.98 -105.5 8240 18 20.7 43 51 55
Never Summer 40.4 -105.96 10280 13 98.4 173 NA NA
New Raymer 40.6 -103.85 4793 38 4.4 15 16 28
New Raymer 21 N 40.93 -103.87 5180 26 5.1 16 17 31
Niwot 40.04 -105.54 9910 77 71.3 129 127 109
North Inlet Grand Lake 40.28 -105.77 9000 74 46.8 93 83 75
North Lake 37.22 -105.05 8806 27 12.1 31 30 30
North Lost Trail 39.07 -107.14 9200 80 97.2 185 171 148
North Mountain 37.93 -108.4 9360 19 93 154 170 146
Northdale 37.82 -109.03 6655 52 9.6 32 32 38
NORTHGATE 40.93 -106.28 8550 43 34.5 55 58 56
Northglenn 39.9 -105.01 5407 29 5.8 19 23 39
Norwood 38.13 -108.28 7019 51 6.3 32 29 31
Nunn 40.71 -104.78 5196 17 4 8 NA NA
Olathe 38.62 -107.98 5364 7 3.3 10 NA NA
OPHIR LOOP 37.92 -107.83 11320 15 91.1 128 NA NA
Ordway 2 ENE 38.22 -103.72 4315 62 2.1 14 10 18
Ordway 21 N 38.54 -103.7 4759 63 2.7 11 12 21

Colorado Design Snow Loads 93

 Table A5.2 Assembled snow site data
Mean
Annual Max Reliability-
Max Ground 50-year Based
Ground Snow Ground Ground
Long Num Snow Load Snow Snow
Site Name Lat (deg) (deg) Alt (ft) Yrs Load (psf) (psf) Load (psf) Load
Ouray 38.02 -107.67 7828 58 29.9 71 76 66
Overland Res. 39.09 -107.63 9840 26 64.9 110 136 111
Pagosa Springs 37.27 -107.02 7181 49 23.8 62 61 69
Palisade 39.11 -108.35 4751 62 1.7 8 8 12
Palisade Lakes 6 SSE 37.43 -107.15 8094 20 48.4 90 103 92
Palmer Lake 39.12 -104.92 7273 10 19.3 42 NA NA
PANDO 39.47 -106.33 9500 38 52.7 82 88 79
Paonia 1 SW 38.85 -107.62 5614 65 6.4 24 24 34
Paradox 1 E 38.37 -108.96 5383 44 5.1 20 20 32
Park Cone 38.82 -106.59 9600 79 57.1 119 103 97
Park Reservoir 39.05 -107.88 9960 76 146.2 244 241 210
Park View 40.37 -106.1 9160 77 50.3 76 79 72
Parker 6 E 39.52 -104.65 6304 50 7.2 36 29 49
Parshall 10 SSE 39.92 -106.12 8274 8 29.6 127 NA NA
PASS CREEK 37.55 -106.77 9300 32 69.3 141 139 133
Penrose 38.45 -105.07 5413 22 0.5 2 3 5
Perry Park 39.26 -104.97 6326 7 0 0 NA NA
Phantom Valley 40.4 -105.85 9030 80 55.4 99 98 87
PINE CREEK 40.77 -105.5 7900 39 13.8 35 40 51
Pinos Mill 37.05 -106.42 10000 44 130.4 213 241 207
Pitkin 38.6 -106.53 9199 27 56.8 118 118 111
Placerville 38.02 -108.05 7383 59 14.5 61 64 51
Platoro 37.35 -106.55 9876 61 83.6 180 168 154
Pool Table Mountain 37.8 -106.8 9840 64 30.4 81 80 69
Porcupine 37.85 -107.17 10280 62 49.8 119 112 95
Porphyry Creek 38.49 -106.34 10760 76 91.3 158 144 131
Pueblo WSO AP 38.27 -104.56 4739 66 3.3 21 12 22
Pyramid 40.23 -107.09 8030 23 29.4 74 79 71
Rabbit Ears 40.37 -106.74 9400 65 150.3 231 245 219
Ralston RSVR 39.83 -105.24 5900 35 10.2 39 36 61
Rand 40.43 -106.17 8630 10 18.4 34 NA NA
Rangely 1 E 40.09 -108.77 5285 63 5.2 25 24 37
Rawah 40.71 -106.01 9020 13 55.3 87 NA NA
Red Feather 40.82 -105.65 9000 57 39.5 76 73 67
Red Feather Lakes 2 SE 40.79 -105.56 8207 34 16.1 33 38 45
Red Mountain Pass 37.89 -107.71 11125 65 152.9 255 267 231
Red Wing 1 WSW 37.72 -105.32 7900 15 0 0 NA NA
Redstone 4 W 39.2 -107.3 8066 12 43 92 NA NA

