Remote Sensing of 2000–2016 Alpine Spring Snowline Elevation in Dall Sheep Mountain Ranges of Alaska and Western Canada

Article Remote Sensing of 2000–2016 Alpine Spring Snowline Elevation in Dall Sheep Mountain Ranges of Alaska and Western Canada

David Verbyla 1,* ID , Troy Hegel 2, Anne W. Nolin 3, Madelon van de Kerk 4, Thomas A. Kurkowski 5 and Laura R. Prugh 4 ID 1 School of Natural Resources and Extension, University of Alaska Fairbanks, Fairbanks, AK 99775, USA 2 Yukon Department of Environment, Whitehorse, YT Y1A 4Y9, Canada; [email protected] 3 College of Earth, Ocean and Atmospheric Sciences, Oregon State University, Corvallis, OR 97331, USA; [email protected] 4 School of Environmental and Forestry Sciences, University of Washington, Seattle, WA 98195, USA; [email protected] (M.v.d.K.); [email protected] (L.R.P.) 5 Scenarios Network for Alaska and Arctic Planning, University of Alaska Fairbanks, Fairbanks, AK 99775, USA; [email protected] * Correspondence: [email protected]; Tel.: +1-907-474-5553

Received: 18 August 2017; Accepted: 7 November 2017; Published: 11 November 2017

Abstract: The lowest elevation of spring snow (“snowline”) is an important factor influencing recruitment and survival of wildlife in alpine areas. In this study, we assessed the spatial and temporal variability of alpine spring snowline across major Dall sheep mountain areas in Alaska and northwestern Canada. We used a daily MODIS snow fraction product to estimate the last day of 2000–2016 spring snow for each 500-m pixel within 28 mountain areas. We then developed annual (2000–2016) regression models predicting the elevation of alpine snowline during mid-May for each mountain area. MODIS-based regression estimates were compared with estimates derived using a Normalized Difference Snow Index from Landsat-8 Operational Land Imager (OLI) surface reflectance data. We also used 2000–2009 decadal climate grids to estimate total winter precipitation and mean May temperature for each of the 28 mountain areas. Based on our MODIS regression models, the 2000–2016 mean 15 May snowline elevation ranged from 339 m in the cold arctic class to 1145 m in the interior mountain class. Spring snowline estimates from MODIS and Landsat OLI were similar, with a mean absolute error of 106 m. Spring snowline elevation was significantly related to mean May temperature and total winter precipitation. The late spring of 2013 may have impacted some sheep populations, especially in the cold arctic mountain areas which were snow-covered in mid-May, while some interior mountain areas had mid-May snowlines exceeding 1000 m elevation. We found this regional (>500,000 km2) remote sensing application useful for determining the inter-annual and regional variability of spring alpine snowline among 28 mountain areas.

Keywords: spring snow; alpine; Dall sheep; MODIS; MODCAG; snowline elevation; snow mapping

1. Introduction Climate warming is altering the duration, amount and timing of snowfall and these changes are especially substantial in arctic and boreal regions [1]. Changing spring snow conditions likely impact many alpine mammals. For example, earlier spring plant phenology initiated earlier emergence from hibernation, earlier weaning, increased body mass and increased survival in yellow-bellied marmots (Marmota ﬂaviventris)[2]. Reduced snow depth during calving allowed parturient caribou (Rangifer tarandus) females to disperse up in elevation away from predator populations, thus increasing recruitment rates [3]. During later springs, upward movement may be limited by snow, resulting in

Remote Sens. 2017, 9, 1157; doi:10.3390/rs9111157 www.mdpi.com/journal/remotesensing Remote Sens. 2017, 9, 1157 2 of 18 Remote Sens. 2017, 9, 1157 2 of 18 higher predation ratesrates on neonates [[3].3]. In addition, late late-winter-winter snowfall has been shown to negatively affect caribou calf growthgrowth ratesrates inin AlaskaAlaska [[4].4]. Dall sheep (Ovis dalli dalli) areare anan iconiciconic alpinealpine speciesspecies distributeddistributed acrossacross northwesternnorthwestern NorthNorth America, includingincluding major major mountain mountain ranges ranges in Alaska, in Alaska, Yukon, Yukon, the Northwest the Northwest Territories Territories and northwest and Britishnorthwest Columbia. British Columbia. The species The is aspecies rare example is a rare of example a North of American a North Americ large mammalan large thatmammal occupies that mostoccupies of its most native of rangeits native , with range genetic , with isolation genetic among isolation mountain among rangemountain populations range populations for thousands for ofthousands years [5]. of Because years there[5]. Because is limited there dispersal is limited among dispersal mountain among populations, mountain some populations, populations some may bepopulations vulnerable may as “islandbe vulnerable populations”. as “island There populations”. are many There survival are many challenges survival in thesechallenges high latitudein these mountainshigh latitude including mountains extreme including weather extreme events weather such as events heavy snowfall,such as heavy winter snowfall, icing events winter and icing late springevents snowand late delaying spring springsnow delaying forage production spring forage [6,7] production (Figure1). [6,7] (Figure 1).

(a) (b)

Figure 1. (a) Typical Dall sheep habitat from Wrangell Mountains, Alaska; (b) Dall sheep ewe and Figure 1. (a) Typical Dall sheep habitat from Wrangell Mountains, Alaska; (b) Dall sheep ewe and lamb lamb (both photographs were from late August by D. Verbyla). (both photographs were from late August by D. Verbyla).

Dall sheep populations have declined in recent decades in some areas. Range-wide, Dall sheep populationsDall sheep have populations declined by have 21% declined since 1990 in recent[8] but decades the causes in some remain areas. unknown Range-wide, [9]. Harsh Dall spring sheep populationsweather was haveshown declined to reduce by horn 21% sincegrowth 1990 rates [8 ]in but the the Yukon causes [10] remain and pregnant unknown Dall [9 ].sheep Harsh in springDenali weatherNational was Park shown were tofound reduce to hornavoid growth snow-covered rates in theareas Yukon [11]. [ 10However,] and pregnant no studies Dall sheephave examined in Denali Nationalimpacts of Park spring were snow found conditions to avoid on snow-covered Dall sheep populations areas [11]. across However, their norange, studies in part have due examined to a lack impactsof available of spring regional snow snow conditions products on that Dall capture sheep populations important acrossfeatures their of range,snow-covered in part due landscapes, to a lack ofsuch available as the elevation regional snow of spring products snowline. that capture important features of snow-covered landscapes, such as theThe elevation elevation of springof spring snowline. snowline is important to all alpine herbivores in this region since above this elevationThe elevation forage of is spring restricted snowline to low is importantquality fora toge all and alpine is covered herbivores with in snow. this region Below since snowline, above thissnowmelt elevation has forageoccurred is restrictedand new shoot to low growth quality provides forage and high is quality covered forage with snow.that is Belowavailable snowline, for the snowmeltfirst time after has occurreda long winter. and new Theshoot snowline growth elevation provides may high thus quality affect foragesheep by that altering is available the quantity for the ﬁrstand quality time after of forage. a long winter.Snowline The elevation snowline may elevation also affect may the thus food affect-safety sheep tradeoff by altering for sheep, thequantity because andseveral quality of their of forage. primary Snowline predators elevation occur mainly may also at lower affect elevations the food-safety [12]. tradeoff for sheep, because severalA spring of their warming primary trend predators has occurred occur mainly in recent at lower decades elevations with earlier [12]. spring snowmelt and river and Alake spring ice-out warming in northern trend hasAlaska occurred and Canada in recent [13 decades–15]. However, with earlier despite spring this snowmelt trend andin earlier river andspring lake snowmelt, ice-out inan northernoccasional Alaska late spring and could Canada impact [13– 15sheep]. However, populations despite in some this mountain trend in ranges. earlier springSpring snowmelt,2013 was unusually an occasional late lateat lower spring elevations, could impact with sheep river populationsice-out in Canada in some and mountain interior Alaska ranges. Springoccurring 2013 11 was–20 days unusually later than late atthe lower 22-year elevations, average with(Figure river 2). ice-outSnow conditions in Canada and and impact interior on Alaska lamb occurringsurvival in 11–20 the spring days later of 2013 than likely the 22-year varied average among (Figuremountain2). Snowareas. conditionsIncreased andvariability impact in on lambinter- survivalannual alpine in the springsnow may of 2013 impact likely populations varied among to mountaina greater extent areas. Increasedthan shifting variability mean conditions in inter-annual and alpineadditional snow tools may and impact products populations are therefore to a greater needed extent to than characterize shifting mean this conditionsvariability and across additional broad toolsregions and [16]. products are therefore needed to characterize this variability across broad regions [16].

