Impact of Global Warming on Snow in Ski Areas - AMS Journals

Impact of Global Warming on Snow in Ski Areas A Case Study Using a Regional ClimateSimulation over the Interior Western United States

CHRISTIAN PHILIPP LACKNERab BART GEERTSa AND YONGGANG WANGc

aDepartment of Atmospheric Science University of Wyoming Laramie Wyomingb Institute for Atmospheric Physics Johannes Gutenberg University Mainz Germany

cDepartment of Atmospheric and Geological Sciences State University of New York at Oswego Oswego New York

(Manuscript received 11 July 2020 in final form 10 February 2021)

ABSTRACT A high-resolution (4 km) regional climate simulation conducted with theWeather Research and Forecasting

Model is used to investigate potential impacts of global warming on skiing conditions in the interior western United States

(IWUS) Recent-past and near-future climate conditions are compared The past climate period is from November 1981 to

October 2011 The future climate applies to a 30-yr period centered on 2050 A pseudondashglobal warming approach is used

with the driver reanalysis dataset perturbed by the CMIP5 ensemble mean model guidance Using the 30-yr retrospective

simulation a vertical adjustment technique is used to determinemeteorological parameters in the complex terrain where ski

areas are located For snow water equivalent (SWE) Snowpack Telemetry sites close to ski areas are used to validate the

technique and apply a correction to SWE in ski areas The vulnerability to climate change is assessed for 71 ski areas in the

IWUS considering SWE artificially produced snow temperature and rain 20 of the ski areas will tend to have fewer than

100 days per season with sufficient natural and artificial snow for skiing These ski areas are located at either low elevations

or low latitudes making these areas the most vulnerable to climate change Throughout the snow season natural SWE

decreases significantly at the low elevations and low latitudes At higher elevations changes in SWE are predicted to not be

significant in the midseason In mid-February SWE decreases by 118 at the top elevations of ski areas and decreases by

258 at the base elevations

KEYWORDS Snowpack Climate change Adaptation

1 Introduction

Snow is a valuable resource that supports many industries

Thus there is a need to evaluate how changing snow cover due

to climate change will impact these industries (Sturm et al

2017) The skiing industry is one such industry that heavily

relies on snow to ensure operations Therefore it can be con-

sidered vulnerable to climate change Years with early melting

periods and inadequate snow cover will becomemore common

in theNorthernHemisphere toward the late twenty-first century

(Rhoades et al 2018) and economic distress from low snow

yearsmight increase within the next three decades (Diffenbaugh

et al 2013) Ski resorts in the western United States can be ex-

pected to be affected by these changes caused by warmer tem-

peratures and decreases in the snow-to-precipitation ratio

(Ashfaq et al 2013) The impact may vary regionally for in-

stance in maritime regions the sensitivity of the snow-cover

duration to global warming was higher than in continental re-

gions (Brown and Mote 2009)

Considering these future changes in snow cover the vul-

nerability of the skiing industry to climate change has been

shown in many studies in different parts of the world Different

areas include but are not limited to Australia (eg Hennessy

et al 2008) Austria (eg Steiger 2010) France (eg Pons et al

2015) and Switzerland (eg Koenig and Abegg 1997) The

various studies project decreased natural snow amounts in-

creased requirements and challenges for snowmaking and

shortened and more variable ski seasons lengths (Steiger et al

2019) Low-elevation ski areas are most impacted by snow-

deficient winters causing less demand for skiing there (eg

Koenig and Abegg 1997) High-elevation ski areas may profit

from the decreased demand in low-elevation ski areas and

experience increasing demand consistent with observations in

Austria that winters with poor snow conditions tend to de-

crease (increase) demand in low-elevation (high elevation) ski

areas (Toumleglhofer et al 2011)Snow conditions in the context of climate change in ski areas

in North America have been investigated in different regions

Studies focusing on Ontario (Scott et al 2003 2019) the

northeastern United States (eg Scott et al 2007 Beaudin and

Huang 2014 Scott et al 2019) Arizona (Bark et al 2010) and

the whole contiguous United States (CONUS) (Wobus et al

2017) concur that potential season lengths are decreasing and

that snowmaking is gaining interest as an adaption technique to

mitigate the effects of climate change Warm winters have

impacted the ski industry in the United States in the past The

winter of 201112 was one of the warmest on record in the

CONUS and as a result saw the fewest ski visits (51 million) in

the period from 199293 to 201920 (NSAA 2021) Beaudin and

Huang (2014) found that in New England climate change may

already have significantly contributed to changes in the local

ski industry Scott et al (2003) pointed out the particular im-

portance of snowmaking in studies on climate impacts in ski

areas to display skiing operations realistically

The objective of this study is to compare downhill skiing

conditions in recreational ski areas in the interior western

United States (IWUS) under past and future climate conditionsCorresponding author Bart Geerts geertsuwyoedu

MAY 2021 LACKNER ET AL 677

DOI 101175JAMC-D-20-01551

2021 American Meteorological Society For information regarding reuse of this content and general copyright information consult the AMS CopyrightPolicy (wwwametsocorgPUBSReuseLicenses)

Unauthenticated | Downloaded 060322 0452 PM UTC

In the winter of 201819 over 41 of the 593 million ski visits in

the United States were in the Rocky Mountain region (NSAA

2021) making it the most visited skiing region in the country

The Rocky Mountains have diverse climate conditions varying

orography and few studies have looked at ski areas in this region

(eg Zimmerman et al 2006 Bark et al 2010 Battaglin et al

2011Wobus et al 2017) Thus combinedwith the importance of

the skiing industry in the region it is of interest to assess the

vulnerability of the local skiing industry to climate change This

study focuses on vulnerability in the sense of exposure to envi-

ronmental stress caused by climate change and adaptive capacity

to mitigate this stress (Adger 2006) For instance exposure to

environmental stress for ski areas could be decreasing natural

snow amounts or rising temperatures negatively impacting ski-

ing operations in the futureArtificial snowmaking is an adaptive

capacity ski areas have to cope with this exposure A regional

climate simulation over the IWUS conducted with the Weather

Research and Forecasting (WRF)Model is used in this study In

the analysis of skiing conditions we examine natural snowfall

snowmaking potential excessive temperature and rain

Additionally so-called snow indicators as suggested by

Abegg et al (2021) will be used for this analysis

Section 2 will describe the methods of this study including

the regional climate simulation determination validation and

correction of meteorological parameters in the ski areas and

definitions of snow indicators Section 3 will present the results

from the analysis of natural snowfall thresholds for tempera-

ture and rain changes in production potential for artificial

snow and the snow indicators In section 4 implications and

limitations of this study are discussed and choices of certain

parameters are justified The main findings will be summarized

in section 5

2 Methods

a WRF regional climate simulation over the IWUS

This study uses a high-resolution (4km) convection-permitting

regional climate simulation over the IWUS conducted with the

WRF Model (Skamarock et al 2019) For details on the model

configuration see Table 1 More details can be found in Wang

et al (2018) The 4-km resolution and similar WRF physics

choices were used inRasmussen et al (2011 2014) who show that

such setup captures the cold-season precipitation distribution and

amount over the Colorado Headwaters region well with a bias

of 10ndash15 compared to Snowpack Telemetry (SNOTEL)

measurements Because of this good performance in simulating

orographic precipitation over complex terrain similar WRF

simulations have been used to assess changes in orographic pre-

cipitation in a changing global climate For instance Li et al

(2019) and Newman et al (2021) both use convection-permitting

4-km WRF simulations to explore the sensitivity of precipitation

and snowpack to climate change in western Canada and Alaska

respectively Liu et al (2017) extended a 4-kmWRF simulation to

cover the entire contiguous United States although over fewer

years than our simulation The surface temperature and pre-

cipitation for the simulation used in our study have been vali-

dated byWang et al (2018) The cold-season precipitation over

the mountains in the IWUS was validated by Jing et al (2017)

Comparing 10 years of model data against SNOTEL data

across the IWUS Jing et al (2017) find a correlation coefficient

of 095 a mean bias of213mm and root-mean-square bias of

65mm of water-equivalent precipitation in the winter (DJF)

They conclude that simulated seasonal precipitation over

mountains can be more accurate than the numerous gridded

gauge-based precipitation datasets in existence [seeHenn et al

(2018) for a discussion of these datasets] a statement con-

firmed in a broader study by Lundquist et al (2019) Thus

these high-resolution regional climate models of the recent

past are almost equivalent to reanalysis data but of sufficient

resolution to capture finescale orographic snowfall patterns

The model output is used to investigate the skiing condi-

tions both in the recent past and the near future The retro-

spective simulation spans the period from November 1981 to

October 2011 This retrospective climate will be referred to as

historic climate To examine the same patterns in the near-

future climate YWang et al (2020 unpublished manuscript)

used a pseudondashglobal warming (PGW) approach with the

driver dataset perturbed by the CMIP5 ensemble mean

model guidance for 2050 under the representative concen-

tration pathway (RCP) 85 (Pachauri et al 2014) The basic

idea of the PGW approach is to apply the guidance from

global climate models (in this case the ensemble mean CMIP5

climate change signal) to the driver dataset of the regional

climate model (Schaumlr et al 1996) This approach has been

used widely including in studies of changes in orographic

precipitation in the Colorado Headwaters (Rasmussen et al

2011 2014 Eidhammer et al 2018) Thus the historic and

future climate conditions correspond to 1990s and 2050s

conditions The use of 30 years of simulations builds statistical

significance in terms of the mean and the spread of snow

years at any ski resort

Relevant for this study are the hourly model output of sur-

face air temperature rain and snow water equivalent (SWE)

Modeled SWE was chosen over modeled snow depth since it

more accurately describes the total amount of available snow

Furthermore wet-bulb temperature is needed to incorporate

production potential for artificial snow (Olefs et al 2010)Wet-

bulb temperature is calculated using air temperature and rel-

ative humidity (Sadeghi et al 2013) The relative humidity is

TABLE 1 Model configurations

Model parameter Model configuration

Spatial resolution 4 km 3 4 km (420 3 410 grid points)

Vertical levels 51 (topped at 50 hPa)

Driver dataset NCEP Climate Forecast System

Reanalysis (Saha et al 2010)

Microphysics Thompson scheme (Thompson et al 2008)

Radiation Rapid Radiative Transfer Model for

GCMs (Iacono et al 2008)

Planetary

boundary layer

Yonsei University scheme (Hong and

Pan 1996)

Surface layer Revised MoninndashObukhov scheme

(Jimeacutenez et al 2012)Land surface Noah-MP scheme (Niu et al 2011 Yang

et al 2011)

678 JOURNAL OF APPL IED METEOROLOGY AND CL IMATOLOGY VOLUME 60


calculated from gridpoint pressure temperature and water

vapor mixing ratio

b Ski areas and SNOTEL sites

Skiing conditions are investigated in different ski areas

throughout the IWUS Snow amounts measured by SNOTEL

stations (Serreze et al 1999) operated by the Natural

Resources Conservation Service (NRCS) are used to vali-

date modeled snow amounts in the ski areas All investigated

ski areas and SNOTEL sites are shown in Fig 1a

A total of 71 ski areas across eight different states (Table 2)

are part of the final study With 22 and 13 ski areas Colorado

andUtah have themost whileArizona and SouthDakota have

only one Not all ski areas within the domain are investigated

Very small ski areas and ski areas closer than 10 grid points to

the edge of the model domain buffer zone were left out

Furthermore seven ski areas where the historic simulation did

not validate well against proximity SNOTEL data were re-

moved from the analysis and are not part of the 71 ski areas

presented in the study

With the help of the online software Google Maps a coor-

dinate close to the spatial midpoint of the ski areas was de-

termined by handMinimum andmaximum elevation of the ski

areas were retrieved from the areasrsquo websites These elevations

will be referred to as base and top these elevations are shown

in Figs 1c and 1d and can be found in Table 2 Ski areas have

the highest elevations in Colorado with top elevations up to

4000m Elevations decrease to the northwest with ski areas in

Idaho having bottom elevations and sometimes also top ele-

vations below 2000m

Each ski area was assigned a SNOTEL site for the validation

of modeled SWE The SNOTEL sites were chosen based on

proximity (close to the ski areas) elevation (between the base

and top elevations) and data completeness (the full 30 years of

the historic climate simulation) On average the SNOTEL sites

were slightly closer to the top elevation (305m below) than the

base elevation (383m above) Only 61 SNOTEL sites were

used since some ski areas are so close to each other that the

same SNOTEL site is used In some cases where no other close

SNOTEL site was found we used a SNOTEL site whose ele-

vation was slightly outside the elevation range of the ski area

which was the case for 10 ski areas or whose data record did not

cover the full 30 years which was the case for 29 ski areas al-

though only for 4 of those it was less than 15 years The ele-

vations of the SNOTEL sites are shown in Fig 1b SNOTEL

data for all sites including elevation coordinates and daily

SWE was retrieved from the NRCS Report Generator 20

(NRCS 2020)

c Meteorological parameters covering the elevation rangeof ski areas

Ski areas are in regions with complex terrain The 4-km

resolution of the model cannot capture all features of the ter-

rain well especially high peaks on which ski areas are often

FIG 1 Study area (a) Red circles indicate ski areas blue triangles indicate select SNOTEL sites (b) Elevations

of SNOTEL sites (c) Base elevations of ski areas (d) Top elevations of ski areas In these maps and in the maps

shown below three clusters of ski areas are zoomed into with insert maps All ski areas are listed in Table 2