Colorado Design Snow Loads 94

 Table A5.2 Assembled snow site data
Mean
Annual Max Reliability-
Max Ground 50-year Based
Ground Snow Ground Ground
Long Num Snow Load Snow Snow
Site Name Lat (deg) (deg) Alt (ft) Yrs Load (psf) (psf) Load (psf) Load
RICO 37.68 -108.03 8760 63 42.6 144 112 89
Ridgway 38.15 -107.76 7034 31 13 28 30 36
Rifle 39.53 -107.79 5337 79 6.4 42 28 43
Rio Blanco 40.03 -107.28 8500 71 77.8 125 124 112
Ripple Creek 40.11 -107.29 10340 29 128.8 201 209 181
RIVER SPRINGS 37.07 -106.27 9300 53 35.6 103 82 76
Roach 40.87 -106.05 9700 76 98.5 162 153 136
Rocky Ford 2 SE 38.05 -103.6 4176 68 4 18 18 31
Rough And Tumble 39.03 -106.08 10360 17 32.4 54 NA NA
Rush 38.83 -104.08 6020 54 3.2 20 14 25
Rustic 9 WSW 40.71 -105.71 7700 20 13.6 47 55 70
Ruxton Park 38.84 -104.97 9050 46 25.5 76 99 78
Rye 37.92 -104.93 6796 32 10.2 39 48 80
Saguache 38.08 -106.14 7701 59 4.8 27 18 21
Saint Elmo 38.7 -106.37 10420 50 64.9 114 113 101
Salida 38.53 -106.02 7160 48 12.6 55 59 56
Salida 3 W 38.53 -106.05 7488 11 11.7 37 NA NA
San Luis 1 E 37.18 -105.43 8060 24 3.8 7 10 13
Santa Maria 37.82 -107.12 9600 74 24.8 70 62 57
Sargents 38.4 -106.42 8460 40 14.2 53 70 66
Sargents 6W 38.4 -106.5 8136 10 39.6 85 NA NA
Sargents Mesa 38.29 -106.37 11530 7 46.5 58 NA NA
Sawtooth 40.13 -105.58 9740 11 103.3 156 NA NA
Schofield Pass 39.02 -107.05 10700 30 174.1 254 263 237
Scotch Creek 37.65 -108.01 9100 29 60.6 110 127 104
Sedalia 4 SSE 39.41 -104.93 6077 54 10 48 36 61
Sedgwick 5 S 40.88 -102.52 3886 65 6.6 21 20 42
Sharkstooth 37.5 -108.11 10720 11 97.1 153 NA NA
Sheep MTN 37.72 -105.24 7754 26 12.9 35 45 49
Shoshone 39.57 -107.23 5930 60 17.4 43 49 65
Shrine Pass 39.53 -106.22 10700 71 101.9 145 147 133
Silver Lakes 37.38 -106.4 9500 75 34.4 78 72 70
Silverton 37.81 -107.66 9285 92 55.7 118 119 102
Slumgullion 37.99 -107.2 11440 37 69.4 97 103 94
Snake River 39.63 -105.9 10000 62 42.8 83 79 70
South Colony 37.97 -105.54 10800 37 95 160 166 146
SOUTH COLONY 37.97 -105.55 11140 21 125.3 202 222 193
South Platte 39.41 -105.18 6156 10 0.8 4 NA NA