Remote Sens. 2017, 9, 1157 3 of 18

Figure 2. Spring river ice-out dates observed from 1994–2016 at Nenana, interior Alaska (ﬁlled circles) Figure 2. Spring river ice-out dates observed from 1994–2016 at Nenana, interior Alaska (filled circles) andand Dawson, Dawson, Yukon Yukon (open (open circles). circles). TheThe 1994–20161994–2016 mean mean date date of ofspring spring ice-out ice-out is show is show as dot as-dashed dot-dashed lineline for Nenanafor Nenana and and ﬁner finer dashed dashed line line for for Dawson. Dawson. Sources: http://www.nenanaakiceclassic.com http://www.nenanaakiceclassic.com, , http://yukonriverbreakup.comhttp://yukonriverbreakup.com..

DallDall sheep sheep inhabit inhabit a a wide widerange range ofof alpine areas areas in in Alaska Alaska and and western western Canada, Canada, ranging ranging from from arcticarctic tundra tundra alpine alpine areas areas north northof of latitudinallatitudinal tree tree line line,, to to alpine alpine areas areas where where glaciers glaciers dominate dominate the the landscape,landscape, to relativelyto relatively warm warm alpine alpine areas.areas. How How does does winter winter precipitation precipitation and andspring spring temperature temperature varyvary among among these these alpine alpine areas? areas? ObjectivesObjectives of of this this research research were were to:to: 1. Use gridded climate products to estimate winter precipitation and May temperature among 28 1. Usemajor gridded mountain climate areas products in Alaska to and estimate northwest winter Canada. precipitation and May temperature among 2.28 majorDevelop mountain 2000–2016 areas remotely in Alaska sensed and estimates northwest of spring Canada. snowline elevation during the typical 2. Developlambing 2000–2016 period of mid remotely-May among sensed the estimates 28 major mountain of spring areas. snowline elevation during the typical 3.lambing Model period mean of2000 mid-May–2016 spring among snowline the 28 majorelevation mountain as a function areas. of decadal mean winter precipitation and mean spring temperature. 3. Model mean 2000–2016 spring snowline elevation as a function of decadal mean winter 4. Assess the regional variability of remotely sensed spring snowline during the unusually cold precipitation and mean spring temperature. and late spring of 2013. 4. Assess the regional variability of remotely sensed spring snowline during the unusually cold and Regional snowline estimation varies with application such as the lowest elevation of alpine snow late spring of 2013. at the end of the summer [17] or the equilibrium line altitude for estimation of glacier mass balance [18].Regional Spring snowline snowline estimationis the regional varies elevation with applicationwhere at lower such elevations as the lowest snowmelt elevation has occurred. of alpine In snow the French Alps spring snowline was estimated using a time series of MODIS data as locations with at the end of the summer [17] or the equilibrium line altitude for estimation of glacier mass balance [18]. first snow-free day with an uncertainty of ± 3 days [19]. In Slovakia, a time series of daily MODIS Springsnow snowline cover was is used the regionalto estimate elevation snowline where as the elevation at lower for elevations which the snowmelt area of snow has pixels occurred. below In the Frenchand Alpsnon-snow spring pixels snowline above was is minimized estimated for using each a timeday [20]. series The of method MODIS was data above as locations 75 percent with first snow-freeaccurate day during with spring an uncertainty months when of ± compared3 days [19 to]. field In Slovakia, data, however a time the series method of daily may MODIS not be snow coverapplicable was used to regions to estimate where snowline perennial asglaciers the elevation below the forregional which snowline the area would of snow be misclassified pixels below as and non-snowsnow pixels pixels each above day. is In minimized this study we for use each a regression day [20]. Theapproach method to estimate was above the spring 75 percent regional accurate duringsnowline spring elevation months across when 28 compared mountain to areas field for data, 17 howeveryears. This the application method may may not be beuseful applicable for to regionsexamining where the perennial inter-annual glaciers and below regional the effects regional of changing snowline spring would snowline be misclassified on alpine as wildfire snow pixels eachpopulations. day. In this study we use a regression approach to estimate the spring regional snowline elevation across 28 mountain areas for 17 years. This application may be useful for examining the inter-annual and regional effects of changing spring snowline on alpine wildfire populations. Remote Sens. 2017, 9, 1157 4 of 18

2. Methods

2.1. Study Region The study region included major Dall sheep mountain areas across northwestern North America, which encompasses this species’ global distribution. Dall sheep generally inhabit the elevation zone at or above tree line in arctic and subarctic mountains where they typically feed in alpine meadows near steep, rocky slopes that serves as escape terrain from predators such as wolves, coyotes and bears. The study region included 28 alpine areas (Table1) in Alaska, Yukon, the Northwest Territories and British Columbia in Canada. Because of a maritime to continental climate gradient, we delineated several study areas within the Alaska Range, Brooks Range and Wrangell mountains. Typically, spring snowline is at lower elevations on the south side relative to the north side of the Alaska Range and Wrangell mountains due to the maritime inﬂuence of the Paciﬁc Ocean. Both the Alaska and Brooks Range have a west-to-east gradient of maritime to continental climate, with spring snowline typically at lower elevation in western portions of these ranges (Figure3).

Table 1. Characteristics of Dall Sheep mountain areas. See Section 2.2 for methods used to estimate May temperature and winter precipitation.

Elevation Mean May Total Winter Mountain Area Area km2 Climatic Class Range (m) Temp. (◦C) Precip. (mm) (1) Alaska Range East, South Slope 12,177 500–2500 5.2 124 Interior (2) Alaska Range East, North Slope 13,284 300–2500 6.6 112 Interior (3) Alaska Range Central, South Slope 15,314 50–2500 4.8 442 High Snowpack (4) Alaska Range Central, North Slope 28,124 100–2500 6.7 145 Interior (5) Alaska Range West, South Slope 18,583 0–2500 4.8 936 High Snowpack (6) Alaska Range West, North Slope 21,181 0–2500 5.8 458 High Snowpack (7) Brooks Range East, South Slope 65,367 200–2400 3.8 64 Cold Arctic (8) Brooks Range East, North Slope 32,698 50–2500 0.2 83 Cold Arctic (9) Brooks Range Central, South Slope 33,797 50–2300 2.0 185 Cold Arctic (10) Brooks Range Central, North Slope 31,667 100–2200 −0.1 156 Cold Arctic (11) Brooks Range West, South Slope 23,367 50–1500 1.8 248 Cold Arctic (12) Brooks Range West, North Slope 20,311 150–1400 0.1 235 Cold Arctic (13) Chugach Mountains 23,937 0–2500 4.7 807 High Snowpack (14) Coast Mountains 18,688 350–2350 5.8 254 Interior (15) Dawson Range, 9715 600–2000 6.2 113 Interior (16) Kenai Mountains 11,317 0–1900 7.6 914 High Snowpack (17) Kluane Mountains 9342 500–2500 5.5 112 Interior (18) Mackenzie Mountains 60,535 200–2700 5.8 232 Interior (19) Olgilvie Mountains 39,956 200–2200 5.2 125 Interior (20) Pelly Mountains 39,118 600–2300 5.7 185 Interior (21) Richardson Mountains 17,099 150–1700 3.0 161 Interior (22) Ruby Range 14,256 600–2300 6.4 116 Interior (23) Selwyn Mountains 33,511 500–2500 5.9 215 Interior (24) Talkeetna Mountains 22,634 100–2500 5.1 326 High Snowpack (25) Tanana Uplands 60,034 150–1900 7.0 84 Interior (26) Wernecke Mountains 27,722 350–2500 4.0 174 Cold Arctic (27) Wrangell Mountains, North Slope 18,895 500–2500 6.6 100 Interior (28) Wrangell Mountains, South Slope 32,994 50–2500 6.3 480 High Snowpack Remote Sens. 2017, 9, 1157 5 of 18