TABLE 2 All ski resorts with base and top elevation and median total ski days

Median total ski days

State Ski resort Base elev (m) Top elev (m) Historic Future

AZ Arizona Snowbowl 28040 35050 150 114

CO Arapahoe Basin 32063 39774 216 194

CO Aspen Highlands 24505 37769 176 148

CO Aspen Mountain 24215 34173 161 131

CO Beaver Creek Resort 24688 34867 167 141

CO Breckenridge Ski Resort 29259 39616 200 176

CO Buttermilk Ski Area 23987 30174 134 95

CO Copper Mountain 29601 37528 199 179

CO Crested Butte 28574 37068 192 172

CO Eldora Mountain Resort 28040 32307 183 153

CO Keystone Resort 28284 37818 181 154

CO Loveland Ski Area 32917 39653 225 202

CO Monarch Ski Area 32886 36428 200 173

CO Powderhorn Mountain Resort 24992 30021 143 111

CO Purgatory Resort 26800 32984 160 124

CO Ski Cooper 32002 35660 205 179

CO Snowmass Ski Area 24700 38129 175 147

CO Steamboat Ski Resort 21030 32210 163 139

CO Sunlight Mountain Resort 24032 30158 144 115

CO Telluride Ski Resort 26593 40079 186 157

CO Vail Ski Resort 24749 35264 173 147

CO Winter Park Resort 27431 36757 202 178

CO Wolf Creek Ski Area 31393 36282 214 184

ID Brundage Mountain 17927 23782 186 156

ID Kelly Canyon 17068 20116 107 63

ID Lookout Pass 13715 17220 176 128

ID Pebble Creek 19384 28257 134 92

ID Pomerelle 23651 26705 160 120

ID Schweitzer Mountain 12191 19506 172 134

ID Silver Mountain 12496 19201 142 72

ID Soldier Mountain 17531 21874 124 78

ID Sun Valley 17525 27888 135 105

ID Tamarack 14934 23468 151 97

MT Big Sky 20725 34032 192 169

MT Blacktail Mountain 15959 20347 139 95

MT Bridger Bowl Ski Area 18592 26821 161 130

MT Discovery 20878 24840 193 167

MT Great Divide 17464 22045 145 94

MT Lost Trail 19506 24992 181 155

MT Maverick Mountain 19811 25145 164 142

MT Montana Snowball 15178 23103 177 146

MT Red Lodge Mountain 21384 28699 186 145

MT Showdown 20725 24992 187 162

MT Whitefish Mountain 13606 20777 168 120

NM Angel Fire Resort 26212 32542 126 87

NM Pajarito Mountain 27431 31820 118 80

NM Red River Ski Area 26669 31545 114 77

NM Sipapu 24992 28208 98 58

NM Ski Santa Fe 31545 36803 171 136

NM Taos Ski Valley 28040 38040 164 137

SD Terry Peak 17982 21640 127 93

UT Alta 25998 33734 214 184

UT Beaver Mountain 21823 27004 169 136

UT Brian Head Ski Resort 29259 33435 178 147

UT Brighton Ski Resort 26669 32002 208 175

UT Cherry Peak Resort 17601 21487 120 83

UT Deer Valley 20024 29168 139 99

UT Eagle Point Ski Resort 27735 32307 172 145



located However the elevation dependency of temperature

and snow amounts is an important factor influencing skiing

conditions Therefore a vertical adjustment technique (VAT)

of these parameters was used A 5 3 5 box of grid points

around the grid point closest to the coordinate of each ski area

was determined At these 25 grid points the values of tem-

perature wet-bulb temperature and SWE are used for a linear

regression against elevation For the temperatures this is done

hourly and for SWE daily The linear regression for SWE only

uses each of the 13 grid points with the lowest elevations when

these have above zero SWE This is done since on many days

there is no snow at low elevations and much snow at high el-

evations Thus using the grid points with zero SWE at low

elevations would skew the linear regression to lower SWE

values and the values at higher elevations might be under-

estimated by the linear regression The 12 grid points with the

higher elevations are always used even if SWE at these grid

points is zero to retain information about the elevation of the

snow line Using this technique an approximate value of these

parameters can be determined at every elevation in the area of

the box

Rain in the ski areas is determined by the mean over a 33 3

box of grid points around the grid point closest to the ski area

coordinate Using rain amounts the number of days exceeding

rain of 1mm (lsquolsquowet daysrsquorsquo for simplicity) in each ski area can be

determined This is of interest since wet days might have a

decreased demand for skiing due to the negative impacts of

rain on snow quality for skiing

d Validation and correction of SWE values

While it can be assumed that temperature and wet-bulb

temperature have a linear lapse rate with height this might not

be the case for snow parameters It can be expected that snow

amounts have a positive elevation gradient (eg Lehning et al

2011 Gruumlnewald et al 2013) however the technique using a

linear regression should be validated The VAT is applied to

different SNOTEL sites the same way as described before For

each day of the historic climate simulation the simulated SWE

amount at the elevation of each SNOTEL site is compared to

the measured SWE value at the same time This is illustrated in

Fig 2a showing the SNOTEL site Tower which is the site used

for the ski area Steamboat Ski Resort Colorado Since very

small snow amounts are not of interest in this study only days

are compared on which both the SNOTEL value and the

modeled value exceeded 1mm of SWE At all sites the model

has an overall negative mean bias underestimating the snow-

pack compared to SNOTEL as is the case for SNOTEL site

Tower (Fig 2a) For this reason a correction is applied to the

modeled daily SWE values SWEWRF Since the relative bias is

increasing the lower the SWE values are this correction is

applied in bins of 100mm of SWE The correction is a mean

bias correction (Maraun 2016) First the means of SWE cal-

culated with theVAT SWEWRFbin are determined for each bin

Thereafter the SNOTEL SWE values measured on the same

days as the values in each SWEWRFbin are used to calculate the

corresponding SWESNOTELbin These two means are used to

determine a correction factor for every bin By multiplying

each model value SWEWRF with the correction factor of its

corresponding bin corrected SWE values SWEWRFcorr are

obtained

SWEWRFcorr

5 SWEWRF

3SWE

SNOTELbin

SWEWRFbin

(1)

The results of this correction are illustrated for the same

SNOTEL site in Fig 2b Because of the nature of the correc-

tion themean bias of SWEWRFcorr is 0 Averaged over all sites

the model underestimated SWE by 1431mm relative to

SNOTEL meaning the correction adds on average this value

to the model values Next a linear regression is applied to the

corrected SWE values If the correlation squared R2 is below

01 the ski area corresponding to the SNOTEL site is elimi-

nated from the study due to the insufficient validation of the

simulation This was the case for seven ski areas in the IWUS

(Bogus Basin Idaho Lee Canyon Nevada Nordic Valley

Utah three ski areas in eastern Washington and Hogadon

Wyoming) There are different reasons why the model might

not validate well against SNOTEL at certain sites In

Washington the reason is likely the proximity to the up-

stream boundary where precipitation in the simulation has a

TABLE 2 (Continued)



UT Park City Mountain Resort 20725 30558 151 115

UT Powder Mountain 21039 28717 171 142

UT Snowbasin Resort 19659 28848 122 84

UT Snowbird 23651 33526 202 171

UT Solitude Mountain Resort 24365 31966 197 167

UT Sundance Resort 18592 25145 101 62

WY Grand Targhee 22578 30058 213 184

WY Jackson Hole 19235 31850 179 154

WY Pine Creek 20802 25069 122 91

WY Sleeping Giant Ski Resort 20174 22639 125 96

WY Snow King Mountain 19009 23798 127 95

WY Snowy Range 27431 30479 183 152

WY White Pine 25754 28955 184 162



strong dry bias (Wang et al 2018) For the other sites the

reason might be related to local terrain features influencing

precipitation that are not captured well by the model

For all other 71 ski areas the corrections from the SNOTEL

sites were applied individually to the daily SWE values in the ski

areas assigned to each site Moreover the correction is used in

both historic and future climates Figure 3 shows theR2 for all ski

areas given by their SNOTEL site 52 of the ski areas and all ski

areas in Colorado have R2 values above 05 whereas 13 have

values below 03 including all ski areas in New Mexico The

results at those locations potentially have a larger uncertainty

e Artificial snow

To realistically capture the potential for skiing operations it is

important to consider artificial snow in a climate vulnerability

study (eg Scott et al 2003 Steiger et al 2019) Since artificially

made snow is not modeled in the land surface model it must be

accounted for differently Here the snowmaking production

potential as defined by Olefs et al (2010) will be used From

data from snow gunmanufacturers Olefs et al (2010) calculated

how much snow could be artificially produced by a snow gun

depending on the ambient wet-bulb temperature Tw They gave

values for so-called fan guns and airndashwater guns For simplicity

we use the average of the two with an estimated loss of 10

from sublimation and loss through wind (Olefs et al 2010)

Other than weather (Tw) the production potential is only lim-

ited bywater availability The production potential pp of snow in

cubic meters per hour per gun is given by

pp5 09(24385Tw2 0145) (2)

This equation is valid for 2148C Tw 228C The produced

snow has a density of 400 kgm23 (Olefs et al 2010) Daily

production potential dpp can be calculated from the sum of the

individual hours on the same day

Since artificial snow is not physically modeled it must be

estimated We define artificially provided SWE SWEAPd to

estimate how much artificial snow is on the surface on a given

day Thus this parameter does not describe howmuch artificial

snow is produced on a day SWEAPd is defined as the mean of

the daily production potential over an area of 1000m2 inte-

grated over the previous 7 days

SWEAPd

5dppd27d21

3400 kgm23

1000m2(3)

This area is comparatively small For instance the new snow-

making facilities of Vail Mountain in Colorado have 421 snow

FIG 3 Correlation (R2) between the corrected modeled SWE

values and SWE values measured at the proximity SNOTEL site

FIG 2 SWE validation and correction The red dashed line indicates the 11 line and the red solid line is the linear regression line

Shown are (a) uncorrected and (b) corrected daily SWE values from the VAT at the SNOTEL site against daily SWE values measured by

the same SNOTEL site (Tower Colorado)



guns serving an area of more than 500 acres (Vail Mountain

2020) This means that one snow gun must on average provide

snow for roughly 5000m2 of terrain The small area is chosen to

account for previously produced artificial snow that is still on

the surface Recent snowmaking conditions are accounted for

by the 7-day mean The selection of the area will be discussed

further in section 4

f Snow indicators

To assess the skiing conditions in the different climates a

few specific snow indicators are examined Our approach is

guided by Abegg et al (2021) another climate study focusing

on ski areas The snow indicators used here can be found in

Table 3 For these snow indicators a so-called snow day

(Abegg et al 2021) must be defined where a certain threshold

of natural andor artificial snow is present on the ground and

skiing would be possible To avoid confusion with other cli-

matological terms in the literature we use the term ski day

instead of snow day In this study a ski day is defined as a day

on which the sum of artificially provided SWE SWEAP and

modeled SWE from precipitation SWEWRF exceeds 20 cm

20 cm of SWE correspond to 50-cm snow depth at a snow

density of 400 kgm23 which is a typical value for groomed ski

slopes (Olefs et al 2010) Other studies use 30 cm of snow

depth for this threshold (eg Scott et al 2003 Steiger 2010)

