Mapping Temperature-Sensitive Snowpacks, Frequency of Warm Winters, and Winter Precipitation Variability in the Western U.S.

1
Mapping Temperature-Sensitive Snowpacks,
Frequency of Warm Winters, and Winter
Precipitation Variability in the Western U.S.

Anne Nolin Department of Geosciences Oregon State University
2
Climate Warming Impacts on Snow and Water
Resources
(From Barnett et al., 2005 Nature)
3
Fractional change in winter snowfall water
equivalent (adjusted for changes in
precipitation) for WY 1949 - 2004. Knowles et
al., 2006
Trends in measured winter temperature
Courtesy Phil Mote, UW/CIG
4
Research Goals

• Map temperature sensitive snowcover in the
  Western US
• Quantify the relative frequency of warm winters
  (recent and projected) for selected subregions
• Consider impacts on
• hydrology
• ski industry
• Explore variability of winter precipitation on a
  watershed scale

5
Model output is too coarse (10 x 12 km) for
watershed-scale hydrology Data-driven approach
can provide higher resolution
Payne et al., 2004
6
Mapping temperature sensitive snowcover

• Snow classification based on Sturm et al., 1995
• Used temperature, precipitation, and wind speed
  to define snow classes
• Original scheme used 0.5 x 0.5 degree grid
  resolution

(Data courtesy National Snow and Ice Data Center)
7
Focus Areas
Background image PRISM digital elevation
8
DATA

• PRISM gridded temperature and precipitation
  (interpolated from station data)
• Historical monthly averages for 1971-2000
• 4 km x 4 km
• MODIS Vegetation Cover Fraction (VCF) product
  (proxy for wind speed)

9
Precipitation is classified based on a
temperature threshold, Tsnow, above which all
precipitation is considered to fall as rain
0oC
0oC
(a)
Colder than 0oC
Colder than 0oC
(b)
Because this threshold temperature is somewhat
arbitrary, we use a range of temperatures in the
snow classification exercise
10

• Now
• Lets assume climate warming over the next 40-60
  years
• Using the IPCC Climate Model output for the
  Pacific Northwest, the models are in good general
  agreement that temperatures will continue to warm
  at the rate of 0.2-0.6oC per decade
• Here, we modify the transition temperature for
  warm vs. cold snow by 0.5 degree increments for a
  total warming of 2oC

11
Decision tree thresholds

• Snow vs. No Snow
• DJF Tmean -2.0 to 2.0oC, in 0.5oC increments
• Warm snow vs. cold snow
• DJF Tmean -2.0 to 0oC, in 0.5oC increments
• High precip vs. low precip
• DJF P ? 2mm/day
• Low wind vs. high wind
• Forest cover density ? 35

12
(No Transcript)
13
US Snowcover Classification
14
(No Transcript)
15
(No Transcript)
16
Sensitivity to Rain-Snow Temperature Threshold
(6.5 km3 of water)
17
Percent of Snow-Covered Area That is At-Risk

• Pacific Northwest study arealt3
• Oregon Cascades..22
• Washington Cascades..12
• Olympic Range...61

18
Total snow area 24,128 km2 At-risk snow area
7872 km2 At-risk snow percent 32 2300 - 2700
m elevation
19
Total snow area 1600 km2 At-risk snow area
640 km2 At-risk snow percent 40 2400 - 2600 m
elevation
20
What is the relative frequency of warm winters?

• First, what is a warm winter?
• Winter DJF
• Warm When at least one winter month has a mean
  temperature above the 0oC
• If Tmean LE 0oC in December and January and
  February then it is not a warm winter
• Relative Frequency
• The number of times (N) an event occurs within a
  number of N trials
• Thus, the relative frequency of an event is N/N

21
We use monthly DJF Tmean from PRISM data
(1971-2000) Evaluate relative frequency of DJF
Tmean below a threshold temperature Shift
threshold temperature upwards by increments of
0.5oC (going from -2oC to 0oC)
22
(No Transcript)
23
Nolin and Daly, 2006
24
California Ski Areas
25
Hydrologic Implications

• Temporal centroid of hydrograph will continue to
  shift to earlier date (Stewart et al., 2005)
• Snowmelt is a significant contributor to
  mountainfront groundwater recharge
• Snowmelt vs. rainfall runoff
• Occurs during season of low evapotranspiration
• How will landscape controls (geology, vegetation)
  interact with climate controls to change the
  spatial and temporal patterns of streamflow?

26
Monthly discharge for the Clear Lake, OR
watershed in two historical periods (1948-1952,
2001-2005) and a predicted future discharge
from Jefferson et al., submitted to Hydrological
Proc.
27
Winter Precipitation Variability

• For individual watersheds, how variable has
  winter precipitation been over the past 30 years?
  
• 30-year PRISM precipitation data, computed DJF
  total DJF precipitation for each year
• Preliminary analysis only (mean and variance by
  watershed)

28
Oregon Mean of aggregate December-January-Februa
ry Precipitation (1971-2000)
mean DJF aggregate ppt m
29
Oregon Mean of aggregate December-January-Februa
ry Precipitation (1971-2000) per Watershed
mean DJF aggregate ppt m per watershed
30
Oregon Variance of aggregate December-January-Fe
bruary Precipitation (1971-2000) per Watershed
variance of DJF aggregate ppt m2 per watershed
31
To summarize

• Data-driven approach is useful for sensitivity
  studies
• At risk snow represents a large proportion of
  the Oregon and southern Washington Cascades,
  Olympic range, CA Sierra Nevada, and AZ White
  Mountains
• Relative frequency of warm winters will likely
  influence lower elevation ski areas across the
  Western US
• Hydrologic impacts are already evident
• Mapping efforts such as this can help identify
  sensitive areas that need to be integrated into
  climate measurement networks

See Nolin, A. and C. Daly, 2006. Mapping
at-risk snow in the Pacific Northwest, USA, J.
Hydrometeorology, in press.