Figure 3. Alpine climate classes based on regional climatology. (1) Cold arctic mountains in the Tundra biome; (2) High snowpack mountains in the Boreal biome influenced by southerly storm tracks from the Pacific Ocean; (3) Interior Mountains in the Boreal biome. Alpine areas included in this study: (1) Alaska Range East, South Slope; (2) Alaska Range East, North Slope; (3) Alaska Range Central, South Slope; (4) Alaska Range Central, North Slope; (5) Alaska Range West, South Slope; (6) Alaska Range West, North Slope; (7) Brooks Range East, South Slope; (8) Brooks Range East, North Slope; (9) Brooks Range Central, South Slope; (10) Brooks Range Central, North Slope; (11) Brooks Range West, South Slope; (12) Brooks Range West, North Slope; (13) Chugach Mountains; (14) Coast Mountains; (15) Dawson Range; (16) Kenai Mountains; (17) Kluane Mountains; (18) Mackenzie Mountains; (19) Olgilvie Mountains; (20) Pelly Mountains; (21) Richardson Mountains; (22) Ruby Range; (23) Selwyn Mountains; (24) Talkeetna Mountains; (25) Tanana Uplands; (26) Wernecke Mountains; (27) Wrangell Mountains North; (28) Wrangell Mountains South. Arrows portray major winter storm tracks from the Pacific Ocean.

2.2. Climatic Variation among Mountain Areas Based on well-known regional climate [21–23], we grouped our mountain areas into 3 classes (Figure3): (1) cold arctic mountains; (2) high snowpack mountains; and (3) interior mountains. These cold arctic mountains were within the global tundra biome, with a monthly mean May temperature of 4 ◦C or less (Table1). The other mountain areas were within the global boreal forest biome [ 24]. The high snowpack mountain areas intercept winter moisture from the Pacific Ocean air mass flow associated with the Aleutian Low, resulting in heavy winter snow accumulation and substantial alpine glaciers. The interior mountain areas receive less Pacific Ocean air mass flow, resulting in a more continental climate with lower precipitation, warmer spring/summer temperatures and most of these mountains lack glaciers. Mountain areas from the high snowpack class had winter precipitation ranging from 326–936 mm, while mountain areas from the interior mountain class ranged from 84–254 mm (Table1). The elevation of spring snowline is dependent on the depth of snow pack accumulated over the winter and spring temperature above freezing. Therefore, we expected spring snowline elevation

Remote Sens. 2017, 9, 1157 6 of 18 to vary with the amount of snow accumulated over the winter (October–April) and with spring temperature. We created 2-km grids of winter precipitation and mean May temperature from the most recent decade of 2000–2009 for each of the 28 mountain areas. Gridded climate data were not available for the entire study region after 2009. These grids were based on 0.5 degree monthly climate products (Climate Research Unit—CRU TS 3.1) [25] which were downscaled by the Scenarios Network for Alaska and Arctic Planning (SNAP) program at the University of Alaska Fairbanks [26]. These products were downscaled via the delta method [27,28] using Param-elevation Relationships on Independent Slopes Model (PRISM) [29] with 1961–1990 2-km resolution climate normals (monthly temperature and precipitation) as baseline climate. The delta method was implemented by calculating climate anomalies applied as differences in temperature and quotients in precipitation, between monthly CRU data and PRISM climate normals for 1961–1990. These coarse-resolution anomalies were then interpolated to PRISM spatial resolution via a spline technique and then added to (temperature) or multiplied by (precipitation) the PRISM climate normals. The PRISM climatology product was used as baseline climate in the downscaling procedure because it accurately represents the elevational effects on precipitation patterns across mountainous regions [30] and it is based on extensive use of weather stations across Alaska: 455 for precipitation, 316 for temperature, as well as the European Center for Medium-range Weather Forecasts’ reanalysis of temperatures at the 500 mb height [31]. We used the 2-km decadal climate grids to estimate the mean May temperature and total October–April precipitation for each of the 28 areas (elevation zone from 500–1000 m). Virtually all October–April precipitation in these areas is snow. We then developed a linear model predicting MODIS-based mean 15 May snowline elevation (see Section 2.4) as a function of decadal mean May temperature and October–April precipitation.

2.3. Elevation Zones To delineate alpine elevation zones, we used the Global Multi-resolution Terrain Elevation Data 2010 produced by the U.S. Geological Survey (USGS) and the National Geospatial-Intelligence Agency (NGA) (http://earthexplorer.usgs.gov) at 30 arc second spacing. To match the other geospatial datasets, the elevation data were projected to the Alaska Albers NAD83 coordinate system using bilinear interpolation to 500-m pixel size. For each mountain range, we then created a raster of elevation zones at 100-m elevation intervals. These elevation zone grids were used with the time series of daily remotely sensed snow fraction to estimate the mean day of year of last spring snow for each zone.

2.4. Remotely Sensed Snow Fraction We used a daily snow fraction product, from the Moderate Resolution Imaging Spectrometer (MODIS) sensor that covered our entire study area at 500-m pixel size (global tiles H10V02, H11V02, H12V02). The MODIS Snow-Covered Area and Grain size product (MODSCAG) adapts methods originally developed from imaging spectrum [32] to use the spectral information from MODIS to estimate subpixel snow properties—fractional snow-covered area, grain size and albedo. The MODSCAG snow fraction product has been validated with fractional snow cover derived using Landsat sensors [33] resulting in a root mean squared error of 5% averaged across 31 Landsat scenes. This product is superior to the MOD10A1 snow extent products, especially during the period of spring snowmelt [34]. The physically-based MODSCAG algorithm uses spectral mixture analysis on a pixel-by-pixel basis to derive gridded 500 m daily fraction of snow cover estimates. The linear spectral mixture procedure is based on land surface endmembers (e.g., snow, soil, rock, vegetation, lake ice) that matches surface reflectance from the Terra MOD09GA (MODIS daily gridded surface reflectance) product. The MODSCAG model uses the relative shape of the snow’s spectrum and is applicable to mountainous areas where the local solar illumination angle on a slope is often unknown because of co-registration errors between the image and a digital elevation model. The MODSCAG model estimates the fraction of each pixel that is covered by snow [33]. A threshold value of >0.15 to flag Remote Sens. 2017, 9, 1157 7 of 18 Remote Sens. 2017, 9, 1157 7 of 18 each daily 500-m pixel as snow covered has been used to accurately map snow extent for treeless eachareas daily in the 500-m Colorado pixel asRocky snow Mountains, covered has upper been Rio used Grande, to accurately Himalayas map snow in Nepal extent [35] for and treeless the Sierra areas inNevada the Colorado Mountains Rocky in Mountains,California [36]. upper Based Rio Grande,on these Himalayas results, we in used Nepal a fraction [35] and threshold the Sierra value Nevada of Mountains>0.15 to flag in each California daily 500 [36].-m Based pixel onas thesesnow results,covered. we used a fraction threshold value of >0.15 to flag eachFor daily 2000 500-m through pixel as2016, snow the covered. three daily snow fraction tiles were mosaicked to one daily grid, coveringFor 2000the entire through study 2016, area. the The three daily daily snow snow fraction fraction grid tileswas werethen projected mosaicked to tothe one Alaska daily Albers grid, coveringNAD83 projection the entire studyusing area. nearest The dailyneighbor snow resampling fraction grid at was500 then-m pixel projected size toto thematch Alaska the Albers other NAD83geospatial projection datasets. using For nearesteach day neighbor from April resampling 1 through at 500-m July pixel 31, each size to500 match-m pixel the otherwas flagged geospatial as datasets.“Snow” if For the each snow day fraction from 1exceeded April through 0.15, resulting 31 July , eachin a time 500-m series pixel of was daily flagged snow asgrids. “Snow” Any ifpixel the snowthat did fraction not have exceeded a valid 0.15, snow resulting fraction invalue a time (primarily series of due daily to snowcloud grids.contamination) Any pixel was that excluded did not havefrom aanalysis. valid snow From fraction this daily value time (primarily series, duethe last to cloud day of contamination) snow grid was was computed excluded as from day analysis. of year Fromranging this from daily 91 time (April series, 1) to the 213 last (July day 31) of for snow each grid pixel. was Any computed pixel with as daysnow of detected year ranging on 1- fromAugust 91 (1was April) likely to from 213 (31 perennial July) for snowfields each pixel. or Any glaciers pixel and with these snow pixels detected were on excluded 1-August from was the likely analysis. from perennialFor each day snowfields from April or glaciers 1 through and theseJuly 31, pixels each were 500 excluded-m pixel fromwas flagged the analysis. as “Snow” For each if the day snow from 1fraction April through exceeded 31 0.15, July, resulting each 500-m in pixela time was series flagged of daily as“Snow” snow grids. if the From snow this fraction time exceededseries, the 0.15, last resultingday of snow in a grid time was series computed of daily as snow day grids. of year From ranging this timefrom series,91 (April the 1) last to day 213 of(July snow 31) grid for each was computedpixel (Figure as day4). of year ranging from 91 (1 April) to 213 (31 July) for each pixel (Figure4).