Colorado Design Snow Loads 95

 Table A5.2 Assembled snow site data
Mean
Annual Max Reliability-
Max Ground 50-year Based
Ground Snow Ground Ground
Long Num Snow Load Snow Snow
Site Name Lat (deg) (deg) Alt (ft) Yrs Load (psf) (psf) Load (psf) Load
Spicer 40.46 -106.46 8368 52 33.3 75 75 69
Springfield 7 WSW 37.38 -102.68 4535 54 4.1 18 14 27
SPRUCE CREEK 38.22 -105.68 10940 11 58.3 92 NA NA
Spud Mountain 37.7 -107.78 10660 64 152 320 311 258
Squaw MTN 39.68 -105.5 11509 13 43.4 76 NA NA
STATE TURKEY EXP FAR 37.22 -107.27 6663 11 11.8 39 NA NA
Steamboat Springs 40.49 -106.82 6865 94 43.4 86 79 92
Sterling 40.63 -103.21 3974 58 2.9 14 14 27
Stillwater Creek 40.23 -105.92 8720 67 43 67 67 64
Stonington 37.34 -102.24 3895 64 3.5 26 17 30
Stratton 39.3 -102.65 4441 59 5.3 26 22 37
Stump Lakes 37.48 -107.63 11200 29 91 180 207 163
Sugarloaf RSVR 39.25 -106.37 9871 60 53.2 110 121 106
Summit Ranch 39.72 -106.16 9400 63 51.7 91 89 79
SUMMITILLE (DISC.) 37.43 -106.6 11500 36 112 195 213 184
Sundance 39.57 -105.73 11100 33 51.4 80 83 76
Tacoma 37.52 -107.78 7300 38 32.1 77 72 68
Tacony 10 SE 38.38 -104.07 4960 56 2.5 7 9 16
Taylor Park 38.82 -106.61 9179 48 41.5 122 136 121
Telluride 37.94 -107.81 8788 96 39.7 72 71 66
TENNESSEE PASS 39.35 -106.34 10226 73 57.3 99 98 90
Tower 40.54 -106.68 10500 50 250.4 412 402 355
Trapper Lake 40 -107.24 9700 30 95 166 173 148
TRICKLE DIVIDE 39.13 -107.9 10000 47 159.4 256 243 216
Trinchera 37.35 -105.23 10860 49 52.4 81 85 79
Trinidad 37.18 -104.49 6024 61 6.1 30 19 35
Trinidad FAA AP 37.25 -104.33 5742 53 3.8 12 11 20
Trinidad Lake 37.15 -104.56 6310 23 6 16 21 40
Trout Creek Pass 38.92 -106.05 9720 39 26.4 48 54 53
TROUT LAKE 37.83 -107.88 9766 64 74.7 126 129 113
Troy 1 SE 37.13 -103.3 5428 62 4.5 28 20 34
Twin Lakes Reservoir 39.09 -106.35 9205 59 13 55 46 42
Twin Lakes Tunnel 39.08 -106.53 10513 78 59.7 98 100 90
Two Buttes 37.56 -102.39 4090 20 1.3 7 10 17
TWO MILE 40.38 -105.67 10500 41 89.5 134 142 127
University Camp 40.03 -105.58 10300 78 107 202 188 163
UPPER RIO GRANDE 37.72 -107.26 9447 78 42.1 112 94 87
Upper San Juan 37.49 -106.84 10200 78 171.9 313 303 264