Durand et al (2009) describe 30 cm of snow cover as margin-

ally sufficient for skiing and 50 cm as good SWE is analyzed in

this study instead of snow depth since the modeled snow depth

does not account for compaction of snow on ski slopes

For ski areas with large vertical extents as in this study

skiing conditions should be evaluated at the mean elevation of

the skiable terrain (Scott et al 2017) For simplicity the base

elevation plus one-third of the vertical extent of the ski area is

used as an approximation of themean elevation This elevation

will be referred to as the investigated elevation The higher

50-cm snow-depth threshold is chosen to balance that skiing

conditions at the base elevation are not directly evaluated

Furthermore a key period especially important for the skiing

industry from 15 November to 15 April is examined In other

studies this core season starts on 1 December (eg Koenig and

Abegg 1997) but it is extended here to include the Thanksgiving

holiday period in the United States in late November A mini-

mum of 100 and 120 ski days in the core season have been used

as thresholds indicating the possibility of commercially viable ski

operations from a snow-cover perspective However it should

be mentioned that commercial viability can depend on other

factors than snow cover (Abegg et al 2021) For the interpre-

tation of the snow indicators median values of ski days will be

analyzed since they are more representative of the year-to-year

conditionsMean values of ski days can be strongly influencedby

outlier years in the 30-yr climatology Twomore snow indicators

concern the Christmas (late December) and Thanksgiving (late

November) periods These periods are important for ski areas

because of increased demand for skiing

Aside from these snow indicators we also examine changes

in wet days and days exceeding a mean temperature of 08C(lsquolsquowarm daysrsquorsquo for simplicity) Both parameters are important

as they impact snow conditions and thus skier decisions More

rain on snow and higher temperatures may deteriorate snow

quality

3 Results

a Natural snow

Natural snow amounts in the form of SWE are experiencing

changes between the historic and the future climate As an

example for this Fig 4 shows SWE values at top and base el-

evations of Steamboat Ski Resort for both climates Steamboat

is the northernmost ski resort in Colorado (see Fig 1) With 18

lifts and roughly 12 km2 of skiable terrain it is one of the

largest Top and base elevations are chosen to show the ex-

tremes throughout the ski area In both climates at the top

elevation SWE starts to accumulate in October and reaches its

maximum in April or May in most years The median seasonal

peak in SWE at the top elevation is similar in both climates

about 1200mm However this is reached in late April in the

future climate as compared to mid-May in the historic climate

The time with above zero median SWE is shorter in the future

climate than in the historic climate in October it starts one

week later in June it ends two weeks earlier From January to

March absolute values of SWE are comparable between both

climates Changes are more pronounced at Steamboatrsquos base

elevation Most years only have snow in the months from

November to March in both climates In the historic climate

there is a continuous period of above zero median SWE from

mid-December to late February giving 16 weeks with snow

cover This number halves to 8 in the future climate and the

median in these weeks is always lower than in the historic

climate

For further investigation Fig 5a shows the difference be-

tween Figs 4b and 4a A Studentrsquos t test was conducted to

determine if the means of the weekly SWE distributions be-

tween both climates have a statistically significant difference

from each other on a confidence level of 95 Figure 5a shows

that at the base elevation the mean and median differences are

TABLE 3 Snow indicators and their description

Snow indicator Description

Start date snow period Start date of the longest continuous period

of ski days

End date snow period End date of the longest continuous period

of ski days

Length snow period No of days in the longest continuous

period of ski days

Core-season ski days No of ski days in the core season (15Novndash

15 Apr)

Total ski days No of ski days in a year starting on 15 Sep

Natural ski days No of ski days without artificial

snowmaking in a year starting on

15 Sep

Snow years

Thanksgiving

period

Percentage of years with at least 8 ski days

between 22 Nov and 1 Dec (10 days)

Snow years Christmas

period


between 23 Dec and 1 Jan (10 days)



always negative This means that there is less snow at the base

elevation in the future climate in most years Furthermore the

negative differences between themeans are significant in almost

all weeks except one frommid-November to mid-March At the

top elevation the differences between future climate and his-

toric climate are not comparable to the base elevation While

there is less snow inmost years inNovember andDecember and

after early April the mean and median differences are close to

zero (not significant) from late December to early April These

differences can be positive in some weeks in January February

and March In two weeks in March this positive difference is

statistically significant The negative difference is largest in late

May and early Junewhich can be explainedwith an earlier onset

of strong melting in the future climate

For comparison with a lower-elevation ski area Fig 5b shows

the difference of SWEbetween the future andhistoric climates for

Whitefish Mountain the northernmost ski area in Montana It is

chosen to compare with Steamboat because both are comparable

in size yet Whitefish Mountain is at a considerably lower eleva-

tion on average In this ski area there is significantly less snow in

all weeks between November and May at both top and base el-

evations The comparison indicates that lower elevations aremore

vulnerable to climate change since the natural snow amounts

decrease at all elevations of this ski area Comparisons between

the base and top elevations at other ski areas (not shown) indicate

that the impact of climate change on natural snow is most severe

at the ski areasrsquo base In other words the lack of snow at the base is

expected to increasingly become a bottleneck to ski operations

FIG 4 Box-and-whisker plots of weekly averaged SWE at the Steamboat Ski Resort ski area containing all 30 years of each simulation

Black lines indicate the median and red diamonds are the mean The boxes include values between the lower quartile (25th percentile)

and upper quartile (75th percentile) Whiskers extend up to 15 times the interquartile range Outliers are indicated by circles Shown are

SWE at top and base elevation for the (a) historic and (b) future climates

FIG 5 As in Fig 4 but for the difference between SWE in the future climate SWEfut and historic climate SWEhis at (a) Steamboat Ski

Resort and (b) Whitefish Mountain Red boxes indicate that the difference between the weekly means of both climates is significant as

based on a Studentrsquos t test



Figure 6 shows the relative change of median SWE for all ski

areas at base and top elevations in three select weeks between

historic and future climates December February and April

weeks were chosen as examples for early mid and late-season

weeks The decrease is larger at the base elevations of the ski

areas compared to the top elevations in all weeks shown In

most cases where snow decreases the relative difference is

larger in the weeks in December and April relative to the

February week Averaged over all ski areas the decrease is

118 at the top elevation and 258 at the base elevation in

the February week In the December week the numbers are

146 and 262 in the April week 315 and 318 This

indicates that snow at lower elevations and early and late-

season snow is impacted more by climate change The largest

relative decrease is found in ski areas in Idaho northwestern

Montana and northern Utah where the ski areas have the

lowest elevations In the February week (Fig 6c) SWE in-

creases at most of the high top elevations (over 3000m) in

central Colorado Montana and Wyoming In some ski areas

this increase is also present in the April week No such increase

materializes near the top of ski areas in Arizona southern

Colorado New Mexico and Utah several of which also have

top elevations over 3000m Although the relative decrease is

lower compared to lower-elevation ski areas other factors than

elevation for example latitude may play a major role in the

reduced snow cover in these areas in the future climate

b Temperature rain and snowmaking potential

Figure 7a shows the absolute change in mean warm days at

the investigated elevation in the core season (15 November to

15 April) In the future climate the number of these days is

larger throughout the domain The smallest absolute increase

is in ski areas where both the investigated elevation is over

3000m and the latitude is north of 378N especially in central

Colorado In Loveland Ski Area Colorado the increase is the

smallest from amean of 1 warm day in the historic climate to a

mean of 4 warm days in the future climate With an investi-

gated elevation of 3516m Loveland is the highest of all ski

areas in the domain The largest increase in warm days is found

in ski areas in Idaho and New Mexico The ski areas with the

largest increases are Soldier Mountain Idaho and Sipapu

New Mexico where the mean number of days increases from

43 to 71 and from 50 to 78 respectively The highest total

number of above-freezing days (86 in total) in the future cli-

mate is found in the ski area Sundance Resort Utah Ski areas

in Arizona New Mexico and southern Utah have a larger

increase in above-freezing days than ski areas in Colorado

Montana and Wyoming with comparable investigated eleva-

tions This might be a reason why in the previous section a

difference between these areas in the change in median SWE

was found despite the similar elevations

Figure 7b presents the absolute change in mean wet days

in the core season In Arizona Colorado New Mexico and

high elevations in Utah (investigated elevation over 2600m)

the number of wet days increases on average by only 2 days

Farther northwest this change is larger In Idaho and

Montana the number of wet days increases on average by

8 days The largest increase is found in the ski area Lookout

Pass in northern Idaho The mean number of wet days in the

core season there used to be 21 while it is 35 in the future

climate Wet days are encountered most frequently at

Schweitzer Mountain Idaho in both climates (39 days in the

future increasing from 28 in the historic climate) The in-

crease in wet days and warm days impacts not only SWE

(less snow through melting) but also snow lsquolsquoqualityrsquorsquo for

skiing Since the increase in those days is lower at higher

FIG 6 Relative difference between the median weekly averaged SWE in all ski areas in the historic and future climates during three

different weeks for (a)(c)(e) top elevations and (b)(d)(f) base elevations for the same weeks Black crosses indicate that median SWE is

zero in both climates



elevations snow amounts and snow quality are less af-

fected there

Figures 7c and 7d show the change in mean snowmaking

production potential for two different periods October to

December and January to March While during the first

period the production potential decreases between 4

(Loveland) and 35 (Silver Mountain Idaho) in all ski

areas during the second period production potential in-

creases in 7 very high-elevation ski areas (investigated ele-

vations over 3100 m) in Colorado by up to 14 (Loveland)

The reason for this is that the wet-bulb temperature at these

high elevations was below the lower threshold of 2148C for

snowmaking more often In a warmer climate snowmaking

will become more difficult and the decrease in production

potential is larger early in the season (OctoberndashDecember)

than later (JanuaryndashMarch) This challenge compounds the

decrease in natural snow in a warmer climate snowmaking

is especially important in the early season when only small

accumulations of natural snowfall must be balanced by ar-

tificial snow production to ensure skiing operations Overall

high-elevation ski areas in Colorado have the smallest de-

creases in production potential (or increases in the second

period as mentioned) while low-elevation ski areas in Idaho

northwestern Montana and Utah have a larger decrease

c Snow indicators

Figure 8a shows the first 6 snow indicators from Table 3 for

Steamboat Themedians of all snow indicators have fewer days

in the future climate and the median start date of the snow

period is later while the end date is earlier The median start

date of the snow period used to be 16 November it is antici-

pated to be 11 days later (27 November) in the future climate

The median end date falls on 15 April in the historic climate

and on 2 April in the future climate Consequently the median

snow period length decreases by 22 days from 149 to 127 days

Similar conclusions apply to the number of ski days in the core

season (decreasing from 145 to 132 days median values) the

number of total ski days (decreasing from 163 to 139 days) and

the number of natural ski days (decreasing from 142 to

118 days) The low difference between total and core-season

ski days in the future climate means that ski days will be almost

exclusively confined to the core season The decreasing amount

of natural ski days signifies the increased importance of

snowmaking in the future climate Furthermore the difference

FIG 7 (a) Total change in days exceeding a mean temperature of 08C at the investigated elevation (b) Total

change in days exceeding rain amount of 1mm over the ski area (c)(d) Change in total production potential in two

different periods



betweenmaximum andminimum values increases in the future

climate indicating that there might be more interannual vari-

ability in the snow indicators

For comparison with a lower-elevation ski area Fig 8b

shows Whitefish Mountain In this ski area changes are

stronger than for Steamboat While in the historic climate the

median snow indicators all have higher values at Whitefish

Mountain compared to Steamboat (except natural ski days)