FigureFigure 4.4. Last day of springspring snowsnow detecteddetected fromfrom dailydaily MODSCAGMODSCAG datadata forfor 2016.2016. Mountain areas areare outlinedoutlined asas blackblack polygons.polygons. TheThe namesnames ofof eacheach mountainmountain areaarea areare presentedpresented inin FigureFigure1 .1.

ForFor eacheach 100-m100-m elevationelevation zonezone withwith atat leastleast 1000 pixels, thethe mean day of year of last spring snow waswas thenthen computed.computed. We usedused aa samplesample thresholdthreshold ofof 10001000 pixelspixels toto minimizeminimize locallocal effectseffects inin smallersmaller elevationelevation zoneszones thatthat werewere fromfrom inin aa mountainmountain area.area. ForFor eacheach ofof thethe 2828 mountainmountain areas,areas, wewe estimatedestimated thethe MayMay snowlinesnowline elevationselevations byby regressionregression equationsequations predictingpredictingthe theelevation elevation of of last last snow snow as as a a function function of of mean mean day day of of last last snow. snow. We We developed developed a Maya May regression regression for for each each of of the the 17 17 years years (2000–2016) (2000–2016) for for each each of of the the28 28 mountainmountain areas.areas. AllAll regression 2 equationsequations had an an R R2 exceedingexceeding 0.95, 0.95, with with the the mean mean day day of last of lastspring spring snow snow for each for eachelevation elevation zone zonebased based on at least on at 1000 least pixels 1000 pixelsper zone per (Figure zone (Figure 5). Because5). Because our goal our was goal to develop was to developa regional a regionalsnowline snowlineelevation elevationproduct to product enable tobroad enable-scale broad-scale spatial comparisons spatial comparisons (e.g. among (e.g. mountain among mountainareas) and areas)inter-

Remote Sens. 2017, 9, 1 1157157 8 of 18 Remote Sens. 2017, 9, 1157 8 of 18 annual comparisons, we did not account for localized variation in factors such as slope and aspect annual comparisons, we did not account for localized variation in factors such as slope and aspect thatand inter-annualwould result comparisons,in local-scale weelevational did not accountvariations. for localized variation in factors such as slope and aspectthat would that wouldresult in result local in-scale local-scale elevational elevational variations. variations.

Figure 5. Example of MODIS-based 2000–2016 regression trend lines predicting snowline elevation FigureFigure 5.5. ExampleExample ofof MODIS-basedMODIS-based 2000–20162000–2016 regressionregression trendtrend lineslines predictingpredicting snowlinesnowline elevationelevation from mean day of last snow within each elevation zone, Chugach Mountain alpine area. The earliest fromfrom meanmean dayday ofof lastlast snowsnow withinwithin eacheach elevationelevation zone,zone, ChugachChugach MountainMountain alpinealpine area.area. TheThe earliestearliest (2009) and latest spring (2000) above 700 m are portrayed as thicker trend lines. (2009)(2009) andand latestlatest springspring (2000)(2000) aboveabove 700700 mm areare portrayedportrayedas as thicker thicker trend trend lines. lines. Figure 6 summarizes our methods in a flowchart. FigureFigure6 6 summarizes summarizes our our methods methods in in a a ﬂowchart. flowchart.

Figure 6. Flowchart of geospatial methods used for each year (2000 (2000–2016).–2016). Figure 6. Flowchart of geospatial methods used for each year (2000–2016).

Remote Sens. 2017, 9, 1157 9 of 18 Remote Sens. 2017, 9, 1157 9 of 18

2.52.5.. Validation Validation We usedused the the Landsat-8 Landsat Operational-8 Operational Land Land Imager Imager (OLI) data(OLI) to data validate to ourvalidate MODSCAG our MODSCAG regression regressionapproach. Atapproach. this high At latitude this high May latitude solar elevation May solar at Landsat elevation overpass at Landsat can be overpass as low as can 40 be degrees as low thus as 40a scene degrees with thus 25percent a scene cloudwith 25 coverage percent will cloud have coverage >50 percent will areahave covered >50 percent by cloud area andcovered cloud by shadow. cloud andWe selectedcloud shadow. Landsat We scenes selected with Landsat less than scenes 10 percent with less cloud than coverage 10 percent from cloud 10–31 coverage May, 2013–2016from May 10throughout–31, 2013– the2016 study throughout area (Figure the study7). area (Figure 7).

FigureFigure 7. ThirtyThirty Landsat Landsat-8-8 OLI OLI scenes scenes used used to to validate MODSAG MODSAG-based-based regression models.

To mapmap snow snow pixels, pixels, we we computed computed a Normalized a Normalized Difference Difference Snow IndexSnow (NDSI) Index for(NDSI) each Landsat-8for each Landsatscene. We-8 scene. computed We computed NDSI as: NDSI as: NDSI = (ρBand3 − ρBand6)/(ρBand3 + ρBand6) NDSI = (ρBand3 − ρBand6)/(ρBand3 + ρBand6) where ρ is the surface reflectance at 0.525–0.600 μm (Band3) and 1.56–1.66 μm (Band6). A 30-m pixel waswhere mappedρ is the as surface snowreﬂectance covered when at 0.525–0.600 NDSI >0.40,µm which (Band3) is anda standard 1.56–1.66 thresholdµm (Band6). [30,31 A]. 30-m We pixelthen classifiedwas mapped each as 500 snow-m pixel covered as “Snow” when NDSI if the >0.40, pixel whichcontained is a at standard least 50threshold percent of [30 the,31 30].- Wem snow then pixels.classiﬁed We each then 500-m determined pixel as the “Snow” percent if theof snow pixel pixels contained within at leasteach 50500 percent-m elevation of the 30-mpixel snowand plotted pixels. theWe thenmean determined percent snow the percentby elevation of snow zones pixels to within estimate each the 500-m elevation elevation with pixel 50 percent and plotted snow the pixels. mean (Figurepercent 8c). snow For by the elevation same date, zones we toused estimate the MODSCAG the elevation based with regression 50 percent to snowestimate pixels. the elevation (Figure8 c).of springFor the snowline same date, (Figure we used8d). More the MODSCAG than one mountain based regression area could tobeestimate in one Landsat the elevation scene. When of spring this occurred,snowline (Figurewe estimated8d). More the than elevation one mountain of spring area snowline could be infor one each Landsat mountain scene. area, When resulting this occurred, in 53 snowlinewe estimated estimates the elevation based on ofthe spring Landsat snowline NDSI and for MODSCAG each mountain regressions. area, resulting in 53 snowline estimates based on the Landsat NDSI and MODSCAG regressions.