Colorado Design Snow Loads 96

 Table A5.2 Assembled snow site data
Mean
Annual Max Reliability-
Max Ground 50-year Based
Ground Snow Ground Ground
Long Num Snow Load Snow Snow
Site Name Lat (deg) (deg) Alt (ft) Yrs Load (psf) (psf) Load (psf) Load
Upper Taylor 38.99 -106.75 10640 6 68.8 110 NA NA
Uravan 38.38 -108.74 5021 50 1.2 6 9 11
Ute Creek 37.62 -105.37 10650 20 60.9 106 127 109
UTE PASS 39.82 -106.1 9550 13 62.1 100 NA NA
Utleyville 37.27 -103.03 5003 8 1.7 3 NA NA
Vail 39.64 -106.35 8304 27 55.1 110 109 94
Vail Mountain 39.62 -106.38 10300 38 106.2 159 174 153
Vallecito 37.49 -107.51 10880 35 92 168 174 149
Vallecito Dam 37.38 -107.58 7758 61 26.1 78 94 76
Vasquez 39.85 -105.82 9600 56 75.7 126 119 105
Victor 38.72 -105.15 9708 7 14.3 31 NA NA
Wager Gulch 37.88 -107.36 11100 4 39.7 51 NA NA
Wagon Wheel Gap 3 N 37.8 -106.83 8507 22 18 44 63 64
Walden 40.74 -106.28 8056 64 12.1 44 38 34
Walsenburg 37.63 -104.79 6188 64 7.5 28 27 48
Ward 40.07 -105.52 9500 62 34.2 67 67 62
Waterdale 40.43 -105.21 5230 65 5.8 21 19 34
Weminuche Creek 37.52 -107.32 10740 5 73.7 79 NA NA
Westcliffe 38.13 -105.47 7860 65 13.9 45 45 48
WESTCLIFFE 38.12 -105.58 9400 47 40.1 69 72 71
Weston 39.07 -106.02 9300 30 18.4 45 41 41
Wetmore 2 S 38.22 -105.1 6585 19 10.3 26 33 55
Wetmore 9 S 38.13 -105.08 7365 6 20.9 25 NA NA
Whiskey Ck 37.21 -105.12 10220 55 54 98 95 84
Wiggins 7 SW 40.15 -104.18 4715 11 2.5 7 NA NA
Wild Basin 40.2 -105.6 9595 79 68.5 127 124 108
Williams Fork Dam 40.04 -106.2 7618 31 19.6 44 48 43
Willow Creek Pass 40.35 -106.09 9540 78 72 107 105 96
Willow Park 40.43 -105.73 10700 38 96.8 161 175 151
Winfield Middle 38.98 -106.45 10340 4 43.9 63 NA NA
Winter Park 39.87 -105.76 9108 65 73.6 129 119 105
WOLF CREEK PASS 37.47 -106.78 10320 50 160.1 288 269 253
Wolf Creek Pass 1 E 37.48 -106.78 10640 35 160.8 583 496 380
Wolf Creek Pass 4 W 37.48 -106.87 9436 21 167.9 428 447 344
Wolf Creek Summit 37.48 -106.8 10934 65 173.7 333 323 276
Woodrow 6 NNE 40.08 -103.57 4374 20 3.4 8 9 20
Wootton RCH 37 -104.48 7582 23 11.8 35 42 65
Wray 40.06 -102.22 3582 90 3.4 12 11 25

Colorado Design Snow Loads 97

 Table A5.2 Assembled snow site data
Mean
Annual Max Reliability-
Max Ground 50-year Based
Ground Snow Ground Ground
Long Num Snow Load Snow Snow
Site Name Lat (deg) (deg) Alt (ft) Yrs Load (psf) (psf) Load (psf) Load
Wurtz Middle 39.4 -106.37 10563 5 94.8 139 NA NA
Yampa 40.16 -106.91 7857 64 26 47 47 47
Yampa View 40.37 -106.77 8200 62 82 127 121 114
Yellow Jacket 2 W 37.52 -108.75 6860 40 16.3 43 54 57
Yuma 40.19 -102.78 4162 65 5.1 28 20 35
Zirkel 40.79 -106.6 9340 12 150.7 219 NA NA

Colorado Design Snow Loads 98

Appendix 6. Comparison of 2015 Design Ground Snow Loads to
Previous Editions

Table A6.1 Comparison of 2015 design ground snow loads to previous editions

1971 Tabulated Flat Roof

1971 Tabulated Flat Roof

Snow Load Converted to
2007 Tabulated Pg (psf)
2015 Tabulated Pg (psf)

Difference from 2015

Difference from 2015

Equivelent Pg (psf)
Snow Load (psf)
Longitude (deg)
Latitude (deg)