they are all lower in the future climate For instance the me-

dian number of core-season ski days decreases by 31 days from

147 to 116 days Median natural ski days decrease by 60 days

from 111 to 51 days This indicates that at this location (and

other similar low-elevation ski areas) a sufficient amount of ski

days (at least 100 days) cannot be achieved without consider-

able snowmaking efforts in the future The comparison of these

two cases shows again that lower-elevation ski areas such as

Whitefish Mountain are more vulnerable to climate change

than higher-elevation areas such as Steamboat

Figure 9 presents the median number of ski days in the core

season for the historic and future climates and the absolute

change of this snow indicator In the historic climate 18 ski

areas (most at high elevation 10 in Colorado 4 in Utah 2 in

Montana and 1 each in Idaho andWyoming) have a median of

152 core-season ski days meaning that all days in this period

are ski days In the future climate only 3 ski areas achieve this

distinction (2 in Colorado 1 in Wyoming) In Colorado and

some ski areas with high elevations in Montana Utah and

Wyoming the decrease in ski days in the core season is lower

compared to Arizona Idaho and New Mexico ski areas The

ski area with the largest decrease is Silver Mountain where the

median number of core-season ski days almost halves from 135

to 71 days The lowest median number of ski days can be found

at Sipapu ski area with 96 and 58 ski days in the historic and

future climate respectively In the historic climate the 100-day

threshold for median core-season ski days is reached in 70 ski

areas and the 120-day threshold in 61 ski areas In the future

climate these numbers decrease to 51 and 44 Table 4 shows

the average median core-season ski days by states in both cli-

mates Idaho has the largest decrease with 24 and Colorado

the lowest decrease with 7 In Montana Utah and South

DakotaWyoming the decreases are all comparable with 13

Based on a Studentrsquos t test the change in all snow indicators is

significant on a 95confidence level except for a few ski areas for

the start date (7) and core-season ski days (2) Comparingmedian

core-season ski days with the total number of ski days (Table 2)

the number of ski areas where ski days are almost exclusively in

the core season increases in a warmer climate a minority of ski

areas (31) have at least 10 ski days outside the core season in the

future climate whereas most of them (52) did in the past

Figure 9d displays the absolute decrease in median core-

season ski days against the investigated elevation with an in-

dication of the latitude It corroborates that low-elevation and

low-latitude ski areas are most vulnerable to climate change A

relation between elevation and decrease in ski days could be

inferred but is not clear since latitude plays a role as well for

comparable investigated elevations ski areas with lower lati-

tudes have larger decreases For comparable decreases the ski

areas with higher investigated elevations are at lower latitudes

The elevation dependency becomes clearer when only com-

paring ski areas with similar latitudes With a few exceptions

ski areas with higher investigated elevations have smaller de-

creases than ski areas of comparable latitude with lower in-

vestigated elevations The exceptions imply that there are

other factors such as the local terrain that could influence the

vulnerability of ski areas to climate change

To signify the increased importance of snowmaking in the

future Fig 10 shows the median natural ski days Averaged

over all ski areas the number of median natural ski days de-

creases from 107 to 76 days between the two climates In the

future climate only 22 ski areas have more than 100 median

natural ski days Therefore the majority of ski areas will need

snowmaking to sustain sufficiently long ski seasons in the fu-

ture Similar to Fig 9d Fig 10d shows that there is a relation

between the decrease in natural ski days elevation and lati-

tude Note that a low decrease in natural ski days can be caused

by an already low amount in the historic climate

The last two snow indicators from Table 3 concern the

Christmas period and the Thanksgiving period Figure 11

FIG 8 Various snow indicators in both climates for (a) Steamboat Ski Resort and (b) Whitefish Mountain For each pair the left box-

and-whisker plot is for the historic climate and the right box-and-whiskers plot is for the future climate The box-and-whiskers plots have

same definitions as in Fig 4



shows the percentage of the 30 years in both simulations that

have at least 8 ski days in each 10-day period These snow in-

dicators quantify the fraction of years when ski operations are

possible during the defined periods Ski operations during the

Thanksgiving period become more difficult or impossible in a

warmer climate (Figs 11ab) The largest decrease is found at

Schweitzer Mountain with a decrease from 87 to 43 of

the years with enough natural or artificial snow around

Thanksgiving Three ski areas used to reach 100 (every

Thanksgiving was skiable) Loveland and Arapahoe Basin

have the highest percentage in the future climate with 93

Sipapu has the lowest percentages in both climates with 37

in the historic and 13 in future climate In the historic cli-

mate 48 ski areas had at least 70 of the years reach the

threshold of 8 days Compared to this only 15 ski areas reach

70 in the future climate 11 of which are in Colorado and 2

each in Wyoming and Montana Averaged over all ski areas

the percentage drops from 77 of the years to 50 These

numbers indicate that in most years in the future climate

skiing will be confined to fewer and higher-elevation ski areas

during the Thanksgiving period

The Christmas period (Figs 10cd) looks better not all ski

areas have a decreasing percentage of years with at least 8 ski

days in the period In 23 ski areas the percentage does not

decrease (for 7 of those ski areas it increases in the future cli-

mate) These 23 ski areas are in Colorado (15) Utah (4)

Montana (2) and Wyoming (2) 15 of these have an investi-

gated elevation of at least 2800m the lowest of them is Lost

Trail Montana For all other ski areas the percentage of good

years decreases around Christmas by up to 47 at Silver

Mountain (93 of the years reached the threshold in the his-

toric climate compared to 46 of the years in the future cli-

mate) In the historic climate 67 ski areas reached the 8-day

threshold in at least 90 of the years This number decreases

to 38 in the future climate most of which are in Colorado (18)

Montana (5) Utah (8) or Wyoming (4) Averaged over all ski

areas the percentage of years with adequate snow around

Christmas drops from 95 of the years to 86 This decrease

is not as large as during the Thanksgiving period The per-

centage of years only decreases substantially at low elevations

(mostly Idaho) and the low latitudes of Arizona and New

Mexico At higher elevations the Christmas period sees no or

FIG 9 Map with all ski areas showing median core-season ski days for (a) the historic climate and (b) the future

climate along with (c) the absolute difference between (b) and (a) The range in which the investigated elevation

falls is indicated by the shape of the symbols Also shown is a plot of the absolute difference between (b) and (a) vs

investigated elevation with an indication of the state(symbol shapes) and latitude (color shades)



little change between the climates Overall Christmas skiing

could become more challenging in the future but is not as im-

pacted as the Thanksgiving period

4 Discussion

The results presented herein have implications for the future

of the skiing industry Themost serious implication is that 20 of

the ski areas fall below a median of 100 core-season ski days

and an additional 7 ski areas fall below 120 such days If the

applied emissions scenario (RCP85) materializes insufficient

amounts of snow pose risks to the viable operations of those ski

areas Moreover the increased interannual variability of ski

days mentioned in section 3c will make it harder for ski areas to

consistently plan their ski season ahead of time Ski areas could

try to further expand their snowmaking capabilities adjust

their lift infrastructure to be less dependent on lower terrain or

expand into higher terrain (Scott et al 2006) to counter climate

impacts The possibility that these ski areasmight have to cease

operating could seriously impact local economies and com-

munities through lost revenue and lost employment The ski

industry directly and indirectly (eg hotels restaurants retail

travel) is of great importance to the economies of many

mountain communities in the IWUS (Burakowski and

Magnusson 2012) These communities largely depend on the

employment tied to the skiing industry or the proximity to a

ski area for leisure purposes For these reasons closures of

ski areas could threaten the existence of whole mountain

communities as economic alternatives are limited (Steiger

et al 2019)

Snowmaking is critical in achieving a sufficient number of ski

days for most ski areas in this study If the production potential

can be completely utilized 44 ski areas still have over 120

median core-season ski days Furthermore ski areas that fall

below that threshold could increase snowmaking capabilities

TABLE 4 Average (by state) median core-season ski days

State(s) Historic climate Future climate

AZNM 123 95 (223)

CO 148 137 (27)

ID 133 101 (224)

MT 145 126 (213)

SDWY 135 118 (213)

UT 138 120 (213)

FIG 10 As in Fig 9 but for natural ski days



to achieve more ski days Many ski areas in the domain such as

Vail (Vail Mountain 2020) have recently made improvements

to their snowmaking facilities or are planning to make im-

provements in the near future However snowmaking is not

only limited by the production potential but also by costs and

water availability (Scott et al 2017) Thus ski areas may not be

able to fully develop their production potential since the costs

of producing enough snowmight be too high to reach economic

viability or water availability might be limited Water avail-

ability might be limited for different reasons For instance the

risk of severe droughts in the Southwest United States will

increase during the twenty-first century (eg Ault et al 2016)

limiting water availability physically Water availability for

snowmaking may also be limited legislatively given the tightly

restricted water use rights in the IWUS especially in the

Colorado River basin The Colorado River has to provide

water for irrigation and around 40 million people in the

southwestern United States (Udall and Overpeck 2017)

Snowmaking may also be restricted due to environmental

concerns (eg Baron et al 2000)

Studies in other regions have identified three behaviors of

tourists responding to marginal snow conditions or ski area

closures (eg Behringer et al 2000 Dawson and Scott 2010

Rutty et al 2015 Steiger et al 2019) The most common be-

havior is spatial substitution (skiing somewhere else) as

compared to temporal substitution (skiing at a different time)

or activity substitution (replacing skiing with a different ac-

tivity) This means that in the future many skiers might travel

to the ski areas that remain in operation increasing the number

of visitors there These ski areas could profit economically from

the increased demand but could also face challenges caused by

overcrowding (Steiger et al 2019) The increasing demand and

requirements for snowmaking might lead to an increase in lift

ticket prices at these locations making skiing affordable for

fewer people

Recently Scott et al (2019) investigated ski season length

changes in Ontario and Quebec Canada and the northeastern

United States They project that under theRCP85 midcentury

ski season lengths in these regions will be 15ndash22 shorter

compared to a baseline climate from 1981 to 2010 This is very

FIG 11Mapwith all ski areas showing the percentage of years having at least 8 ski days in the (top) Thanksgiving

period from 22 Nov to 1 Dec and (bottom) the Christmas period from 23 Dec to 1 Jan for the (a)(c) historic and