Remote Sens. 2017, 9, 1157 10 of 18 Remote Sens. 2017, 9, 1157 10 of 18

(a) (b)

FigureFigure 8. ((aa)) Landsat Landsat OLI OLI image image from from 18 18-May-2015-May-2015 with with 940 940 m m and and 1065 1065-m-m contours contours;; ( (bb)) MODSCAG snow/nosnow/no snow snow classification classiﬁcation for for same same date date and and area area;; ( (c)c) Snowline Snowline elevation elevation estimated estimated at at 1065 1065 m m based based onon 18 18-May-2015-May-2015 Landsat NDSI NDSI;; ( d)d) Snowline elevation elevation estimated at 940 m based on 2015 2015 MODSCAG MODSCAG lastlast day of spring spring snow snow linear linear regression.

3. Results 3. Results

3.1.3.1. Comparison Comparison of of Landsat Landsat NDSI NDSI and and MODSCAG Based Snowline Elevation Estimates EstimatesEstimates ofof MayMay snowline snowline elevation elevation from from Landsat Landsat NDSI NDSI and MODSCAGand MODSCAG regressions regressions were similar were similar(Figure 9(Figure), with 9), amean with a absolute mean absolute error of error 106 m of (standard 106 m (standard deviation deviation of 77 m). of 77 The m). Landsat The Landsat NDSI NDSIbased based snowline snowline estimates estimates were typically were typically at higher at elevation, higher elevation, perhaps dueperhaps to the due ﬁner to spatialthe finer resolution spatial resolution(30-m pixels (30 for-m Landsat pixels for NDSI, Landsat 500-m NDSI, pixels 500 for- MODSCAG)m pixels for orMODSCAG) due to the differentor due to methodological the different methodologicalapproaches. The approaches. uncertainty The of elevation uncertainty due of to elevation use of a regional due to use elevation of a regional raster likely elevation introduced raster likelyerror inintroduced both the MODSCAG error in both regression the MODSCAG and Landsat regression NDSI validationand Landsat datasets. NDSI Alsovalidation because datasets. of local Alsoconditions because such of local as slope conditions direction, such topographic as slope direction, position, topographic topographic position, control topographic of wind-blown control snow of winddistribution,-blown snow there isdistribution, no contour elevationthere is no that contour perfectly elevation matches that snowline perfectly at amatches local scale snowline (Figure at8a). a local scale (Figure 8a).

Remote Sens. 2017, 9, 1157 11 of 18 Remote Sens. 2017, 9, 1157 11 of 18

Figure 9. Comparison of May snowline elevation estimates from Landsat NDSI and MODSCAGMODSCAG regressions (n = 53). These estimates were from May 2013, May 2014, May 2015 covering Alaska and Yukon mountainmountain areasareas (Figure(Figure7 7).). Day Day of of year year ranged ranged from from 5 May May 5 through through 30 May May. 30.

3.2. Mountain Area Climates Based on the gridded climatic products, the 3 regionalregional climaticclimatic classes were distinct.distinct. The Cold Arctic Mountain class always had a meanmean MayMay temperaturetemperature ofof lessless thanthan 4.04.0 ◦°CC (as(as lowlow asas −−0.1 °◦C).C). These mountainsmountains included included all all six six areas areas of the ofBrooks the Brooks Range Range in Alaska in andAlaska the Richardsonand the Richardson and Wernecke and MountainsWernecke inMountains northern Yukon.in northern The other Yukon. mountain The areasother had mountain warmer meanareas Mayhad temperature;warmer mean all aboveMay 4.5temperature;◦C and as all high above as 7.6 4.5◦C. °C The and High as high Snowpack as 7.6 °C class. The always High Snowpack had a winter class precipitation always had exceeding a winter 300precipitation mm, with exceeding a high of 936 300 mm. mm, These with werea high major of 936 glaciered mm. These mountains were major of Alaska glaciered including mountains the south of slopesAlaska of including the central the Alaska south Rangeslopes andof the Wrangell central Mountains, Alaska Range the entireand Wrangell western Mountains, Alaska Range the and entire the Chugach,western Alaska Talkeetna Range and and Kenai the Mountains.Chugach, Talkeetna The Interior andMountains Kenai Mountains. class had The a winter Interior precipitation Mountains lessclass than had 300a winter mm, precipitation with a low of less 84 mm.than 300 These mm, included with a low most of of84 the mm. ranges These in included central Alaskamost of and the northwestranges in central Canada Alaska that were and lessnorthwest influenced Canada by Pacific that were Ocean less storm influenced flows. Theby Pacific warmest Ocean and drieststorm areaflows. was The the warmest Tanana and Uplands driest which area was is in the interior Tanana Alaska, Uplands with which mean is May in interior temperature Alaska, of with 7.0 ◦ Cmean and lessMay than temperature 85 mm of of total 7.0 °C winter and precipitation.less than 85 mm of total winter precipitation.

3.3. 2000 2000–2016–2016 Mean May 15 May 15 Snowline Elevation among Mountain Areas Based on the daily MODSCAG snow fraction product, the mean 152000 May–2016 2000–2016 May 15 snowline elevation of snow line ranged from 0 to over 2000 m (Figure 1010a)a) withwith thethe interiorinterior mountainsmountains class having the highest snowline elevation. Areas with mean May 15 May 15 snowline elevation less than 400 m were from cold arctic areasareas (Richardson(Richardson Mountains, Brooks Range) or areas with high winter snow accumulation due to storm flow flow from the PacificPacific Ocean (Alaska Range central and western region, south slopes) (Figure 1010b).b). The mean 2000 2000–2016–2016 snowline elevation was significantly significantly related to mean May temperature and winter precipitation (Table(Table2 2).).

Remote Sens. 2017, 9, 1157 12 of 18 Remote Sens. 2017, 9, 1157 12 of 18

2000 Cold Arctic High Snowpack Interior

1500

)

(

e 1000

S 500

s s s s s s

e e e e e e e e e e e e e e e e

n n n n n n n n d n n n

p p p p p p p p p p p p p p g g

t t t t t t t t t t t

o o o o o o o o o o o o o o n n

l l l l l l l l l l l l l l

M M

M M M M M M M a M M

a a

s s s s l

S S S S S S S S S S

y p R R

n h e a e n e e

N N N

− −

S S S a S i S s i

N N N

c y

o l n n

t t U

l v z y

t t l n t t a n

t t t

l e

s a

a a

s s a

s s a s s n

s a e s o o

d g

a g a

l u

e e a a e s

e e l

r e l

n C

E E

a E E

w K

h O

W W

e e

T n

i g g

g g

n n

a a

R R

s K

B B (a)

(b)

Figure 10.10. ((aa)) Box Box plots of estimated 15May May 15 snowlinesnowline elevations.elevations. The The lower lower and upper hinges correspond toto thethe ﬁrstfirst and and third third quartiles quartiles (the (the 25th 25th and and 75th 75th percentiles). percentiles). The The horizontal horizontal line line in the in boxthe boxrepresents represents the median.the median. The The upper/lower upper/lower whisker whisker lines lines extend extend from from the the hinge hinge to theto the largest largest value value no nofurther further than than 1.5 1.5 * IQR * IQR from from the the hinge hinge (where (where IQR IQR is the is the inter-quartile inter-quartile range, range, or distance or distance between between the theﬁrst first and and third third quartiles); quartiles) (b); 2000–2016(b) 2000–2016 mean mean 15 May May snowlines 15 snowlines estimated estimated from from regressions. regressions.

