wyoming fater equivalent variability based on...

1
7 8 Wyoming SnowFWater Equivalent Variability Based on Historical SNOTEL9 Data ;enitza ;q Voutchkova & Scott jq >iller Wyoming ,enter for {nvironmental ~ydrology and 'eophysics & ;epartment of {cosystem Science and >anagement University of Wyoming <endall |S {ast |im ;ivide jew }ork +ake ,ottonwo od ,reek Willow ,reek Spring ,reek ;ivide )lind )ull Sumq Tri ple @eak 'ranite ,reek +o omis @ark ~ams }ork Snider )asin /ndian ,reek <elley |S Salt |iver Summit )ig Sandy #pening {lkhart @ark 'S )ase ,amp 'rassy +ake Snake |iver Station ,anyon @hilli ps )ench +ewis +ake ;ivide Thumb ;ivide ,asper >tnq |eno ~ill )attle >ountain Windy @eak South )rush ,reek Sandstone |S Whiskey @ark Webber Springs Sand +ake jorth }rench ,reek ;ivide @eak #ld )attle +aprele )ro oklyn +ake >arquette #wl ,reek Timber ,reek ,old Springs Stq +awrence =ltq South @ass ~obbs @eak Townsend ,reek @arker @eak Two #cean @lateau Wolverine Sylvan |oad )earto oth +ake Younts @eak )lackwater {vening Star Sylvan +ake 'ross Ventre Summit )urroughs ,reek ,loud @eak |eservoir +ittle Warm <irwin )ear Trap >eadow ~ansen Sawmill @owder |iver @ass )ald >tnq )one Springs ;ivide Shell ,reek >iddle @owder ;ome +ake )urgess ?unction Sucker ,reek 46w 09w 06w 04w 42w 09w 42w 4Yw 44w 4Yw 64w 49w 4Yw 4%w 2Yw 68w 66w 04w 48w 48w 64w 62w 0%w 6%w 4Jw 00w 6%w 07w 00w 04w 0Yw 44w 09w 47w 0Yw 0%w 49w 6%w 09w 00w 07w 22w 07w 48w 42w 0Yw 00w 0Jw 0Jw 24w 47w 42w 26w 0Yw 07w 04w 04w 4%w 4Jw 49w 42w 64w 4Jw 07w 00w 04w 08w 4%w 49w Y ~eight 9 % 2 0 JY J9 J% ,luster dendrogram j#T{S data5 SW{max J44Yb9YJ7 Bn:97 x p:6YT BJ464bJ404O J44%O 9YJ%O 9YJ8 excludedT transformation5 centered and scaled1 |5scale {base} distance5 euclideanO |5dist {stats} clustering5 hierarchical aglomeartion5 wardq;9 methodO |5hclust {stats} uncertainty5 approximately unbiased pbvalues BwTO computed by multiscale bo otstrap resampling1 |5pvclust {pvclust} with replication ):JYLYYYO sample sizes nL:cBJJO J7O J2O J0O 9YO 97O 98 96O 94O 79TO reproducibility iseed:JY Wyoming Wyoming %L9Y6 mqaqsql j 4%% mqaqsql Y %Y 0Y km BJ7L0Y% ftT B7LY44 ftT Wyoming Sj#T{+ stations @owderbTongue ,heyenne South @latte jorth @latte WhitebYampa 'reat ;ividebUpper 'reen )ig ~orn Upper Snake )ear Upper Yellowstone >issouri ~eadwaters >issourib +ittle >issouri jiobrara jot included ,luster J ,luster 9 ,luster 7 ,luster % ,luster 8 {levation ;istance B%4O6 miT ~ydrologic subregions B~U%T Bshort recordT 7L9YY 7LYYY 9L0YY 9L2YY 9L%YY 9L9YY mean median Jst quartile 7rd quartile min max J44Y J44J J449 J447 J448 J444 9YYY J442 J446 J440 9YYJ 9YY9 9YY7 9YY% 9YY8 9YY2 9YY6 9YY0 9YY4 9YJY 9YJJ 9YJ9 9YJ7 SW{ max dataset used for cluster analysis 6Y stations year SW{ max d : maxBSW{ max T:J d 97 years colors5clusters Jb8 SW{ max t max SW{ aprqJ Sj#T{+ is automated snowpack monitoring network5 b operated by the jatural |esources ,onservation Service Bj|,STO b 67Y remote high elevation stations in JJ Western statesO b openbaccessO nearbreal time & ~7Y years of historical dataq @U|@#S{5 forecasting of water supplies in the West xf xx x = h / k 6 7 8 9 WORK FLOW download daily data n:6Y WY stations period5J464b9YJ8 exclude incomplete water years & convert units Bfeet in mmT extract SW{ signatures all stations & years check SW{ signatures for errors Bto o lowAhighO earlyAlateO negative periodsT check raw data & maintenance comments select common period for all 6Y stations hierarchical clustering SW{max variability pbvalues via multiscale bo otstrap resampling linear trends for 8 regions in WY multiannual variability 8year moving window implications & future work number of Sj#T{+ sites 6Y 2Y 8Y %Y 7Y 9Y JY Y Pressure transducer types 6Y 2Y 8Y %Y 7Y 9Y JY Y Pillow type JY Y 8 new pressure transducers Bno change in typeT new site Bn:%T or replaced pillow Bno change in typeT Jq8 @si Usbr 2YLL @illow Usbr b Validyne JYYLL Transducer b ;ruck 8YLL Transducer b Sensotec JYYLL Transducer b Sensotec unconfirmed from archives @V, )utyl >etal ~ypalon not specified = {qui pment & maintenance h6 @retreatment & selection {levation BmqaqsqlT / SW{ signatures t aprqJ t Y @ melt :t Y bt max Jd n mm s #ctq J Sepq 7Y Snowbwater equivalent BSW{T accumulation SW{ max annual peak accumulation SW{ aprqJ snow water equivalent accumulation on =pril Jst @ melt duration of snowbablation period S rate of ablation in mmAday