Modeled Snow Cover Product Snow Water Equivalent

1
Modeled Snow Cover Product (Snow Water
Equivalent)
The snow water equivalent (SWE) model is a
spatially distributed snowmelt model driven by
remotely sensed and assimilated meteorological
data. The model is based upon the Utah Energy
Balance (UEB) snowmelt model developed by
Tarboton (Tarboton, 1994 Tarboton et al., 1995).

The (daily) snow water equivalent maps show the
spatial distribution of the modeled water content
of the snow pack. As such, the maps also portray
the spatial distribution of snow cover extent,
and provide an indication of relative snow depth
and water available for irrigation when the snow
melts.
For more information about the SWE model please
see the explanation on the FEWS NET Afghanistan
website http//edcintl.cr.usgs.gov/Afghan/readme/
SWE_Readme_0411v1.doc
2
Snow Water Equivalent Volume Estimates
Snow Water Equivalent (SWE) volume estimates were
created for selected basins above known dam
locations. Each cell of the daily SWE grids
(with units of depth) was multiplied by its area
to obtain water volume estimates. Basin-level
statistics were calculated to create tables
summing up the total volume in each of the
reservoir basins. An Access database is updated
daily with new totals for snow water volume, and
new time series charts are produced in Excel to
provide a seasonal comparison (e.g., 2005, 2004,
2003).
The charts for specific sub-basins can be
selected from a map of 16 known dam locations
throughout Afghanistan
3
Snow Water Volume Charts
The Snow Water Equivalent (SWE) model produces
uncalibrated estimates of snow water equivalent
since data for calibration are not available in
Afghanistan at the current time. Thus, snow
water volumes derived from the SWE model can be
used to make relative comparisons between
seasons, although the values should not be
regarded as absolute. The Snow Water Volume
charts are designed to provide insight into where
flooding hazards due to snowmelt might exist.
4
Afghanistan Sistan Wetlands Time Series Analysis
1976
2001
5
Afghanistans Sistan Wetlands Time Series
Analysis
A 3-year time series analysis of the Sistan
Wetlands was completed for the years 1989 through
1991 in an effort to characterize the Helmand
River flood of February 1991.
SPOT imagery was acquired for August 12, 1991,
the earliest available date after the flood
event. The SPOT images are panchromatic (single
band, black and white) with 10-m spatial
resolution. The SPOT imagery was used to train
and validate results from the much coarser
resolution 1-km AVHRR data available historically
on a daily basis.
AVHRR scenes were selected based on date and
cloud cover. One scene was selected for each
month between 1989 and 1991 for a total of 36
scenes.
6
The SPOT image was clipped to a screen-digitized
water boundary because it was more effective than
deriving an water-only classification from the
panchromatic image. 10-m pixel resolution
provides reasonable accuracy.
Areas within the original water boundary that
revealed a large variety in unique pixel values
were then considered mixed water/land pixels.
Road
The grid was then reclassified such that pure
(water or land) pixels were distinguished from
the mixed pixels. A Percent Water grid was
created by assigning each mixed pixel a water
value, determined from the ratio of land (0) to
water (1).
7

• The surface area of water from the SPOT image was
  determined by the equation
• (h2o) (cellsize)2 (pixel count),
• where h2o is the percent water for each pixel
  from the Percent Water grid.
• The calculated surface area for the Hamoun-e
  Saberi and the Daryache-ye Sistan water bodies on
  12 Aug 1991 is 2834 km2.
• The volume of the 2 lakes was determined
  according to
• ? (max. lake elevation pixel
  elevation) (cell size)2 (h2o),
• with the condition if (pixel elevation) gt (max
  lake elevation), then the depth 0.5 meters.
• The volume of the 2 lakes on 12 August 1991 was
  calculated to be 3.1269 km3 (or bcm, billion
  cubic meters).