Remote Sens. 2017, 9, 1157 13 of 18

Remote Sens. 2017, 9, 1157 13 of 18 Table 2. Regression model predicting mean 2000–2016 snowline elevation as a function of 2000–2009 2 decadalTable 2. meanRegression May temperature model predicting and mean mean total 2000 winter–2016 precipitationsnowline elevation (p < 0.001, as a Rfunction= 0.71, ofn 2000= 28).–2009 decadal mean May temperature and mean total winter precipitation (p < 0.001, R2 = 0.71, n = 28). Coefﬁcient Std. Error p-Value InterceptCoefficient 337.96 Std. 117.53Error 0.00812p-Value MayIntercept temperature (◦C)337.96 148.36 117.53 22.032 <0.00010.00812 May temperatureWinter Precip (°C) (mm) 148.36−0.9837 22.032 0.1934 <0.0001<0.0001 Winter Precip (mm) −0.9837 0.1934 <0.0001

Due to the the influence inﬂuence of of storms storms from from the the south, south, the the snowline snowline elevation elevation was was consistently consistently lower lower on onsoutherly southerly slopes slopes of the of theAlaska Alaska Range Range and and Wrangell Wrangell Mountains Mountains (Figure (Figure 11). 11In). the In Alaska the Alaska Range, Range, the thedifference difference in snowline in snowline elevation elevation was was least least in ineastern eastern areas areas due due to to a a west west-to-east-to-east maritime to continental climate gradient.

(a)

(b)

Figure 11. 20002000–2016–2016 snow line mean elevation on May 15 May 15 for north versus south south-facing-facing slopes within (a) western Alaska Range and ((b)) WrangellWrangell Mountains.Mountains. There was also a west-to-eastwest-to-east maritime to continental climatic gradient along the Alaska Range.

3.4. Spring 2013 Snowline Elevation among Mountain Areas 3.4. Spring 2013 Snowline Elevation among Mountain Areas The late spring conditions of 2013 where the May 15 snow line elevation was less than 100 m, The late spring conditions of 2013 where the 15 May snow line elevation was less than 100 m, were primarily in the cold arctic mountain areas (Brooks Range, Richardson Mountains, Wernecke were primarily in the cold arctic mountain areas (Brooks Range, Richardson Mountains, Wernecke Mountains) and the high winter precipitation areas of the south slopes of the western and central Mountains) and the high winter precipitation areas of the south slopes of the western and central Alaska Range (Figure 12). Eastern mountain areas with continental climate such as the north slope of the Wrangells and mountain ranges in southern Yukon had May 15 snowline elevations above 700 m.

Remote Sens. 2017, 9, 1157 14 of 18

Alaska Range (Figure 12). Eastern mountain areas with continental climate such as the north slope of Remotethe Wrangells Sens. 2017, and9, 1157 mountain ranges in southern Yukon had 15 May snowline elevations above14 700 of m.18

FigureFigure 12. 12. ModeledModeled elevation elevation of of snow snow line line on on May 15 May 15, 2013 (Day of Year 135).135).

4. Discussion We developed a new application of the MODSCAG snow fraction product that accurately characterizes spatialspatial andand temporal temporal regional regional variation variation in in the the elevation elevation of springof spring snowline. snowline. We usedWe used Dall Dallsheep sheep as a caseas a studycase study because because this alpine this alpine species species is declining is declining and thought and thought to be sensitiveto be sensitive to spring to spring snow snowconditions conditions [8]. However, [8]. However, the elevation the elevation of spring of snowline spring snowline is important is important for many alpinefor many herbivores, alpine herbivores,because it affects because access it affects to forage access as towell forage as as forage well qualityas forage and quality quantity. and quantity. New plant New shoot plant growth shoot growthtypically typically begins after begins snowmelt after snowmelt and forage and quality forage is quality highest isin highest new plant in new growth plant because growth ofbecause high cell of highsoluble cell content, soluble whichcontent, declines which asdeclines plants as matures plants andmatures ﬁber and accumulates fiber accumulates [19]. In addition, [19]. In addition, nitrogen nitrogencontent of content forage of species forage typically species typically declines declines with time with after time snowmelt after snowmelt [37,38]. [37,38 Higher]. Higher quality quality forage forageallows allows for shorter for shorter rumination rumination time and time thus and more thus more time totime forage, to forage, the so-called the so-called multiplier multiplier effect effect [39]. Winter[39]. Winter forage, forage, or forage or forage above above snowline, snowline, is typically is typically low quality: low quality: higher higher in ﬁber in andfiber lower and lower in crude in proteincrude protein [40]. Thus [40]. if Thus spring if snowlinespring snowline elevation elevation is relatively is relatively low, forage low, quantity forage andquantity quality and would quality be wouldreduced be relative reduced to relative a warm to spring a warm when spring the snowlinewhen the elevationsnowline iselevation higher. is higher. The variation among the 28 mountain areas was substantial in terms of decadal climate and elevation of spring snowline. For example, based on the gridded May temperature product, the cold ◦ arctic mountain areasareas hadhad aa meanmean MayMaytemperature temperature of of 1.9 1.9 C,°C, while while the the interior interior mountain mountain areas areas had had a ◦ amean mean May May temperature temperature of of 6.1 6.1 C.°C. Based Based on on the theMODSCAG MODSCAG dailydaily snowsnow fractionfraction product,product, the mean 15May May 15 snowline elevation for the cold arctic mountain areas was 339 m, while the interior mountain areas had a mean 15May May 15 snowline elevation of 11451145 m. Variation in spring snow conditions can be substantial, forfor exampleexample in in 2013 2013 the the 15 May May elevation15 elevation of snowline of snowline in the in Chugach the Chugach Mountains Mountains was above was above500 m and500 m lamb and survival lamb survival was high was [41 high] relative [41] torelative the Brooks to the Range Brooks areas Range where areas 15 where May snowline May 15 snowlineelevation elevation was 0 m and was no 0 m lambs and no were lambs observed were observed in this region in this during region 2013 during lamb 2013 survey lamb [42 survey]. [42]. The magnitude of the observed inter-annualinter-annual variability in spring snow conditions is especially relevant because the frequency and magnitude of climate extremes are expected to increase under climate changechange [ 43[43–45–45]. This]. This increased increased variability variability mayhave may stronger have stronger impacts onimpacts ecosystem on functioningecosystem functioningthan gradually than shifting gradually means, shifting because means, wildlife because populations wildlife may populations be resilient tomay occasional be resilient extreme to weatheroccasional conditions extreme weather but unable conditions to persist but if unable these conditions to persist if become these conditions more frequent become [16]. more For example,frequent [inter-annual16]. For example, variation inter- inannual snowline variation was in highest snowline in thewas centralhighest andin the eastern central Brooks and eastern Range Brooks areas Range areas (Figure 10a), which is the region experiencing the steepest decline in sheep populations and emergency harvest closures [42]. Dall sheep populations recover slowly from extreme weather events, since each adult ewe produces a maximum of one lamb per year [46]. Therefore, the effect of harsh years occurring at a higher frequency could be detrimental to their persistence.