(b)(d) future climate periods The range in which the investigated elevation falls is indicated by the shape of the

symbols



similar to the 24and 23decrease inmedian core-season ski

days found for Idaho and ArizonaNew Mexico It can be in-

ferred that these states have a similar vulnerability to climate

change as the regions in Scott et al (2019) Colorado is less

vulnerable with a decrease of only 7 Wobus et al (2017)

investigated ski areas in the CONUS which includes ski areas

investigated in our study Under the same emission scenario

the authors predict a 19 decrease of mean ski season lengths

in Colorado and a 47 decrease in Idaho by 2050 These are

much higher decreases than the decreases we found for mean

core-season ski days in these two regions which are the same as

for median core-season ski days (7 and 24) A reason for

the difference might be their definition of a ski season their

season starts when either 10 cm of SWE or 450 h of snow-

making conditions (Tw 228C) are reached at the base ele-

vation of a ski area and ends when SWE falls below 10 cm at the

top elevation of a ski area While snowmaking plays a role in

initiating a season the lack of snowmaking during the season is

likely the reason why Wobus et al (2017) found much higher

decreases Because of that lack of representation of snow-

making they likely overestimate the decrease in ski season

length Furthermore it should be mentioned that they over-

estimate ski season length in general since their criterion for

the end of the season only accounts for snow at the top ele-

vations which ignores the possible lack of snow at lower

elevations

In the following paragraphs we list a few caveats of this

study First the parameter SWEAP (artificially provided SWE)

is defined the same for every ski area However this parameter

likely differs between ski areas since ski areas have different

snowmaking facilities and capabilities Therefore the area of

1000m2 in Eq (3) is a factor that could be defined individually

for every ski area To test the sensitivity to changes of this

parameter results were also examined for areas of 500 and

2000m2 Generally these changes produce more (500m2) or

fewer (2000m2) ski days For 500m2 the 100- and 120-day

thresholds are reached in 69 and 62 ski areas in the future cli-

mate This is not much different from the 70 and 61 ski areas

that reach the thresholds for 1000m2 in the historic climate

This highlights the importance and potential of snowmaking as

an adaption technique to mitigate the impacts of climate

change and to increase the number of ski days The choice fell

on 1000m2 since at 500m2 many ski days with no or little

natural snow occur The Tw can be close to the upper threshold

for snowmaking (228C) and enough snow for a ski day can be

produced For 2000m2 the opposite is trueTwmust be close to

the lower limit for snowmaking (2148C) for an extended pe-

riod and the number of ski days is not influenced much by the

addition of snowmaking Another simplification with snow-

making is that it is always applied This might not be in line with

typical operations at ski areas

Second the investigated elevationmaynot be the best choice for

all ski areas This elevationwas chosen as an approximation of the

mean elevation of the skiable terrain However suitable eleva-

tions for thismight differ between ski areasA similar point can be

made for the threshold of 200mm of SWE for a ski day since the

amount of snow required for skiing could differ between ski areas

depending on the terrain and underlying soil cover

For the above two reasons the results of this study should

not be understood to accurately display the vulnerability of

each individual ski area with their own specifications for

snowmaking snow requirements and elevation Rather this

study provides an overview of how vulnerable ski areas in the

IWUS are in general with predefined specifications depending

on their location Thus for some ski areas the presented

numbers of ski days might differ substantially from real-world

values depending especially on the ski arearsquos snowmaking

capabilities

Third the snowmaking capabilities are assumed to remain

unchanged This neglects that there might be advances in

snowmaking technology in the future Also snowmaking was

not as widespread in the historic climate (1981ndash2011) Thus the

historic climate snow indicators describe the possibilities under

these climate conditions with current snowmaking technology

While the wet-bulb temperature threshold of 228C is a basic

physical condition advances in snowmaking technology are

possible but it is not possible to estimate to which extent

Fourth the future climate simulation itself has uncertainties

The PGW technique has been used before to investigate

changes in orographic precipitation and snowpack in the

IWUS region in a warming climate (eg Rasmussen et al 2011

2014 Eidhammer et al 2018) The PGW technique assumes

that essential weather patterns (such as the midlatitude storm

track) and low-frequency global atmospheric variability do not

change (Schaumlr et al 1996) While this assumption remains

uncertain there is evidence that thermodynamic changes (ie

the warming and moistening of winter storms) in a globally

warming climate overwhelm sustained changes driven by in-

ternal climate variability (Scalzitti et al 2016) Furthermore

only one RCP (RCP85) is examined in this study referring to

conditions in the mid-twenty-first century or later (if green-

house gas emissions are cut more drastically) The RCP85

scenario implies relatively little action on reducing greenhouse

gas emissions compared to other RCP scenarios This sce-

nario often referred to as the lsquolsquobusiness as usualrsquorsquo scenario in

terms of greenhouse gas emissions may prove to exaggerate

the rate of global warming (Hausfather and Peters 2020) in

which case the lsquolsquofuturersquorsquo climate conditions depicted here may

apply not around 2050 but rather a few decades later Those

interested in a detailed assessment of snow conditions specifi-

cally around 2050 for example for infrastructure planning

purposes are encouraged to compare results with those from

other RCPs such as RCP45 for circa 2050

Fifth while the regional climate model used here estimates

the seasonal snowfall quite well (Jing et al 2017) the seasonal

SWE tends to be underestimated possibly on account of the

land surface scheme (Wang et al 2018) We corrected this

SWE bias (section 2d) but this neglects possible measurement

errors and biases in the SNOTEL data themselves and the

different validations of the simulation depending on the lo-

cation For instance in New Mexico the R2 values were lower

than in most other areas Consequently the snow indicators

have a larger uncertainty in NewMexico ski areas and other ski

areas with equally low R2

Sixth the artificial snow is not incorporated in the simula-

tionrsquos surface model More detailed studies could explicitly



model the interactions between natural and artificial snow

making the definition of artificially provided SWE redundant

However this is beyond the scope of this study

5 Conclusions

The results of this study are in line with findings from pre-

vious studies on the vulnerability of ski areas to climate change

in the sense that snow conditions for skiing generally are

negatively impacted by climate change This study examines 71

ski areas in the interior western United States and finds that

these impacts are larger at relatively low elevationslatitudes

This applies to all parameters investigated herein

d Natural SWE decreases significantly at low elevations and in

the early and late season at most locations At high eleva-

tions (over 3000m) in Colorado Montana and Wyoming

decreases in SWE are not significant in the midseason

(mainly JanuaryndashMarch) and at a few places there is a ro-

bust signal of midseason SWE increase in a warmer climate

At low latitudes in Arizona New Mexico and Utah the

decrease can be significant despite high elevations Averaged

over all ski areas SWE decreases by 118 at the top eleva-

tions and by 258 at the base elevation in mid-February

(midseason) in the future climate In late December (early

season) the decreases are 146 and 262 and in early April

(late season) they are 315 and 318d Similar findings apply to the number of warm days in the

core season which increase more at lower elevations and

latitudes In high-elevation ski areas in Colorado the in-

crease can be as low as 3 days For ski areas with low lati-

tudes or low elevations the increase can be as high as

28 daysd The number of days with rain on snow increases especially at

low elevations in the northwest of the domain In Idaho and

Montana ski areas the number of these days increases on

average by 8 days while in Arizona Colorado NewMexico

and high-elevation ski areas in Utah the average increase is

only 2 daysd Production potential for snowmaking decreases in every ski

area between 4 and 35 from October to December

when artificial snowmaking is especially important From

January to March decreases are lower and production

potential in this time frame could increase at the very highest

elevations in Colorado by up to 14 Again the decrease is

larger at lower elevationlatitude ski areasd Fewer ski areas reach the 100- and 120-day thresholds of

median core-season ski days in the future climate In the his-

toric climate 70 and 61 ski areas reach these thresholds as

compared to 51 and 44 in the future climate The ski areas that

do not reach the thresholds are at the low elevationslatitudes

of the domaind Skiing during the Thanksgiving period decreases substan-

tially throughout the domain In the future climate only 15

ski areas (a decrease of 33) have enough snow for skiing

during this period in at least 70 of the years 11 of those 15

ski areas are in Colorado Thus regular skiing during the

Thanksgiving period might be mostly limited to high eleva-

tions in Colorado

d The Christmas period is not impacted as much as the

Thanksgiving period Averaged over all ski areas the per-

centage of years with adequate snow for skiing during

Christmas decreases from 95 to 86 as compared to

77 to 50 for the Thanksgiving period At high elevations

in Colorado Montana northern Utah and Wyoming the

Christmas period is not affected Larger changes between the

climates are limited to lower elevations Overall Christmas

skiing might become more challenging in the future in some

years however it should still be possible in most of the

domain in most years

In summary the low-elevation ski areas in Idaho and north-

westernMontana and the low-latitude ski areas of Arizona and

New Mexico are most vulnerable Most ski areas in Colorado

and the high-elevation areas in Montana Utah and Wyoming

are less vulnerable

Acknowledgments Thanks are given to Corrine Knapp Jeff

Snider Holger Tost Thomas Mazzetti Coltin Grasmick

Robert Capella and Martin Espitalie for their comments and

insights during the course of this research project This work

was funded by theWyomingWater Development Commission

and the US Geological Survey under the auspices of the

University of Wyoming Water Research Program

Data availability statement The WRF Model data are avail-

able from httpsdoiorg105065D6MK6B4K The IWUSmodel

output for the retrospective climate is available from https

doiorg105281zenodo1157112 The future climate data are

available from httpsdoiorg105281zenodo3934896

REFERENCES

Abegg B S Morin O Demiroglu H Franccedilois M Rothleitner

and U Strasser 2021 Overloaded Critical revision and a new

conceptual approach for snow indicators in ski tourism Int

J Biometeor httpsdoiorg101007s00484-020-01867-3 in

press

Adger W N 2006 Vulnerability Global Environ Change 16

268ndash281 httpsdoiorg101016jgloenvcha200602006

Ashfaq M S Ghosh S-C Kao L C Bowling P Mote

D Touma S A Rauscher and N S Diffenbaugh 2013 Near-

term acceleration of hydroclimatic change in the western US

J Geophys Res Atmos 118 10 676ndash10 693 httpsdoiorg

101002jgrd50816

Ault T R J S Mankin B I Cook and J E Smerdon 2016

Relative impacts of mitigation temperature and precipita-

tion on 21st-century megadrought risk in the American

Southwest Sci Adv 2 e1600873 httpsdoiorg101126

sciadv1600873

Bark R B Colby and FDominguez 2010 Snowdays Snowmaking

adaptation and the future of low latitude high elevation skiing in

Arizona USA Climatic Change 102 467ndash491 httpsdoiorg

101007s10584-009-9708-x

Baron J D Theobald and D Fagre 2000 Management of land

use conflicts in the United States Rocky Mountains Mt Res

Dev 20 24ndash27 httpsdoiorg1016590276-4741(2000)020

[0024MOLUCI]20CO2

Battaglin W L Hay andM Steve 2011 Simulating the potential

effects of climate change in two Colorado basins and at two



Colorado ski areas Earth Interact 15 httpsdoiorg101175

2011EI3731

Beaudin L and J-C Huang 2014Weather conditions and outdoor

recreation A study of New England ski areas Ecol Econ 106

56ndash68 httpsdoiorg101016jecolecon201407011

Behringer J R Buumlrki and J Fuhrer 2000 Participatory inte-

grated assessment of adaptation to climate change in Alpine

tourism and mountain agriculture Integr Assess 1 331ndash338

httpsdoiorg101023A1018940901744

Brown R and P Mote 2009 The response of Northern Hemisphere

snow cover to a changing climate J Climate 22 2124ndash2145

httpsdoiorg1011752008JCLI26651

Burakowski E andMMagnusson 2012 Climate impacts on the

winter tourism economy in the United States Natural

Resources Defense Council Rep 32 pp httpswwwnrdcorg

sitesdefaultfilesclimate-impacts-winter-tourism-reportpdf

Dawson J and D Scott 2010 Systems analysis of climate

change vulnerability for the US Northeast ski sector

Tourism Hospitality Plann Dev 7 219ndash235 httpsdoiorg

1010801479053X2010502383

Diffenbaugh N M Scherer and M Ashfaq 2013 Response of

snow-dependent hydrologic extremes to continued global

warming Nat Climate Change 3 379ndash384 httpsdoiorg

101038nclimate1732

Durand Y G Giraud M Laternser P Etchevers L Meacuterindoland B Lesaffre 2009 Reanalysis of 47 years of climate in the