Altitude (ft.)
City/Town

County

Aguilar Las Animas 37.40 -104.65 6390 55 20 -64% 40 57 4%

Air Force
El Paso 39.01 -104.89 7000 55 45 -18%
Academy
Alamosa Alamosa 37.47 -105.87 7540 25 20 -20% 20 29 14%
Alma Park 39.28 -106.06 10360 65 140 115%
Antonito Conejos 37.08 -106.01 7890 40 20 -50% 20 29 -29%
Arvada Jefferson 39.80 -105.09 5350 40 30 -25%
Aspen Pitkin 39.19 -106.82 7890 75 105 40% 75 107 43%
Aurora Adams 39.73 -104.83 5400 40 30 -25%
Avon Eagle 39.63 -106.52 7430 60 120 100%
Bailey Park 39.41 -105.47 7740 80 65 -19%
Basalt Eagle 39.37 -107.03 6610 55 120 118%
Bayfield La Plata 37.23 -107.60 6900 55 55 0%
Beaver Creek Eagle 39.60 -106.52 8080 75 120 60%
Bellvue Larimer 40.63 -105.17 5130 40 35 -13%
Black Hawk Gilpin 39.80 -105.49 8540 85 40 -53%
Blue River Summit 39.38 -106.05 10500 90 70 -22%
Boulder Boulder 40.01 -105.27 5330 40 25 -38% 30 43 7%
Branson Las Animas 37.02 -103.88 6270 55 20 -64%
Breckenridge Summit 39.50 -106.04 9600 80 110 38% 65 93 16%
Brighton Adams 39.99 -104.82 4980 35 20 -43%
Broomfield Broomfield 39.92 -105.09 5390 40 20 -50%
Buena Vista Chaffee 38.84 -106.13 7960 35 20 -43% 40 57 63%
Canon City Fremont 38.44 -105.24 5350 35 20 -43% 30 43 22%
Carbondale Garfield 39.40 -107.21 6170 50 90 80% 40 57 14%
Cascade El Paso 38.90 -104.97 7380 60 80 33%
Castle Rock Douglas 39.37 -104.86 6220 45 30 -33% 40 57 27%
Cedaredge Delta 38.90 -107.93 6230 35 30 -14% 35 50 43%
Centennial Arapahoe 39.58 -104.88 5830 40 30 -25%
Central City Gilpin 39.80 -105.51 8510 85 40 -53% 50 71 -16%

Colorado Design Snow Loads 99

Table A6.1 Comparison of 2015 design ground snow loads to previous editions

1971 Tabulated Flat Roof

1971 Tabulated Flat Roof

Snow Load Converted to
2007 Tabulated Pg (psf)
2015 Tabulated Pg (psf)

Difference from 2015

Difference from 2015

Equivelent Pg (psf)
Snow Load (psf)
Longitude (deg)
Latitude (deg)

Altitude (ft.)
City/Town

County
Collbran Mesa 39.24 -107.96 5980 45 55 22%
Commerce City Adams 39.81 -104.93 5160 35 25 -29%
Conifer Jefferson 39.52 -105.31 8280 100 70 -30%
Cortez Montezuma 37.35 -108.59 6190 30 20 -33% 25 36 19%
Craig Moffat 40.52 -107.55 6200 30 55 83% 35 50 67%
Crawford Delta 38.70 -107.61 6560 45 20 -56%
Creede Mineral 37.85 -106.93 8800 65 55 -15% 65 93 43%
Crested Butte Gunnison 38.87 -106.98 8910 125 100 -20% 100 143 14%
Crestone Saguache 38.00 -105.70 7930 35 30 -14%
Cripple Creek Teller 38.75 -105.18 9490 70 35 -50% 40 57 -18%
De Beque Mesa 39.33 -108.22 4950 30 65 117%
Del Norte Rio Grande 37.68 -106.35 7880 30 20 -33% 20 29 -5%
Denver Denver 39.74 -104.98 5280 35 30 -14% 30 43 22%
Dillon Summit 39.63 -106.04 9110 65 65 0% 50 71 10%
Dinosaur Moffat 40.24 -109.01 5920 35 30 -14%
Dolores Montezuma 37.47 -108.50 6940 55 85 55% 40 57 4%
Dove Creek Dolores 37.77 -108.91 6840 45 50 11% 25 36 -21%
Durango La Plata 37.28 -107.88 6530 55 60 9% 40 57 4%
Eagle Eagle 39.66 -106.83 6600 45 55 22% 40 57 27%
Edwards Eagle 39.64 -106.59 7220 55 100 82%
Elbert Elbert 39.22 -104.54 6720 55 20 -64%
Empire Clear Creek 39.76 -105.68 8620 60 80 33%
Estes Park Larimer 40.43 -105.52 7520 65 45 -31% 40 57 -12%
Evergreen Jefferson 39.63 -105.32 7050 70 45 -36% 40 57 -18%
Fairplay Park 39.22 -106.00 9950 55 95 73% 50 71 30%
Fort Collins Larimer 40.59 -105.08 5000 35 20 -43% 30 43 22%
Fort Garland Costilla 37.43 -105.43 7940 25 40 60%
Fort Morgan Morgan 40.25 -103.80 4330 30 20 -33% 25 36 19%
Fountain El Paso 38.68 -104.70 5550 35 20 -43%
Franktown Douglas 39.39 -104.75 6160 45 25 -44%
Fraser Grand 39.94 -105.82 8580 75 105 40% 65 93 24%
Frisco Summit 39.57 -106.10 9080 65 100 54%
Georgetown Clear Creek 39.71 -105.70 8520 60 65 8% 75 107 79%