J40Y J408 J44Y J448 9YYY 9YY8 9YJY 9YJ8 incomplete water years per cluster Y 9Y %Y 2Y 0Y JYY histogram of incomplete water years per station wy J464b9YJ8 Y J 9 7 % 8 2 6 0 JY incomplete water years BnT frequency Y 8 JY J8 9Y w ;id Wyoming snowbwater equivalent BSW{T change in the period J464b9YJ8- /s there a subbregional difference in SW{ variability in WY- What are the implications for water resources- 9 x979F=fxk significant change Blinear regressionsT s SW{ max t max t Y 82 mmAJYy Bp3YqY8T 8Y mmAJYy Bp3YqY8T 7q2 dAJYy Bp3YqJT %qJ dAJYy Bp3YqJT 7q6 dAJYy Bp3YqJT Jq7% mmAdAJYy Bp3YqY8T JqJ7 mmAdAJYy Bp3YqY8T x99fF=fxk no significant trendsO except5 @melt IJq0 dAJYy Bp3YqJT +inear trends summary mean I J S; I 9 S; b J S; b 9 S; J44Y J448 9YYY 9YY8 9YJY 9YJ8 J40Y J408 central water year Scaled kFyear moving mean of SWEmax J44Y J448 9YYY 9YY8 9YJY 9YJ8 Y 9YY %YY 2YY J08 J4Y J48 9YY 9Y8 9JY 9J8 J44Y J448 9YYY 9YY8 9YJY 9YJ8 99Y 97Y 9%Y 98Y 92Y 96Y J44Y J448 9YYY 9YY8 9YJY 9YJ8 J44Y J448 9YYY 9YY8 9YJY 9YJ8 J44Y J448 9YYY 9YY8 9YJY 9YJ8 98 7Y 78 %Y %8 8Y 88 2 0 JY J9 J% J2 J0 mm SWEmax SWEmax timing dqwqyq SWE=f timing Rate of snowFablation mmAd Duration of snowFablation days dqwqyq xf Variability summary high 9YJ8 central water year low J44Y x Yq8S; x JS; Periods with highplow SWEmax pattern brakes J st period 9 nd period Similarities between clusters SW{ max SW{ max timing SW{:Y timing rate of snowbablation snowbablation duration highest latest longest fastest lowest earliest shortest slowliest Top k @roject funding5 BJT jational Science }oundation B{@S J9Y04Y4T though the Wyoming ,enter for {nvironmental ~ydrology and 'eophysics BWy,{~'T B9T University of Wyoming #ffice of |esearch and {conomic ;evelopment Travel grants5 BJT Wyoming Women in Science and {ngineering BWW/S{T program funded by jS} 'rant K {@S J9Y04Y4 B9T ,onsortium of Universities for the =dvancement of ~ydrologic Science B,U=~S/T xx /mplications & future work >ultiannual variability of SW{ signatures j#T{S data5 subbregional means Bmean from the stations in each clusterT analysis5 centered 8byear moving average ?UST/}/,=T/#j5 We use subbregional means in order to reduce the effect of potential data quality problems for individual stationq ThenO we smo oth the interannual variability of the subbregional means with 8byear moving averageO in order to study the multiannual patterns xf Data limitations BJT Short time record5limits the ability to study longterm trendsO multiannual or decadal variabilityO and the changes in itq |egional linear trends Bsee 4T may be biased by the fewer stations up to J44Yq B9T ;ata quality5data quality evaluation is limited by the lack of or the inconsistency of existing metabdataq The equi pment maintenance comments are not standartizedq The metabdata and the daily observations are not linked in a databaseq twitter5(;enitzaV waterbresearchqinfo Future work BJT {levation control on SW{ trends and variabilityq =ll 6Y stations are at high elevations Bx9YYYmTO nevertheless there may be differences in the longterm trends and variability depending on elevationq B9T ,oupling SW{ with streamflowO preci pitation and data on vegetation disturbances for high elevation headwater watersheds to evaluate potential changes in streamflowq Understanding snowpack dynamics is important step in studying hydrological responce and water resources availability in the Westq }ocusing on subbregional differences in snowpack dynamics helps local water managementO eqgq irrigation and water supply planningq j#T{S5 BJT 96w of the stations Bn:J4T have an active backup pillow and transducer up to end of water year 9YJ81 B9T different types of instrumentation within the spatial and temporal extent of the study1 B7T errors in labelingO maintenance comments in free textO no unification of codesO not a database1 B%T annual summer maintenance includes5zero sensors and pillowsO fixing equi pment due to single damage events BbearsO falling branchesTO moving equi pmentAlocation due to wild firesO or not known reasonsO cutting vegetation around the site n:J% n:J% n:JY n:9% n:0 n:J4 jorthbWest WY5Yellowsone plateauO Wind |iverO WyomingO and Teton |anges Southb{ast WY5Sierra >adreO >edicine )owO and +aramie >ountains {ast slopes of =bsaroka and Wind |iver |anges =bsaroka |ange )ighorn >ountains SW{ max : JqJ9 P SW{ aprqJ I J8q26 | | 9 : Yq47 | pbvalue : 39q9e bJ2 ?