French Alps (1958ndash2005) Climatology and trends for snow

cover J Appl Meteor Climatol 48 2487ndash2512 https

doiorg1011752009JAMC18101

Eidhammer T V Grubisic R Rasmussen and K Ikdea 2018

Winter precipitation efficiency of mountain ranges in the

Colorado Rockies under climate change J Geophys Res

Atmos 123 2573ndash2590 httpsdoiorg1010022017JD027995

Gruumlnewald T and Coauthors 2013 Statistical modelling of the snow

depth distribution in openAlpine terrainHydrol Earth Syst Sci

17 3005ndash3021 httpsdoiorg105194hess-17-3005-2013

Hausfather Z and G Peters 2020 EmissionsmdashThe lsquobusiness as

usualrsquo story ismisleadingNature 577 618ndash620 httpsdoiorg

101038d41586-020-00177-3

Henn B A J Newman B Livneh C Daly and J D Lundquist

2018 An assessment of differences in gridded precipitation

datasets in complex terrain J Hydrol 556 1205ndash1219 https

doiorg101016jjhydrol201703008

Hennessy K P Whetton K Walsh I Smith J Bathols

M Hutchinson and J Sharples 2008 Climate change effects

on snow conditions in mainland Australia and adaptation at

ski resorts through snowmaking Climate Res 35 255ndash270

httpsdoiorg103354cr00706

Hong S-Y and H-L Pan 1996 Nonlocal boundary layer vertical

diffusion in amedium-range forecastmodelMonWeaRev 124

2322ndash2339 httpsdoiorg1011751520-0493(1996)1242322

NBLVDI20CO2

Iacono M J J S Delamere E J Mlawer M W Shephard S A

Clough and W D Collins 2008 Radiative forcing by long-

lived greenhouse gases Calculations with the AER radiative

transfermodels J Geophys Res 113 D13103 httpsdoiorg

1010292008JD009944

Jimeacutenez P A J Dudhia J F Gonzaacutelez-Rouco J Navarro J P

Montaacutevez andEGarciacutea-Bustamante 2012 A revised scheme

for the WRF surface layer formulation Mon Wea Rev 140

898ndash918 httpsdoiorg101175MWR-D-11-000561

Jing X B Geerts Y Wang and C Liu 2017 Evaluating seasonal

orographic precipitation in the interior western United States

using gauge data gridded precipitation estimates and a re-

gional climate simulation J Hydrometeor 18 2541ndash2558

httpsdoiorg101175JHM-D-17-00561

Koenig U andBAbegg 1997 Impacts of climate change onwinter

tourism in the Swiss Alps J Sustainable Tourism 5 46ndash58

httpsdoiorg10108009669589708667275

Lehning M T Gruumlnewald and M Schirmer 2011 Mountain

snow distribution governed by an altitudinal gradient and

terrain roughness Geophys Res Lett 38 L19504 https

doiorg1010292011GL048927

Li Y Z Li Z Zhang L Chen S Kurkute L Scaff andX Pan 2019

High-resolution regional climate modeling and projection over

western Canada using a weather research forecasting model

with a pseudo-global warming approachHydrol Earth Syst Sci

23 4635ndash4659 httpsdoiorg105194hess-23-4635-2019Liu C and Coauthors 2017 Continental-scale convection-

permitting modeling of the current and future climate of

North America Climate Dyn 49 71ndash95 httpsdoiorg

101007s00382-016-3327-9

Lundquist J M Hughes E Gutmann and S Kapnick 2019 Our

skill in modeling mountain rain and snow is bypassing the skill

of our observational networks Bull Amer Meteor Soc 100

2473ndash2490 httpsdoiorg101175BAMS-D-19-00011

Maraun D 2016 Bias correcting climate change simulationsmdashA

critical review Curr Climate Change Rep 2 211ndash220 https

doiorg101007s40641-016-0050-x

Newman A AMonaghanM Clark K Ikeda L Xue E Gutmann

and J Arnold 2021 Hydroclimatic changes in Alaska portrayed

by a high-resolution regional climate simulation Climatic

Change 164 17 httpsdoiorg101007s10584-021-02956-x

Niu G-Y and Coauthors 2011 The community Noah land sur-

face model withmultiparameterization options (Noah-MP) 1

Model description and evaluation with local-scale measure-

ments J Geophys Res 116 D12109 httpsdoiorg101029

2010JD015139

NRCS 2020 NRCS Report Generator 20 USDA accessed

11 April 2020 httpswccscegovusdagovreportGenerator

NSAA 2021 Kottke national end of season survey 201920 Final

report National Ski Areas Association Rep (Appendix B

Skier visit detail by region Table 10) 1 p httpsnsaaorg

webdocsMedia_PublicIndustryStatsHistorical_Skier_Days_

1979_1920pdf

Olefs M A Fischer and J Lang 2010 Boundary conditions

for artificial snow production in the Austrian Alps J Appl

Meteor Climatol 49 1096ndash1113 httpsdoiorg101175

2010JAMC22511

Pachauri R K and Coauthors 2014Climate Change 2014 Synthesis

Report CambridgeUniversity Press 151 pp httpswwwipccch

siteassetsuploads201802SYR_AR5_FINAL_fullpdf

Pons M J Loacutepez-Moreno M Rosas-Casals and E Jover 2015

The vulnerability of Pyrenean ski resorts to climate-induced

changes in the snowpack Climatic Change 131 591ndash605

httpsdoiorg101007s10584-015-1400-8

Rasmussen R and Coauthors 2011 High-resolution coupled

climate runoff simulations of seasonal snowfall over Colorado

A process study of current and warmer climate J Climate 24

3015ndash3048 httpsdoiorg1011752010JCLI39851

mdashmdash and Coauthors 2014 Climate change impacts on the water

balance of the Colorado Headwaters High-resolution regional

climate model simulations J Hydrometeor 15 1091ndash1116


Rhoades A P Ullrich and C Zarzycki 2018 Projecting 21st

century snowpack trends in western USA mountains using



variable-resolution CESM Climate Dyn 50 261ndash288 https

doiorg101007s00382-017-3606-0

Rutty M D Scott P Johnson E Jover M Pons and R Steiger

2015 Behavioural adaptation of skiers to climatic variability

and change in Ontario Canada J Outdoor Recreat Tourism

11 13ndash21 httpsdoiorg101016jjort201507002

Sadeghi S-H R Peters D Cobos H Loescher and C Campbell

2013 Direct calculation of thermodynamic wet-bulb temper-

ature as a function of pressure and elevation J Atmos

Oceanic Technol 30 1757ndash1765 httpsdoiorg101175

JTECH-D-12-001911

Saha S andCoauthors 2010 TheNCEPClimate Forecast System

Reanalysis Bull Amer Meteor Soc 91 1015ndash1058 https

doiorg1011752010BAMS30011

Scalzitti J C Strong and A Kochanski 2016 Climate change

impact on the roles of temperature and precipitation in

western US snowpack variability Geophys Res Lett 43

5361ndash5369 httpsdoiorg1010022016GL068798

Schaumlr C C Frei D Luumlthi and H Davies 1996 Surrogate

climate-change scenarios for regional climate models

Geophys Res Lett 23 669ndash672 httpsdoiorg101029

96GL00265

Scott D G McBoyle and B Mills 2003 Climate change and the

skiing industry in southern Ontario (Canada) Exploring the

importance of snowmaking as a technical adaptation Climate

Res 23 171ndash181 httpsdoiorg103354cr023171mdashmdashmdashmdashAMinogue and BMills 2006 Climate change and the

sustainability of ski-based tourism in eastern North America

A reassessment J Sustainable Tourism 14 376ndash398 https

doiorg102167jost5500

mdashmdash J Dawson and B Jones 2007 Climate change vulnerability

of the US Northeast winter recreationndashtourism sector Mitig

Adapt Strategies Global Change 13 577ndash596 httpsdoiorg

101007s11027-007-9136-z

mdashmdash R Steiger M Rutty M Pons and P Johnson 2017 The

differential futures of ski tourism in Ontario (Canada)

under climate change The limits of snowmaking adapta-

tion Curr Issues Tourism 22 1327ndash1342 httpsdoiorg

1010801368350020171401984

mdashmdash mdashmdash N Knowles and Y Fang 2019 Regional ski tourism

risk to climate change An inter-comparison of eastern

Canada andUSNortheastmarkets J Sustainable Tourism 28

568ndash586 httpsdoiorg1010800966958220191684932

Serreze M C M P Clark R L Armstrong D A McGinnis and

R S Pulwarty 1999 Characteristics of the western United

States snowpack from Snowpack Telemetry (SNOTEL) data

Water Resour Res 35 2145ndash2160 httpsdoiorg101029

1999WR900090

Skamarock W C and Coauthors 2019 A description of

the Advanced Research WRF version 4 NCAR Tech

Rep NCARTN-5561STR 145 pp httpsdoiorg105065

1dfh-6p97

Steiger R 2010 The impact of climate change on ski season length

and snowmaking requirements in Tyrol AustriaClimate Res

43 251ndash262 httpsdoiorg103354cr00941

mdashmdashD Scott B AbeggM Pons and C Aall 2019 A critical review

of climate change risk for ski tourism Curr Issues Tourism 22


Sturm M M A Goldstein and C Parr 2017 Water and life from

snow A trillion dollar science question Water Resour Res

53 3534ndash3544 httpsdoiorg1010022017WR020840

Thompson G P R Field RMRasmussen andWDHall 2008

Explicit forecasts of winter precipitation using an improved

bulk microphysics scheme Part II Implementation of a new

snow parameterization Mon Wea Rev 136 5095ndash5115

httpsdoiorg1011752008MWR23871

Toumlglhofer C F Eigner and F Prettenthaler 2011 Impacts of

snow conditions on tourism demand in Austrian ski areas

Climate Res 46 (1) 1ndash14 httpsdoiorg103354cr00939

Udall B and J Overpeck 2017 The twenty-first century Colorado

River hot drought and implications for the futureWaterResour

Res 53 2404ndash2418 httpsdoiorg1010022016WR019638

Vail Mountain 2020 Vail Mountain Snow Enhancement Project

Accessed 19 June 2020 httpswwwvailcomexplore

snowmakingaspx

Wang Y B Geerts and C Liu 2018 A 30-year convection-

permitting regional climate simulation over the interior western

United States Part I Validation Int J Climatol 38 3684ndash3704

httpsdoiorg101002joc5527

Wobus C and Coauthors 2017 Projected climate change impacts

on skiing and snowmobiling A case study of theUnited States

Global Environ Change 45 1ndash14 httpsdoiorg101016

jgloenvcha201704006

Yang Z-L and Coauthors 2011 The community Noah land

surface model with multiparameterization options (Noah-

MP) 2 Evaluation over global river basins J Geophys

Res 116 D12110 httpsdoiorg1010292010JD015140

Zimmerman G C OrsquoBrady and B Hurlbutt 2006 Climate change

Modeling a warmer Rockies and assessing the implications The

2006 Colorado College State of the Rockies Report Card

Colorado College Publ 89ndash102 httpswwwcoloradocollege

edudotAsseta68bd37f-4ca0-472a-bb73-69ab7e4941aepdf



In the winter of 201819 over 41 of the 593 million ski visits in

the United States were in the Rocky Mountain region (NSAA

2021) making it the most visited skiing region in the country

The Rocky Mountains have diverse climate conditions varying

orography and few studies have looked at ski areas in this region

(eg Zimmerman et al 2006 Bark et al 2010 Battaglin et al

2011Wobus et al 2017) Thus combinedwith the importance of

the skiing industry in the region it is of interest to assess the

vulnerability of the local skiing industry to climate change This

study focuses on vulnerability in the sense of exposure to envi-

ronmental stress caused by climate change and adaptive capacity

to mitigate this stress (Adger 2006) For instance exposure to

environmental stress for ski areas could be decreasing natural

snow amounts or rising temperatures negatively impacting ski-

ing operations in the futureArtificial snowmaking is an adaptive

capacity ski areas have to cope with this exposure A regional

climate simulation over the IWUS conducted with the Weather

Research and Forecasting (WRF)Model is used in this study In

the analysis of skiing conditions we examine natural snowfall

snowmaking potential excessive temperature and rain

Additionally so-called snow indicators as suggested by

Abegg et al (2021) will be used for this analysis

Section 2 will describe the methods of this study including

the regional climate simulation determination validation and

correction of meteorological parameters in the ski areas and

definitions of snow indicators Section 3 will present the results

from the analysis of natural snowfall thresholds for tempera-

ture and rain changes in production potential for artificial

snow and the snow indicators In section 4 implications and

limitations of this study are discussed and choices of certain

parameters are justified The main findings will be summarized

in section 5

2 Methods

a WRF regional climate simulation over the IWUS

This study uses a high-resolution (4km) convection-permitting

regional climate simulation over the IWUS conducted with the

WRF Model (Skamarock et al 2019) For details on the model

configuration see Table 1 More details can be found in Wang

et al (2018) The 4-km resolution and similar WRF physics

choices were used inRasmussen et al (2011 2014) who show that

such setup captures the cold-season precipitation distribution and

amount over the Colorado Headwaters region well with a bias

of 10ndash15 compared to Snowpack Telemetry (SNOTEL)

measurements Because of this good performance in simulating

orographic precipitation over complex terrain similar WRF

simulations have been used to assess changes in orographic pre-

cipitation in a changing global climate For instance Li et al

(2019) and Newman et al (2021) both use convection-permitting

4-km WRF simulations to explore the sensitivity of precipitation

and snowpack to climate change in western Canada and Alaska

respectively Liu et al (2017) extended a 4-kmWRF simulation to

cover the entire contiguous United States although over fewer

years than our simulation The surface temperature and pre-

cipitation for the simulation used in our study have been vali-

dated byWang et al (2018) The cold-season precipitation over

the mountains in the IWUS was validated by Jing et al (2017)