Remote Sens. 2017, 9, 1157 15 of 18

(Figure 10a), which is the region experiencing the steepest decline in sheep populations and emergency harvest closures [42]. Dall sheep populations recover slowly from extreme weather events, since each adult ewe produces a maximum of one lamb per year [46]. Therefore, the effect of harsh years occurring at a higher frequency could be detrimental to their persistence. How might spring snow conditions affect sheep populations? First, spring snow cover affects plant phenology, thus affecting nutrition for lactating ewes and hence lamb survival. For example, lambs born in early May experienced higher mortality rates, likely due to a low snowline restricting access to high-quality forage and increasing vulnerability to predators [47]. However, lambs born very late nurse from females feeding on forage of declining quality and post-weaned lambs have access to high quality forage for a shorter time. For example, delayed plant phenology during a late spring corresponded with a 2-week delay in onset of parturition and thus reduction in time available for lamb growth, weaning and acquisition of body reserves for ewes prior to the onset of plant senescence [48]. Second, foraging distance away from escape terrain may be greater due to a low snowline elevation, leading to greater risk of predation and less efﬁcient foraging. For example, ewes foraged closer to escape terrain during a mild spring when forage was plentiful relative to a late spring when forage availability and quality was lower and ewes foraged less efﬁciently [11]. It is likely that sheep populations are more vulnerable to late spring conditions in the cold arctic mountain areas. For example, our regression estimates of 15 May snowline elevation were close to sea level for all cold arctic mountain areas in 2013. These ranges likely had low lamb recruitment in 2013. Although beyond the scope of this paper, our snowline elevation product can be combined with survey data to quantitatively assess the effects of spring snow conditions on alpine wildlife populations across broad regions. For example, we found that sheep recruitment declined with lower 15 May snowline elevations using a linear mixed model of 1570 Dall sheep aerial surveys and the effect increased with latitude [49]. Our focus was on May snowline because plant growth and high quality forage occurs at elevations just below melting snow. This application may also be useful for estimating the start of alpine plant growing season across the region, since new plant growth occurs below the snowline. True Further development of remotely sensed snow products will improve the ability of wildlife managers to anticipate and mitigate responses of wildlife to climate change.

5. Conclusions To assess the inter-annual and regional variability of spring snowline elevation across 28 mountain areas in Alaska and northwestern Canada, we used a regression model approach with a daily regional remote sensing snow product. Due to frequent extensive cloud and cloud shadow cover in our study areas and a repeat orbit of 16-days, this type of regional assessment was not possible with Landsat sensor data. We found the 2000–2016 MODIS MODSCAG snow fraction product useful for this application and regression-based estimates of regional spring snowline elevation compared favorably with 5 May through 30 May snowline elevation estimates derived from Landsat-OLI scenes across Alaska and northwest Canada.

Acknowledgments: This research was supported by the NASA ABoVE program, Grants NNX15AV86A (DV), NNX15AU21A (LRP) and NNX15AU13A (AWN). We have funds for covering the costs to publish in open access. We thank Mark Melham and the anonymous reviewers for helping us improve the manuscript. Author Contributions: David Verbyla processed the data, Troy Hegel delineated the mountain areas in western Canada and provided expertise about Dall sheep populations in Canada, Anne W. Nolin processed the daily MODSCAG tiles covering the study area, Madelon van de Kerk and Laura R. Prugh provided expertise about Dall sheep populations in Alaska, Thomas A. Kurkowski provided expertise about the gridded temperature and precipitation products. All authors collaborated to write the manuscript. Conﬂicts of Interest: The authors declare no conﬂict of interest. Remote Sens. 2017, 9, 1157 16 of 18

References

1. Callaghan, T.V.; Johansson, M.; Brown, R.D.; Groisman, P.Y.; Labba, N.; Radionov, V.; Barry, R.G.; Bulygina, O.N.; Essery, R.L.H.; Frolov, D.M.; et al. The changing face of Arctic snow cover: A synthesis of observed and projected changes. AMBIO 2011, 40 (Suppl. 1), 17–31. [CrossRef] 2. Ozgul, A.; Childs, D.Z.; Oli, M.K.; Armitage, K.B.; Blumstein, D.T.; Olson, L.E.; Tuljapurkar, S.; Coulson, T. Coupled dynamics of body mass and population growth in response to environmental change. Nature 2010, 466, 482–485. [CrossRef][PubMed] 3. Hegel, T.M.; Mysterud, A.; Huettmann, F.; Stenseth, N.C. Interacting effect of wolves and climate on recruitment in a northern mountain caribou population. Oikos 2010, 119, 1453–1461. [CrossRef] 4. Adams, L.G. Marrow fat deposition and skeletal growth in caribou calves. J. Wildl. Manag. 2003, 67, 20–24. [CrossRef] 5. Worley, K.; Strobeck, C.; Arthur, S.; Carey, J.; Schwantje, H.; Veitch, A.; Coltman, D.W. Population genetic structure of North American thinhorn sheep (Ovis dalli). Mol. Ecol. 2004, 13, 2545–2556. [CrossRef][PubMed] 6. Putkone, J.; Roe, G. Rain-on-snow events impact soil temperatures and affect ungulate survival. Geophys. Res. Lett. 2003, 30.[CrossRef] 7. Pettorelli, N.; Pelletier, F.; von Hardenberg, A.; Festa-Bianchet, M.; Côté, S.D. Early onset of vegetation growth vs. rapid green-up: Impacts on juvenile mountain ungulates. Ecology 2007, 88, 381–390. [CrossRef] [PubMed] 8. Alaska Department of Fish and Game. Trends in Alaska Sheep Populations, Hunting, and Harvests. Division of Wildlife Conservation, Wildlife Management Report 2014, ADFG/DWC/WMR-2014-3. Available online: https://www.adfg.alaska.gov/static/home/library/pdfs/wildlife/mgt_rpts/14_sheep_ report_bog.pdf (accessed on 24 January 2017). 9. Alaska Department of Fish and Game. Dall Sheep Management Report of Survey-Inventory Activities July 2007–June 2010, Project6.0, Harper, P. Editor. Available online: https://www.adfg.alaska.gov/static/home/ library/pdfs/wildlife/mgt_rpts/11_sheep.pdf (accessed on 24 January 2017). 10. Loehr, J.; Carey, J.; O’Hara, R.B.; Hik, D.S. The role of phenotypic plasticity in responses of hunted thinhorn sheep ram growth to changing climate conditions. J. Evol. Biol. 2010, 23, 783–790. [CrossRef][PubMed] 11. Rachlow, J.L.; Bowyer, R.T. Habitat selection by Dall’s sheep (Ovis dalli): Maternal trade-offs. J. Zool. 1998, 245, 457–465. [CrossRef] 12. Arthur, S.M.; Prugh, L.R. Predator-mediated indirect effects of snowshoe hares on Dall’s sheep in Alaska. J. Wildl. Manag. 2010, 74, 1709–1721. [CrossRef] 13. Stone, R.S.; Dutton, E.G.; Harris, J.M.; Longenecker, D. Earlier spring snowmelt in northern Alaska as an indicator of climate change. J. Geophys. Res. 2002, 107.[CrossRef] 14. Brown, R.; Derksen, C.; Wang, L. A multi-data set analysis of variability and change in Arctic spring snow cover extent. J. Geophys. Res. 2010, 115, 1967–2008. [CrossRef] 15. Šmejkalová, T.; Edwards, M.E.; Dash, J. Arctic lakes show strong decadal trend in earlier spring ice-out. Sci. Rep. 2016, 6, 38449. [CrossRef][PubMed] 16. Boyce, M.S.; Haridas, C.V.; Lee, C.T.; The NCEAS Stochastic Demography Working Group. Demography in an increasing variable world. Trends Ecol. Evol. 2006, 21, 141–148. [CrossRef][PubMed] 17. Price, M.F.; Byers, A.C.; Kohler, T.; Price, L.W. (Eds.) Mountain Geography: Physical and Human Dimensions; University of California Press: Los Angeles, CA, USA, 2013; p. 396. 18. Shea, J.M.; Menounos, B.; Moore, R.D.; Tennant, C. An approach to derive regional snow lines and glacier mass change from MODIS imagery, Western North America. Cryosphere 2013, 7, 667–680. [CrossRef] 19. Dedieu, J.-P.; Carlson, B.Z.; Bigot, S.; Sirguey, P.; Vionnet, V.; Choler, P. On the Importance of High-Resolution Time Series of Optical Imagery for Quantifying the Effects of Snow Cover Duration on Alpine Plant Habitat. Remote Sens. 2016, 8, 481. [CrossRef] 20. Krajˇcí, P.; Holko, L.; Perdigão, R.A.; Parajka, J. Estimation of regional snowline elevation (RSLE) from MODIS images for seasonally snow covered mountain basins. J. Hydrol. 2014, 519, 1769–1778. [CrossRef] 21. Hammond, T.; Yarie, J. Spatial prediction of climatic state factor regions in Alaska. Ecoscience 1996, 3, 490–501. [CrossRef] 22. Stafford, J.M.; Wendler, G.; Curtis, J. Temperature and precipitation of Alaska: 50 year trend analysis. Theor. Appl. Climatol. 2000, 67, 33–44. [CrossRef] Remote Sens. 2017, 9, 1157 17 of 18

