MEaSUREs proposal
due 05 March 2007
solicitation
Final Proposal
- 2007-03-05 Dozier
- 2007-03-07 Dozier-Frew-Painter poster at SOGWC
Alpine Snow Earth Science Data Records for Water Resource Management Research
SIGNIFICANCE
Over a billion people worldwide depend on melting snow and glaciers for their water, resources that may be at risk from climate changes. Assessing and managing snow- and glacier-melt cannot, therefore, depend entirely on historical statistics, but must use models driven by remotely sensed data. Similarly, the hydrological and climate research communities need information about snow, but the processing steps to generate that information are often complicated.
Our user community of both hydrologic researchers and operational agencies includes the California Department of Water Resources, U.S. Army CRREL, University of California campuses at Merced and Irvine, National Snow and Ice Data Center, University of Washington, University of Wageningen (Netherlands), and the U.S. Agency for International Development.
PRODUCTS
We propose to provide the most relevant information about snow in the mountains that can be produced from currently operating sensors and those projected to be launched within a decade. The current sensor of choice is MODIS on the Terra and Aqua satellites, to be followed by VIIRS on NPP and NPOESS. These instruments provide global daily coverage, with wavelengths suitable, indeed designed, for measuring snow coverage and properties.
Our core products are daily subpixel snow cover and snow albedo for areas where substantial human populations derive their water supply from melting snow or glaciers. Conventional methods for estimating snow cover classify each pixel as either "snow" or "not snow." In the mountains, these methods underestimate snow at low elevations and overestimate it at high elevations, hence the need for a subpixel method, which estimates the fractional snow cover in each pixel. To use snow cover information in hydrologic models, it is also necessary to estimate its albedo, which varies substantially. Finally, because daily measurements are often occluded by cloud cover, we take advantage of daily time series to smooth and filter the information, and make it available in three data records:
- Daily fractional snow-covered area, grain size, and contaminants.
- Best current assimilated estimate of snow-covered area and broadband albedo.
- End-of-season reanalysis of daily snow-covered area and broadband albedo.
A related application product derives daily spatially variable snow water equivalent from the integration of remotely sensed information with surface measurements, via two approaches:
- Interpolating snow pillow data with fractional snow-covered area.
- Estimating snow-cover depletion with a spatially distributed energy balance model, to provide end-of-season estimates of the melting history of the snowpack.
METHODS
Our "MODSCAG" (MODIS Snow-Covered Area and Grain size) method of estimating fractional snow-covered area and snow grain size, and then deriving the albedo from the grain size, was originally developed with airborne imaging spectrometer data (AVIRIS). MODSCAG uses spectral unmixing, analyzing the spectral signal for the contribution from each of the pixel's components and then solving for the fractional coverage of each "endmember." MODSCAG's adaptation to MODIS and Landsat has been tested with AVIRIS, and the effect of spatial resolution has been tested by comparing MODIS with Landsat.
Data products are delivered through our website http://www.snow.ucsb.edu. Initially, we will provide geoTIFFs of daily snow-covered area and albedo. In consultation with our user community, we will explore more flexible ways of delivering space-time data "cubes", for example via services like OPeNDAP.
Dozier-SOGWC-TH14.pdf