Comparing 10 years of model data against SNOTEL data

across the IWUS Jing et al (2017) find a correlation coefficient

of 095 a mean bias of213mm and root-mean-square bias of

65mm of water-equivalent precipitation in the winter (DJF)

They conclude that simulated seasonal precipitation over

mountains can be more accurate than the numerous gridded

gauge-based precipitation datasets in existence [seeHenn et al

(2018) for a discussion of these datasets] a statement con-

firmed in a broader study by Lundquist et al (2019) Thus

these high-resolution regional climate models of the recent

past are almost equivalent to reanalysis data but of sufficient

resolution to capture finescale orographic snowfall patterns

The model output is used to investigate the skiing condi-

tions both in the recent past and the near future The retro-

spective simulation spans the period from November 1981 to

October 2011 This retrospective climate will be referred to as

historic climate To examine the same patterns in the near-

future climate YWang et al (2020 unpublished manuscript)

used a pseudondashglobal warming (PGW) approach with the

driver dataset perturbed by the CMIP5 ensemble mean

model guidance for 2050 under the representative concen-

tration pathway (RCP) 85 (Pachauri et al 2014) The basic

idea of the PGW approach is to apply the guidance from

global climate models (in this case the ensemble mean CMIP5

climate change signal) to the driver dataset of the regional

climate model (Schaumlr et al 1996) This approach has been

used widely including in studies of changes in orographic

precipitation in the Colorado Headwaters (Rasmussen et al

2011 2014 Eidhammer et al 2018) Thus the historic and

future climate conditions correspond to 1990s and 2050s

conditions The use of 30 years of simulations builds statistical

significance in terms of the mean and the spread of snow

years at any ski resort

Relevant for this study are the hourly model output of sur-

face air temperature rain and snow water equivalent (SWE)

Modeled SWE was chosen over modeled snow depth since it

more accurately describes the total amount of available snow

Furthermore wet-bulb temperature is needed to incorporate

production potential for artificial snow (Olefs et al 2010)Wet-

bulb temperature is calculated using air temperature and rel-

ative humidity (Sadeghi et al 2013) The relative humidity is

TABLE 1 Model configurations

Model parameter Model configuration

Spatial resolution 4 km 3 4 km (420 3 410 grid points)

Vertical levels 51 (topped at 50 hPa)

Driver dataset NCEP Climate Forecast System

Reanalysis (Saha et al 2010)

Microphysics Thompson scheme (Thompson et al 2008)

Radiation Rapid Radiative Transfer Model for

GCMs (Iacono et al 2008)

Planetary

boundary layer

Yonsei University scheme (Hong and

Pan 1996)

Surface layer Revised MoninndashObukhov scheme

(Jimeacutenez et al 2012)Land surface Noah-MP scheme (Niu et al 2011 Yang

et al 2011)




vapor mixing ratio



























vations below 2000m


















(NRCS 2020)





















CO Crested Butte 28574 37068 192 172







CO Ski Cooper 32002 35660 205 179









ID Kelly Canyon 17068 20116 107 63

ID Lookout Pass 13715 17220 176 128

ID Pebble Creek 19384 28257 134 92

ID Pomerelle 23651 26705 160 120




ID Sun Valley 17525 27888 135 105

ID Tamarack 14934 23468 151 97

MT Big Sky 20725 34032 192 169



MT Discovery 20878 24840 193 167

MT Great Divide 17464 22045 145 94

MT Lost Trail 19506 24992 181 155




MT Showdown 20725 24992 187 162





NM Sipapu 24992 28208 98 58

NM Ski Santa Fe 31545 36803 171 136


SD Terry Peak 17982 21640 127 93

UT Alta 25998 33734 214 184





UT Deer Valley 20024 29168 139 99
























the box






































obtained

SWEWRFcorr

5 SWEWRF

3SWE

SNOTELbin

SWEWRFbin

(1)

















TABLE 2 (Continued)






UT Snowbird 23651 33526 202 171



WY Grand Targhee 22578 30058 213 184

WY Jackson Hole 19235 31850 179 154

WY Pine Creek 20802 25069 122 91



WY Snowy Range 27431 30479 183 152

WY White Pine 25754 28955 184 162














e Artificial snow
















pp5 09(24385Tw2 0145) (2)












SWEAPd

5dppd27d21

3400 kgm23

1000m2(3)

















f Snow indicators


















































quality

3 Results

a Natural snow

























climate










of ski days


of ski days


period of ski days


15 Apr)




15 Sep

Snow years

Thanksgiving

period




period

















































































































fected there























c Snow indicators





















different periods









































































































































4 Discussion
























et al 2019)








AZNM 123 95 (223)

CO 148 137 (27)

ID 133 101 (224)

MT 145 126 (213)

SDWY 135 118 (213)

UT 138 120 (213)






































fewer people









symbols




























elevations










































capabilities






















































5 Conclusions















































tions in Colorado
















are less vulnerable













REFERENCES





press







101002jgrd50816





sciadv1600873




101007s10584-009-9708-x




[0024MOLUCI]20CO2






2011EI3731







httpsdoiorg101023A1018940901744











1010801479053X2010502383




101038nclimate1732














101038d41586-020-00177-3













NBLVDI20CO2





1010292008JD009944












httpsdoiorg10108009669589708667275




doiorg1010292011GL048927








101007s00382-016-3327-9







doiorg101007s40641-016-0050-x









2010JD015139







1979_1920pdf




2010JAMC22511







httpsdoiorg101007s10584-015-1400-8














doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006













vapor mixing ratio



























vations below 2000m


















(NRCS 2020)





















CO Crested Butte 28574 37068 192 172







CO Ski Cooper 32002 35660 205 179









ID Kelly Canyon 17068 20116 107 63

ID Lookout Pass 13715 17220 176 128

ID Pebble Creek 19384 28257 134 92

ID Pomerelle 23651 26705 160 120




ID Sun Valley 17525 27888 135 105

ID Tamarack 14934 23468 151 97

MT Big Sky 20725 34032 192 169



MT Discovery 20878 24840 193 167

MT Great Divide 17464 22045 145 94

MT Lost Trail 19506 24992 181 155




MT Showdown 20725 24992 187 162





NM Sipapu 24992 28208 98 58

NM Ski Santa Fe 31545 36803 171 136


SD Terry Peak 17982 21640 127 93

UT Alta 25998 33734 214 184





UT Deer Valley 20024 29168 139 99
























the box






































obtained

SWEWRFcorr

5 SWEWRF

3SWE

SNOTELbin

SWEWRFbin

(1)

















TABLE 2 (Continued)






UT Snowbird 23651 33526 202 171



WY Grand Targhee 22578 30058 213 184

WY Jackson Hole 19235 31850 179 154

WY Pine Creek 20802 25069 122 91



WY Snowy Range 27431 30479 183 152

WY White Pine 25754 28955 184 162














e Artificial snow
















pp5 09(24385Tw2 0145) (2)












SWEAPd

5dppd27d21

3400 kgm23

1000m2(3)

















f Snow indicators


















































quality

3 Results

a Natural snow

























climate










of ski days


of ski days


period of ski days


15 Apr)




15 Sep

Snow years

Thanksgiving

period




period

















































































































fected there























c Snow indicators





















different periods









































































































































4 Discussion
























et al 2019)








AZNM 123 95 (223)

CO 148 137 (27)

ID 133 101 (224)

MT 145 126 (213)

SDWY 135 118 (213)

UT 138 120 (213)






































fewer people









symbols




























elevations










































capabilities






















































5 Conclusions















































tions in Colorado
















are less vulnerable













REFERENCES





press







101002jgrd50816





sciadv1600873




101007s10584-009-9708-x




[0024MOLUCI]20CO2






2011EI3731







httpsdoiorg101023A1018940901744











1010801479053X2010502383




101038nclimate1732














101038d41586-020-00177-3













NBLVDI20CO2





1010292008JD009944












httpsdoiorg10108009669589708667275




doiorg1010292011GL048927








101007s00382-016-3327-9







doiorg101007s40641-016-0050-x









2010JD015139







1979_1920pdf




2010JAMC22511







httpsdoiorg101007s10584-015-1400-8














doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006























CO Crested Butte 28574 37068 192 172







CO Ski Cooper 32002 35660 205 179









ID Kelly Canyon 17068 20116 107 63

ID Lookout Pass 13715 17220 176 128

ID Pebble Creek 19384 28257 134 92

ID Pomerelle 23651 26705 160 120




ID Sun Valley 17525 27888 135 105

ID Tamarack 14934 23468 151 97

MT Big Sky 20725 34032 192 169



MT Discovery 20878 24840 193 167

MT Great Divide 17464 22045 145 94

MT Lost Trail 19506 24992 181 155




MT Showdown 20725 24992 187 162





NM Sipapu 24992 28208 98 58

NM Ski Santa Fe 31545 36803 171 136


SD Terry Peak 17982 21640 127 93

UT Alta 25998 33734 214 184





UT Deer Valley 20024 29168 139 99
























the box






































obtained

SWEWRFcorr

5 SWEWRF

3SWE

SNOTELbin

SWEWRFbin

(1)

















TABLE 2 (Continued)






UT Snowbird 23651 33526 202 171



WY Grand Targhee 22578 30058 213 184

WY Jackson Hole 19235 31850 179 154

WY Pine Creek 20802 25069 122 91



WY Snowy Range 27431 30479 183 152

WY White Pine 25754 28955 184 162














e Artificial snow
















pp5 09(24385Tw2 0145) (2)












SWEAPd

5dppd27d21

3400 kgm23

1000m2(3)

















f Snow indicators


















































quality

3 Results

a Natural snow

























climate










of ski days


of ski days


period of ski days


15 Apr)




15 Sep

Snow years

Thanksgiving

period




period

















































































































fected there























c Snow indicators





















different periods









































































































































4 Discussion
























et al 2019)








AZNM 123 95 (223)

CO 148 137 (27)

ID 133 101 (224)

MT 145 126 (213)

SDWY 135 118 (213)

UT 138 120 (213)






































fewer people









symbols




























elevations










































capabilities






















































5 Conclusions















































tions in Colorado
















are less vulnerable













REFERENCES





press







101002jgrd50816





sciadv1600873




101007s10584-009-9708-x




[0024MOLUCI]20CO2






2011EI3731







httpsdoiorg101023A1018940901744











1010801479053X2010502383




101038nclimate1732














101038d41586-020-00177-3













NBLVDI20CO2





1010292008JD009944












httpsdoiorg10108009669589708667275




doiorg1010292011GL048927








101007s00382-016-3327-9







doiorg101007s40641-016-0050-x









2010JD015139







1979_1920pdf




2010JAMC22511







httpsdoiorg101007s10584-015-1400-8














doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006
































the box






































obtained

SWEWRFcorr

5 SWEWRF

3SWE

SNOTELbin

SWEWRFbin

(1)

















TABLE 2 (Continued)






UT Snowbird 23651 33526 202 171



WY Grand Targhee 22578 30058 213 184

WY Jackson Hole 19235 31850 179 154

WY Pine Creek 20802 25069 122 91



WY Snowy Range 27431 30479 183 152

WY White Pine 25754 28955 184 162














e Artificial snow
















pp5 09(24385Tw2 0145) (2)












SWEAPd

5dppd27d21

3400 kgm23

1000m2(3)

















f Snow indicators


















































quality

3 Results

a Natural snow

























climate










of ski days


of ski days


period of ski days


15 Apr)




15 Sep

Snow years

Thanksgiving

period




period

















































































































fected there























c Snow indicators





















different periods









































































































































4 Discussion
























et al 2019)