Upload: others

Post on 26-May-2020

1 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Wyoming Fater Equivalent Variability Based on …water-research.info/.../uploads/2016/08/POSTER_FINAL.pdfJ44Y J44J J449 J447 J448 J442 J446 J440 J444 9YYY 9YYJ 9YY9 9YY7 9YY% 9YY8

7 8

Wyoming SnowFWater Equivalent Variability Based on Historical SNOTEL9 Data;enitza ;q Voutchkova F Scott jq >iller

Wyoming ,enter for {nvironmental ~ydrology and 'eophysics F

;epartment of {cosystem Science and >anagement

University of Wyoming

<endall |S{ast |im ;ividejew }ork +ake,ottonwood ,reekWillow ,reekSpring ,reek ;ivide)lind )ull SumqTri ple @eak'ranite ,reek+oomis @ark~ams }orkSnider )asin/ndian ,reek<elley |SSalt |iver Summit)ig Sandy #pening{lkhart @ark 'S)ase ,amp'rassy +akeSnake |iver Station,anyon@hilli ps )ench+ewis +ake ;ivideThumb ;ivide,asper >tnq|eno ~ill)attle >ountainWindy @eakSouth )rush ,reekSandstone |SWhiskey @arkWebber SpringsSand +akejorth }rench ,reek;ivide @eak#ld )attle+aprele)rooklyn +ake>arquette#wl ,reekTimber ,reek,old SpringsStq +awrence =ltqSouth @ass~obbs @eakTownsend ,reek@arker @eakTwo #cean @lateauWolverineSylvan |oad)eartooth +akeYounts @eak)lackwater{vening StarSylvan +ake'ross Ventre Summit)urroughs ,reek,loud @eak |eservoir+ittle Warm<irwin)ear Trap >eadow~ansen Sawmill@owder |iver @ass)ald >tnq)one Springs ;ivideShell ,reek>iddle @owder;ome +ake)urgess ?unctionSucker ,reek