Colorado Design Snow Loads 100

Table A6.1 Comparison of 2015 design ground snow loads to previous editions

1971 Tabulated Flat Roof

1971 Tabulated Flat Roof

Snow Load Converted to
2007 Tabulated Pg (psf)
2015 Tabulated Pg (psf)

Difference from 2015

Difference from 2015

Equivelent Pg (psf)
Snow Load (psf)
Longitude (deg)
Latitude (deg)

Altitude (ft.)
City/Town

County
Glenwood
Garfield 39.55 -107.32 5760 40 30 -25% 40 57 43%
Springs
Granby Grand 40.09 -105.94 7980 55 65 18%
Grand Junction Mesa 39.06 -108.55 4590 25 20 -20% 20 29 14%
Grand Lake Grand 40.25 -105.82 8390 70 80 14% 65 93 33%
Greeley Weld 40.42 -104.71 4680 30 30 0% 30 43 43%
Green Mountain
Teller 38.93 -105.02 7760 70 65 -7%
Falls
Gunnison Gunnison 38.55 -106.93 7700 45 40 -11% 50 71 59%
Gypsum Eagle 39.65 -106.95 6310 40 45 13%
Hartsel Park 39.02 -105.80 8870 35 20 -43% 30 43 22%
Hayden Routt 40.50 -107.26 6350 55 70 27% 50 71 30%
Highlands Ranch Douglas 39.54 -104.97 5900 45 30 -33%
Hot Sulphur
Grand 40.07 -106.10 7730 45 90 100%
Springs
Howard Fremont 38.45 -105.84 6720 40 25 -38%
Idaho Springs Clear Creek 39.74 -105.51 7530 60 55 -8% 50 71 19%
Ignacio La Plata 37.12 -107.63 6450 45 30 -33% 40 57 27%
Kremmling Grand 40.06 -106.39 7310 40 75 88% 50 71 79%
La Junta Otero 37.99 -103.54 4080 30 20 -33% 25 36 19%
La Veta Huerfano 37.51 -105.01 7040 50 90 80% 50 71 43%
Lake City Hinsdale 38.03 -107.32 8660 55 70 27% 65 93 69%
Lakewood Jefferson 39.70 -105.08 5520 40 35 -13%
Lamar Prowers 38.09 -102.62 3620 30 30 0% 20 29 -5%
Larkspur Douglas 39.23 -104.89 6730 55 35 -36%
Leadville Lake 39.25 -106.29 10160 75 115 53% 80 114 52%
Littleton Arapahoe 39.61 -105.02 5350 40 40 0%
Livermore Larimer 40.79 -105.22 5900 45 35 -22%
Longmont Boulder 40.17 -105.10 4980 35 20 -43%
Loveland Larimer 40.40 -105.07 4980 35 20 -43%
Lyons Boulder 40.22 -105.27 5370 40 20 -50% 30 43 7%
Mancos Montezuma 37.34 -108.29 7030 45 30 -33% 50 71 59%
Manitou Springs El Paso 38.86 -104.92 6360 50 100 100%
Marble Gunnison 39.07 -107.19 7990 90 155 72%
Meeker Rio Blanco 40.04 -107.91 6240 40 20 -50% 30 43 7%

Colorado Design Snow Loads 101

Table A6.1 Comparison of 2015 design ground snow loads to previous editions

1971 Tabulated Flat Roof

1971 Tabulated Flat Roof

Snow Load Converted to
2007 Tabulated Pg (psf)
2015 Tabulated Pg (psf)

Difference from 2015

Difference from 2015

Equivelent Pg (psf)
Snow Load (psf)
Longitude (deg)
Latitude (deg)

