The Hydrologic Cycle and the Color of Snow

1
The Hydrologic Cycle and the Color of Snow
2
Spectral reflectance of clean snow
3
Snows spectral reflectance, with Landsat bands
4
Snow/cloud discrimination with Landsat
Bands 3 2 1 (red, green, blue)
Bands 5 4 2
5
Topics

• Spectral reflectance of snow and its variability
• Implications for energy balance of snowpack
• Remote sensing of snow-covered area and albedo
• Fractional (subpixel) snow cover and grain size
  from MODIS, every day
• Time series corrections for clouds, viewing
  geometry, and other noise
• Relationship to hydrology
• Net solar radiation on snow over season
• Available for use by others for hydrologic models

6
Measurement of snowpack energy exchange (Mammoth
Mountain)
7
Seasonal solar radiation, Mammoth Mountain
8
Snow is a collection of scattering grains
9
Snow spectral reflectance and absorption
coefficient of ice
10
Spectral solar irradiance
11
Net solar radiation
12
Spectral reflectance of dirty snow and snow with
red algae (Chlamydomonas nivalis)
13
Spectral mixture analysis, generalized

• MODSCAG spectrally mixes with range of snow
  endmembers and chooses the result with the least
  RMS error for that pixel

14
Snow-covered area in the Tokopah Basin (Kaweah
River drainage), Sierra Nevada
AVIRIS
21 May 1997
05 May 1997
18 June 1997
20 km
15
Grain size in the Tokopah Basin (Kaweah River
drainage), Sierra Nevada
21 May 1997
05 May 1997
18 June 1997
20 km
16
Spectra with MODIS land bands
17
Fractional snow-covered area from MODIS
18
Why fractional area?
19
Validation, MODIS snow-covered area
20
MODIS-derived albedo vs field measurements
21
Tuolumne Merced River basin elevations
22
Effect of MODIS view angle
23
Cloud identification (sometimes problematic over
snow)
24
Interpolate and smooth to fill missing values
25
Smoothing spline at a pixel
26
Manual measurement of SWE (snow water equivalent)
27
Typical trends (Tuolumne R drainage)
28
Snow-pillow data for Dana Meadows and Tuolumne
Meadows
29
Accumulation and ablation inferred from snow
pillow data, Tuolumne Meadows and Dana Meadows
30
Tuolumne-Merced 2005, snow-covered area vs
elevation
31
Snowmelt by elevation band June/July 2005
Tuolumne

• Progressive contributions from higher elevation
  bands w/ time.
• Little contribution from lowest or highest
  elevations.

Merced
32
Snow in Tuolumne Merced (2004 2005)

• 36 (Tuolumne) 34 (Merced) snowmelt from above
  3000m
• 13 (Tuolumne) 5 (Merced) snowmelt from below
  2100m
• Highest surface measurement at 2918m
• Lowest surface measurement at 2100m
• Depletion over 2 months 2100-2400m, 4 months
  2700-3000m
• Snow depletion maps provide a better quantitative
  basis for estimating basin-scale snow than do
  snow pillows

33

X
X 1

34
(No Transcript)
35
Applications snowmelt modeling,Marble Fork of
the Kaweah River
Snow Covered Area
net radiation gt 0
degree days gt 0
where mq Energy to water depth conversion,
0.026 cm W-1 m2 day-1 ar Convection parameter,
based on wind speed, temperature, humidity, and
roughness
36
Magnitude of snowmelt Modeled Observed snow
water equivalent
AVIRISalbedo
SWE difference, cm
Tokopah basin, Sierra Nevada
assumedalbedo
assumed w/ update
37
Acknowledgments

• Steve Warren Warren Wiscombe
• Model for the spectral albedo of snow (1980)
• NASA
• Funding on remote sensing of snow since 1977
• UC Santa Barbara
• Great place to think about snow and ice

• Former and current students on this topic
• Bert Davis, Rob Green, Danny Marks, Noah
  Molotch, Anne Nolin, Tom Painter, Karl Rittger,
  Walter Rosenthal, Jiancheng Shi