46w

09w

06w

04w

42w

09w

42w

4Yw

44w

4Yw

64w 49w4Yw

4%w

2Yw68w

66w 04w

48w

48w

64w

62w

0%w

6%w4Jw

00w

6%w07w

00w

04w0Yw

44w

09w

47w

0Yw0%w49w

6%w

09w

00w07w

22w

07w

48w

42w

0Yw

00w

0Jw

0Jw

24w

47w

42w

26w

0Yw

07w

04w

04w

4%w

4Jw

49w

42w

64w

4Jw

07w

00w04w

08w

4%w

49w

Y~eight 9 % 2 0 JY J9 J%

,luster dendrogram

j#T{S

data5 SW{max J44Yb9YJ7 Bn:97 x p:6YT

BJ464bJ404O J44%O 9YJ%O 9YJ8 excludedT

transformation5 centered and scaled1 |5scale {base}

distance5 euclideanO |5dist {stats}

clustering5 hierarchical

aglomeartion5 wardq;9 methodO |5hclust {stats}

uncertainty5 approximately unbiased pbvalues BwTO

computed by multiscale bootstrap resampling1

|5pvclust {pvclust} with replication ):JYLYYYO

sample sizes nL:cBJJO J7O J2O J0O 9YO 97O 98

96O 94O 79TO reproducibility iseed:JY

W y o m i n gW y o m i n g

%L9Y6mqaqsql

j

4%%mqaqsql Y %Y 0Y km

BJ7L0Y% ftTB7LY44 ftT

WyomingSj#T{+ stations

@owderbTongue

,heyenne

South @latte

jorth @latte

WhitebYampa

'reat ;ividebUpper 'reen

)ig ~ornUpper

Snake

)ear

Upper Yellowstone>isso

uri~eadwaters

>issourib

+ittle>issouri

jiobrara

jot included

,luster J

,luster 9

,luster 7

,luster %

,luster 8

{levation ;istance

B%4O6 miT

~ydrologic subregions B~U%T

Bshort recordT

7L9YY

7LYYY

9L0YY

9L2YY

9L%YY

9L9YY

mean

median

Jst quartile

7rd quartile

min

max

J44Y J44J J449 J447 J448

J444 9YYYJ442 J446 J440

9YYJ 9YY9 9YY7 9YY% 9YY8

9YY2 9YY6 9YY0 9YY4 9YJY 9YJJ 9YJ9 9YJ7

SW{max dataset used for cluster analysis

6Y stations

year

SW{max

d : maxBSW{maxT : J

d

97 years

colors5clusters Jb8

SW{max

tmax

SW{aprqJ

Sj#T{+ is automated snowpack monitoring network5

b operated by the jatural |esources ,onservation Service Bj|,STO

b 67Y remote high elevation stations in JJ Western statesO

b openbaccessO nearbreal time F ~7Y years of historical dataq

@U|@#S{5 forecasting of water supplies in the West

xf xxx = h / k 6 7 8 9WORK FLOW

download daily data

n:6Y WY stations

period5J464b9YJ8

exclude incomplete water

years F convert units

Bfeet in mmT

extract SW{ signatures

all stations F years

check SW{ signatures

for errors Btoo lowAhighO

earlyAlateO negative periodsT

check raw data F

maintenance comments

select common period

for all 6Y stations

hierarchical clustering

SW{max variability

pbvalues via multiscale

bootstrap resampling

linear trends for 8

regions in WY

multiannual variability

8year moving window

implications F future work

numberofSj#T{+sites 6Y

2Y

8Y

%Y

7Y

9Y

JY

YPressure transducer types

6Y

2Y

8Y

%Y

7Y

9Y

JY

YPillow type

JY

Y

8 new pressure transducersBno change in typeT

new site Bn:%T or replacedpillow Bno change in typeT

Jq8 @si Usbr

2YLL @illow Usbr b Validyne

JYYLL Transducer b ;ruck

8YLL Transducer b Sensotec

JYYLL Transducer b Sensotec

unconfirmed from archives

@V,

)utyl

>etal~ypalon

not specified

= {quipment F maintenance h 6 @retreatment F selection

{levation BmqaqsqlT

/ SW{ signatures

taprqJ tY

@melt : tY b tmax

Jd

nmms

#ctq J Sepq 7Y

Snowbwater equivalent

BSW{T accumulation

SW{max annual peak accumulation

SW{aprqJ snow water equivalent accumulation on =pril Jst

@melt duration of snowbablation period

S rate of ablation in mmAday

J40Y

J408

J44Y

J448

9YYY

9YY8

9YJY

9YJ8

incomplete water years

per cluster

Y

9Y

%Y

2Y

0Y

JYYhistogram of incomplete

water years per station

wy J464b9YJ8

Y J 9 7 % 8 2 6 0 JY

incomplete water years BnT

frequency

Y

8

JY

J8

9Y

w

;id Wyoming snowbwater equivalent BSW{T change

in the period J464b9YJ8-

/s there a subbregional difference in SW{ variability in WY-

What are the implications for water resources-

9

x979F=fxksignificant changeBlinear regressionsT

s

SW{max

tmax tY

82 mmAJYy Bp3YqY8T

8Y mmAJYy Bp3YqY8T

7q2 dAJYy Bp3YqJT

%qJ dAJYy Bp3YqJT

7q6 dAJYy Bp3YqJT

Jq7% mmAdAJYy Bp3YqY8T

JqJ7 mmAdAJYy Bp3YqY8T

x99fF=fxk no significant trendsO except5 @melt IJq0 dAJYy Bp3YqJT

+inear trends summary

mean

I J S;