AZNM 123 95 (223)

CO 148 137 (27)

ID 133 101 (224)

MT 145 126 (213)

SDWY 135 118 (213)

UT 138 120 (213)






































fewer people









symbols




























elevations










































capabilities






















































5 Conclusions















































tions in Colorado
















are less vulnerable













REFERENCES





press







101002jgrd50816





sciadv1600873




101007s10584-009-9708-x




[0024MOLUCI]20CO2






2011EI3731







httpsdoiorg101023A1018940901744











1010801479053X2010502383




101038nclimate1732














101038d41586-020-00177-3













NBLVDI20CO2





1010292008JD009944












httpsdoiorg10108009669589708667275




doiorg1010292011GL048927








101007s00382-016-3327-9







doiorg101007s40641-016-0050-x









2010JD015139







1979_1920pdf




2010JAMC22511







httpsdoiorg101007s10584-015-1400-8














doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006























e Artificial snow
















pp5 09(24385Tw2 0145) (2)












SWEAPd

5dppd27d21

3400 kgm23

1000m2(3)

















f Snow indicators


















































quality

3 Results

a Natural snow

























climate










of ski days


of ski days


period of ski days


15 Apr)




15 Sep

Snow years

Thanksgiving

period




period

















































































































fected there























c Snow indicators





















different periods









































































































































4 Discussion
























et al 2019)








AZNM 123 95 (223)

CO 148 137 (27)

ID 133 101 (224)

MT 145 126 (213)

SDWY 135 118 (213)

UT 138 120 (213)






































fewer people









symbols




























elevations










































capabilities






















































5 Conclusions















































tions in Colorado
















are less vulnerable













REFERENCES





press







101002jgrd50816





sciadv1600873




101007s10584-009-9708-x




[0024MOLUCI]20CO2






2011EI3731







httpsdoiorg101023A1018940901744











1010801479053X2010502383




101038nclimate1732














101038d41586-020-00177-3













NBLVDI20CO2





1010292008JD009944












httpsdoiorg10108009669589708667275




doiorg1010292011GL048927








101007s00382-016-3327-9







doiorg101007s40641-016-0050-x









2010JD015139







1979_1920pdf




2010JAMC22511







httpsdoiorg101007s10584-015-1400-8














doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006



















f Snow indicators


















































quality

3 Results

a Natural snow

























climate










of ski days


of ski days


period of ski days


15 Apr)




15 Sep

Snow years

Thanksgiving

period




period

















































































































fected there























c Snow indicators





















different periods









































































































































4 Discussion
























et al 2019)








AZNM 123 95 (223)

CO 148 137 (27)

ID 133 101 (224)

MT 145 126 (213)

SDWY 135 118 (213)

UT 138 120 (213)






































fewer people









symbols




























elevations










































capabilities






















































5 Conclusions















































tions in Colorado
















are less vulnerable













REFERENCES





press







101002jgrd50816





sciadv1600873




101007s10584-009-9708-x




[0024MOLUCI]20CO2






2011EI3731







httpsdoiorg101023A1018940901744











1010801479053X2010502383




101038nclimate1732














101038d41586-020-00177-3













NBLVDI20CO2





1010292008JD009944












httpsdoiorg10108009669589708667275




doiorg1010292011GL048927








101007s00382-016-3327-9







doiorg101007s40641-016-0050-x









2010JD015139







1979_1920pdf




2010JAMC22511







httpsdoiorg101007s10584-015-1400-8














doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006
























































































































fected there























c Snow indicators





















different periods









































































































































4 Discussion
























et al 2019)








AZNM 123 95 (223)

CO 148 137 (27)

ID 133 101 (224)

MT 145 126 (213)

SDWY 135 118 (213)

UT 138 120 (213)






































fewer people









symbols




























elevations










































capabilities






















































5 Conclusions















































tions in Colorado
















are less vulnerable













REFERENCES





press







101002jgrd50816





sciadv1600873




101007s10584-009-9708-x




[0024MOLUCI]20CO2






2011EI3731







httpsdoiorg101023A1018940901744











1010801479053X2010502383




101038nclimate1732














101038d41586-020-00177-3













NBLVDI20CO2





1010292008JD009944












httpsdoiorg10108009669589708667275




doiorg1010292011GL048927








101007s00382-016-3327-9







doiorg101007s40641-016-0050-x









2010JD015139







1979_1920pdf




2010JAMC22511







httpsdoiorg101007s10584-015-1400-8














doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006



















































































fected there























c Snow indicators





















different periods









































































































































4 Discussion
























et al 2019)








AZNM 123 95 (223)

CO 148 137 (27)

ID 133 101 (224)

MT 145 126 (213)

SDWY 135 118 (213)

UT 138 120 (213)






































fewer people









symbols




























elevations










































capabilities






















































5 Conclusions















































tions in Colorado
















are less vulnerable













REFERENCES





press







101002jgrd50816





sciadv1600873




101007s10584-009-9708-x




[0024MOLUCI]20CO2






2011EI3731







httpsdoiorg101023A1018940901744











1010801479053X2010502383




101038nclimate1732














101038d41586-020-00177-3













NBLVDI20CO2





1010292008JD009944












httpsdoiorg10108009669589708667275




doiorg1010292011GL048927








101007s00382-016-3327-9







doiorg101007s40641-016-0050-x









2010JD015139







1979_1920pdf




2010JAMC22511







httpsdoiorg101007s10584-015-1400-8














doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006













fected there























c Snow indicators





















different periods









































































































































4 Discussion
























et al 2019)








AZNM 123 95 (223)

CO 148 137 (27)

ID 133 101 (224)

MT 145 126 (213)

SDWY 135 118 (213)

UT 138 120 (213)






































fewer people









symbols




























elevations










































capabilities






















































5 Conclusions















































tions in Colorado
















are less vulnerable













REFERENCES





press







101002jgrd50816





sciadv1600873




101007s10584-009-9708-x




[0024MOLUCI]20CO2






2011EI3731







httpsdoiorg101023A1018940901744











1010801479053X2010502383




101038nclimate1732














101038d41586-020-00177-3













NBLVDI20CO2





1010292008JD009944












httpsdoiorg10108009669589708667275




doiorg1010292011GL048927








101007s00382-016-3327-9







doiorg101007s40641-016-0050-x









2010JD015139







1979_1920pdf




2010JAMC22511







httpsdoiorg101007s10584-015-1400-8














doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006


















































































































































4 Discussion
























et al 2019)








AZNM 123 95 (223)

CO 148 137 (27)

ID 133 101 (224)

MT 145 126 (213)

SDWY 135 118 (213)

UT 138 120 (213)






































fewer people









symbols




























elevations










































capabilities






















































5 Conclusions















































tions in Colorado
















are less vulnerable













REFERENCES





press







101002jgrd50816





sciadv1600873




101007s10584-009-9708-x




[0024MOLUCI]20CO2






2011EI3731







httpsdoiorg101023A1018940901744











1010801479053X2010502383




101038nclimate1732














101038d41586-020-00177-3













NBLVDI20CO2





1010292008JD009944












httpsdoiorg10108009669589708667275




doiorg1010292011GL048927








101007s00382-016-3327-9







doiorg101007s40641-016-0050-x









2010JD015139







1979_1920pdf




2010JAMC22511







httpsdoiorg101007s10584-015-1400-8














doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006































































4 Discussion
























et al 2019)








AZNM 123 95 (223)

CO 148 137 (27)

ID 133 101 (224)

MT 145 126 (213)

SDWY 135 118 (213)

UT 138 120 (213)






































fewer people









symbols




























elevations










































capabilities






















































5 Conclusions















































tions in Colorado
















are less vulnerable













REFERENCES





press







101002jgrd50816





sciadv1600873




101007s10584-009-9708-x




[0024MOLUCI]20CO2






2011EI3731







httpsdoiorg101023A1018940901744











1010801479053X2010502383




101038nclimate1732














101038d41586-020-00177-3













NBLVDI20CO2





1010292008JD009944












httpsdoiorg10108009669589708667275




doiorg1010292011GL048927








101007s00382-016-3327-9







doiorg101007s40641-016-0050-x









2010JD015139







1979_1920pdf




2010JAMC22511







httpsdoiorg101007s10584-015-1400-8














doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006















4 Discussion
























et al 2019)








AZNM 123 95 (223)

CO 148 137 (27)

ID 133 101 (224)

MT 145 126 (213)

SDWY 135 118 (213)

UT 138 120 (213)






































fewer people









symbols




























elevations










































capabilities






















































5 Conclusions















































tions in Colorado
















are less vulnerable













REFERENCES





press







101002jgrd50816





sciadv1600873




101007s10584-009-9708-x




[0024MOLUCI]20CO2






2011EI3731







httpsdoiorg101023A1018940901744











1010801479053X2010502383




101038nclimate1732














101038d41586-020-00177-3













NBLVDI20CO2





1010292008JD009944












httpsdoiorg10108009669589708667275




doiorg1010292011GL048927








101007s00382-016-3327-9







doiorg101007s40641-016-0050-x









2010JD015139







1979_1920pdf




2010JAMC22511







httpsdoiorg101007s10584-015-1400-8














doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006














































fewer people









symbols




























elevations










































capabilities






















































5 Conclusions















































tions in Colorado
















are less vulnerable













REFERENCES





press







101002jgrd50816





sciadv1600873




101007s10584-009-9708-x




[0024MOLUCI]20CO2






2011EI3731







httpsdoiorg101023A1018940901744











1010801479053X2010502383




101038nclimate1732














101038d41586-020-00177-3













NBLVDI20CO2





1010292008JD009944












httpsdoiorg10108009669589708667275




doiorg1010292011GL048927








101007s00382-016-3327-9







doiorg101007s40641-016-0050-x









2010JD015139







1979_1920pdf




2010JAMC22511







httpsdoiorg101007s10584-015-1400-8














doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006





































elevations










































capabilities






















































5 Conclusions















































tions in Colorado
















are less vulnerable













REFERENCES





press







101002jgrd50816





sciadv1600873




101007s10584-009-9708-x




[0024MOLUCI]20CO2






2011EI3731







httpsdoiorg101023A1018940901744











1010801479053X2010502383




101038nclimate1732














101038d41586-020-00177-3













NBLVDI20CO2





1010292008JD009944












httpsdoiorg10108009669589708667275




doiorg1010292011GL048927








101007s00382-016-3327-9







doiorg101007s40641-016-0050-x









2010JD015139







1979_1920pdf




2010JAMC22511







httpsdoiorg101007s10584-015-1400-8














doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006















5 Conclusions















































tions in Colorado
















are less vulnerable













REFERENCES





press







101002jgrd50816





sciadv1600873




101007s10584-009-9708-x




[0024MOLUCI]20CO2






2011EI3731







httpsdoiorg101023A1018940901744











1010801479053X2010502383




101038nclimate1732














101038d41586-020-00177-3













NBLVDI20CO2





1010292008JD009944












httpsdoiorg10108009669589708667275




doiorg1010292011GL048927








101007s00382-016-3327-9







doiorg101007s40641-016-0050-x









2010JD015139







1979_1920pdf




2010JAMC22511







httpsdoiorg101007s10584-015-1400-8














doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006













2011EI3731







httpsdoiorg101023A1018940901744











1010801479053X2010502383




101038nclimate1732














101038d41586-020-00177-3













NBLVDI20CO2





1010292008JD009944












httpsdoiorg10108009669589708667275




doiorg1010292011GL048927








101007s00382-016-3327-9







doiorg101007s40641-016-0050-x









2010JD015139







1979_1920pdf




2010JAMC22511







httpsdoiorg101007s10584-015-1400-8














doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006













doiorg101007s00382-017-3606-0









JTECH-D-12-001911











96GL00265











101007s11027-007-9136-z





1010801368350020171401984









1999WR900090




1dfh-6p97























snowmakingaspx








jgloenvcha201704006












Impact of Global Warming on Snow in Ski Areas - AMS Journals

Documents