Altitude (ft.)
City/Town

County
Mesa Mesa 39.17 -108.14 5640 35 120 243%
Mesa Verde Montezuma 37.15 -108.52 6770 50 60 20% 50 71 43%
Minturn Eagle 39.59 -106.43 7860 70 170 143%
Monte Vista Rio Grande 37.58 -106.15 7660 25 20 -20% 20 29 14%
Montezuma Summit 39.58 -105.87 10310 105 80 -24%
Montrose Montrose 38.48 -107.88 5810 25 30 20% 25 36 43%
Monument El Paso 39.09 -104.87 6980 60 45 -25%
Nederland Boulder 39.96 -105.51 8230 70 40 -43% 50 71 2%
Newcastle Garfield 39.57 -107.54 5600 35 40 14%
Norwood San Miguel 38.13 -108.29 7010 35 25 -29% 30 43 22%
Nucla Montrose 38.27 -108.55 5790 25 75 200%
Oak Creek Routt 40.28 -106.96 7430 70 85 21%
Ophir San Miguel 37.86 -107.83 9700 125 170 36%
Ouray Ouray 38.02 -107.67 7790 65 50 -23% 50 71 10%
Pagosa Springs Archuleta 37.27 -107.01 7130 75 100 33% 65 93 24%
Palmer Lake El Paso 39.11 -104.91 7300 65 45 -31% 40 57 -12%
Paonia Delta 38.87 -107.59 5680 35 20 -43% 25 36 2%
Parker Douglas 39.52 -104.76 5870 45 30 -33%
Pitkin Gunnison 38.61 -106.52 9220 105 105 0%
Poncha Springs Chaffee 38.51 -106.08 7470 45 70 56%
Pueblo Pueblo 38.25 -104.61 4690 30 20 -33% 25 36 19%
Rangely Rio Blanco 40.09 -108.80 5230 35 20 -43% 20 29 -18%
Red Cliff Eagle 39.65 -106.37 8750 85 110 29%
Rico Dolores 37.69 -108.03 8830 100 145 45%
Ridgway Ouray 38.15 -107.76 7050 40 20 -50%
Rifle Garfield 39.53 -107.78 5350 40 20 -50% 30 43 7%
Rye Pueblo 37.92 -104.93 6800 60 35 -42%
Salida Chaffee 38.53 -106.00 7080 45 50 11% 40 57 27%
San Luis Costilla 37.20 -105.42 7980 30 20 -33%
Sawpit San Miguel 37.99 -108.00 7590 55 85 55%
Sedalia Douglas 39.44 -104.96 5840 50 30 -40%
Silt Garfield 39.55 -107.66 5460 35 30 -14%
Silver Cliff Custer 38.14 -105.45 7990 55 35 -36%
Silver Plume Clear Creek 39.70 -105.73 9100 70 65 -7%

Colorado Design Snow Loads 102

Table A6.1 Comparison of 2015 design ground snow loads to previous editions

1971 Tabulated Flat Roof

1971 Tabulated Flat Roof

Snow Load Converted to
2007 Tabulated Pg (psf)
2015 Tabulated Pg (psf)

Difference from 2015

Difference from 2015

Equivelent Pg (psf)
Snow Load (psf)
Longitude (deg)
Latitude (deg)

Altitude (ft.)
City/Town

County
Silverthorne Summit 39.63 -106.07 8760 65 70 8%
Silverton San Juan 37.81 -107.66 9310 105 140 33% 90 129 22%
Snowmass
Pitkin 39.21 -106.94 8210 90 165 83%
Village
South Fork Rio Grande 37.67 -106.64 8210 70 55 -21%
Steamboat
Routt 40.48 -106.83 6730 85 100 18% 75 107 26%
Springs
Sterling Logan 40.63 -103.21 3940 30 20 -33% 20 29 -5%
Telluride San Miguel 37.94 -107.81 8790 75 100 33% 75 107 43%
Thornton Adams 39.87 -104.97 5350 40 25 -38%
Trinidad Las Animas 37.17 -104.50 6030 45 35 -22% 40 57 27%
Victor Teller 38.71 -105.14 9710 80 40 -50%
Walden Jackson 40.73 -106.28 8100 45 50 11% 50 71 59%
Walsenburg Huerfano 37.62 -104.78 6170 45 25 -44% 35 50 11%
Ward Boulder 40.07 -105.51 9150 75 75 0%
Westcliffe Custer 38.13 -105.47 7870 50 35 -30% 40 57 14%
Westminster Adams 39.84 -105.04 5380 40 25 -38%
Winter Park Grand 39.89 -105.76 9050 100 140 40% 100 143 43%
Woodland Park Teller 38.99 -105.06 8480 85 50 -41%
Yampa Routt 40.15 -106.91 7880 60 50 -17% 50 71 19%

Colorado Design Snow Loads 103