I 9 S;

b J S;

b 9 S;

J44Y

J448

9YYY

9YY8

9YJY

9YJ8

J40Y

J408

central water year

Scaled kFyear moving mean of SWEmaxJ44Y

J448

9YYY

9YY8

9YJY

9YJ8

Y

9YY

%YY

2YY

J08

J4Y

J48

9YY

9Y8

9JY

9J8

J44Y

J448

9YYY

9YY8

9YJY

9YJ8

99Y

97Y

9%Y

98Y

92Y

96Y

J44Y

J448

9YYY

9YY8

9YJY

9YJ8

J44Y

J448

9YYY

9YY8

9YJY

9YJ8

J44Y

J448

9YYY

9YY8

9YJY

9YJ8

98

7Y

78

%Y

%8

8Y

88

2

0

JY

J9

J%

J2

J0

mm

SWEmax SWEmax timing

dqwqyq

SWE=f timing Rate of snowFablation

mmAd

Duration of snowFablation

daysdqwqyq

xf Variability summary

high

9YJ8

centralwateryear

low

J44Y

x Yq8S;x JS;

Periods with highplow SWEmax

pattern brakes

Jst period

9nd period

Similaritiesbetweenclusters

SW{max

SW{max timing

SW{:Y timing

rate of snowbablation

snowbablation duration

highestlatestlongestfastest

lowestearliestshortestslowliest

Top k

@roject funding5

BJT jational Science }oundation B{@S J9Y04Y4T though the Wyoming ,enter for {nvironmental ~ydrology and

'eophysics BWy,{~'T

B9T University of Wyoming #ffice of |esearch and {conomic ;evelopment

Travel grants5

BJT Wyoming Women in Science and {ngineering BWW/S{T program funded by jS} 'rant K {@S J9Y04Y4

B9T ,onsortium of Universities for the =dvancement of ~ydrologic Science B,U=~S/T

xx /mplications F future work

>ultiannual variability of SW{ signatures

j#T{S

data5 subbregional means Bmean from

the stations in each clusterT

analysis5 centered 8byear moving average

?UST/}/,=T/#j5

We use subbregional means in order to reduce

the effect of potential data quality problems

for individual stationq

ThenO we smooth the interannual variability

of the subbregional means with 8byear moving

averageO in order to study the multiannual patterns

xf

Data limitationsBJT Short time record5limits the ability to study longterm trendsO multiannual or decadal

variabilityO and the changes in itq |egional linear trends Bsee 4T may be biased by

the fewer stations up to J44Yq

B9T ;ata quality5data quality evaluation is limited by the lack of or the inconsistency of

existing metabdataq The equipment maintenance comments are not standartizedq

The metabdata and the daily observations are not linked in a databaseq

twitter5(;enitzaV waterbresearchqinfo

Future workBJT {levation control on SW{ trends and variabilityq =ll 6Y stations are at high

elevations Bx9YYYmTO nevertheless there may be differences in the longterm trends

and variability depending on elevationq

B9T ,oupling SW{ with streamflowO precipitation and data on vegetation disturbances

for high elevation headwater watersheds to evaluate potential

changes in streamflowq

Understanding snowpack dynamics is important step in studying hydrological responce

and water resources availability in the Westq

}ocusing on subbregional differences in snowpack dynamics helps local water

managementO eqgq irrigation and water supply planningq

j#T{S5

BJT 96w of the stations Bn:J4T have an active backup pillow and transducer up to end of water year 9YJ81

B9T different types of instrumentation within the spatial and temporal extent of the study1

B7T errors in labelingO maintenance comments in free textO no unification of codesO not a database1

B%T annual summer maintenance includes5zero sensors and pillowsO fixing equipment due to single damage

events BbearsO falling branchesTO moving equipmentAlocation due to wild firesO or not known reasonsO

cutting vegetation around the site

n:J%

n:J%

n:JY

n:9%

n:0

n:J4

jorthbWest W

Y5YellowsoneplateauOWind

|iverO W

yomingO and

Teton

|anges

Southb{ast WY5Sierra >adreO >edicine )owO and +aramie >ountains

{ast slopes of =bsaroka and Wind |iver |anges

=bsaroka |ange

)ighorn>ountains

SW{max : JqJ9 P SW{aprqJ I J8q26 | |9 : Yq47 | pbvalue : 39q9ebJ2

?