23. Fleming, M.D.; Chapin, F.S.; Cramer, W.; Hufford, G.L.; Serreze, M.C. Geographic patterns and dynamics of Alaskan climate interpolated from a sparse station record. Glob. Chang. Biol. 2000, 6, 49–58. [CrossRef] 24. Olson, D.M.; Dinerstein, E.; Wikramanayake, E.D.; Burgess, N.D.; Powell, G.V.N.; Underwood, E.C.; D’Amico, J.A.; Itoua, I.; Strand, H.E.; Morrison, J.C.; et al. Terrestrial ecoregions of the world: A new map of life on Earth. Bioscience 2001, 51, 933–938. [CrossRef] 25. Harris, I.; Jones, P.D.; Osborn, T.J.; Lister, D.H. Updated high-resolution grids of monthly climatic observations—the TS3.10 Dataset. Int. J. Climatol. 2014, 34, 62–642. [CrossRef] 26. Scenario Networks for Alaska and Arctic Planning, University of Alaska. 2016. Available online: https: //www.snap.uaf.edu/methods/downscaling (accessed on 24 January 2017). 27. Hay, L.E.; Wilby, R.L.; Leavesley, G.H. A comparison of delta change and downscaled GCS scenarios for three mountain basins in the United States. J. Am. Water Res. Assoc. 2000, 35, 387–397. [CrossRef] 28. Hayhoe, K.A. A Standardized Framework for Evaluating the Skill of Regional Climate Downscaled Techniques. Ph.D. Thesis, University of Illinois, Urbana, IL, USA, 2010. 29. PRISM Climate Group, Oregon State University, Climate Normal Obtained. Available online: http://www. prism.oregonstate.edu (accessed on 24 January 2017). 30. Daly, C.; Halbleib, M.; Smith, J.I.; Gibson, W.P.; Doggett, M.K.; Taylor, G.H.; Curtis, J.; Pasteris, P.P. Pysiographically sensitive mapping of climatological temperature and precipitation across the conterminous United States. Int. J. Climatol. 2008, 28, 2031–2064. [CrossRef] 31. Simpson, J.J.; Hufford, G.L.; Daly, C.; Berg, J.S.; Fleming, M.D. Comparing maps of mean monthly surface temperature and precipitation for Alaska and adjacent areas of Canada produced by two different methods. Arctic 2005, 58, 137–161. [CrossRef] 32. Dozier, J. Spectral signature of alpine snow cover from the Landsat Thematic Mapper. Remote Sens. Environ. 1989, 28, 9–22. [CrossRef] 33. Painter, T.H.; Dozier, J.; Roberts, D.A.; Davis, R.E.; Green, R.O. Retrieval of subpixel snow-covered area and grain size from imaging spectrometer data. Remote Sens. Environ. 2003, 85, 64–77. [CrossRef] 34. Painter, T.H.; Rittger, K.; McKenzie, C.; Slaughter, P.; Davis, R.E.; Dozier, J. Retrieval of subpixel snow covered area, grain size, and albedo from MODIS. Remote Sens. Environ. 2009, 113, 868–879. [CrossRef] 35. Rittger, K.; Painter, T.H.; Dozier, J. Assessment of methods for mapping snow cover from MODIS. Adv. Water Res. 2013, 51, 367–380. [CrossRef] 36. Raliegh, M.S.; Rittger, K.; Moore, C.E.; Henn, B.; Lutz, J.A.; Lundquist, J.D. Ground-based testing of MODIDS fractional snow cover in subalpine meadows and forests of the Sierra Nevada. Remote Sens. Environ. 2013, 128, 44–57. [CrossRef] 37. Van Soest, P.J. Nutritional Ecology of the Ruminant; Cornell University Press: Corvallis, OR, USA, 1982; p. 374. 38. Fox, J. Forage Quality of Carex macrochaeta Emerging from Alaskan Alpine Snowbanks through the Summer. Am. Midl. Nat. 1991, 126, 287–293. [CrossRef] 39. White, R.G. Foraging patterns and their multiplier effects on productivity of northern ungulates. Oikos 1983, 40, 377–384. [CrossRef] 40. Seip, D.R.; Bunnell, F.L. Nutrition of Stone’s sheep on burned and unburned rages. J. Wildl. Manag. 1985, 49, 397–405. [CrossRef] 41. Lohuis, T. Ewe Dall Sheep Survival, Pregnancy and Partition Rates, and Lamb Recruitment in GMU 13D, Chugach Mountains, Alaska. Alaska Department of Fish and Game Annual Progress Report, AKW-4 Wildlife Restoration, FY2015, 10pp. Available online: https://www.adfg.alaska.gov/static/home/library/pdfs/ wildlife/research_pdfs/6.16_10.pdf (accessed on 24 January 2017). 42. Rattenbury, K.; Schmidt, J.; Phillips, L.; Arthur, S.; Borg, B.; Burch, J.; Joly, K.; Lawler, J.; Mangipane, B.; Putera, J. Recent trends in Dall’s sheep populations in Alaska’s National Parks and Preserves. In Proceedings of the Alaska Dall Sheep Working Group Meeting, Anchorage, AK, USA, 20–21 February 2016; Available online: https://www.adfg.alaska.gov/static/regulations/regprocess/gameboard/sheepcomm/ pdfs/20160220/6nps.pdf (accessed on 24 January 2017). 43. Nippert, J.B.; Knapp, A.K.; Briggs, J.M. Intra-annual rainfall variability and grassland productivity: Can the past predict the future? Plant Ecol. 2006, 184, 65–74. [CrossRef] 44. Hansen, B.B.; Grøtan, V.; Aanes, R.; Sæther, B.-E.; Stien, A.; Fuglei, E.; Ims, R.A.; Yoccoz, N.G.; Pedersen, Å.Ø. Climate events synchronize the dynamics of a resident vertebrate community in the high arctic. Science 2013, 339, 313–315. [CrossRef][PubMed] Remote Sens. 2017, 9, 1157 18 of 18

45. IPCC. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change; Stocker, T.F., Qin, D., Plattner, G.-K., Tignor, M., Allen, S.K., Boschung, J., Nauels, A., Xia, Y., Bex, V., Midgley, P.M., Eds.; Cambridge University Press: Cambridge, UK, 2013. 46. Hoefs, M. Twinning in Dall sheep. Can. Field-Nat. 1978, 92, 292–293. 47. Bunnell, F.L. Factors controlling lambing period of Dall’s sheep. Can. J. Zool. 1980, 58, 1027–1031. [CrossRef] [PubMed] 48. Rachlow, J.L.; Bowyer, R.T. Variability in maternal behavior by Dall’s Sheep: Environmental tracking or adaptive strategy. J. Mammal. 1994, 75, 323–337. [CrossRef] 49. Van de Kerk, M.; Verbyla, D.; Nolin, A.W.; Sivy, K.J.; Prugh, L.R. Adverse effects of snow cover on Dall sheep populations increase with latitude. J. Biogeogr. 2016. submitted.

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).