Reconstructing Snow Water Equivalent In The Rio Grande Headwaters Using Remotely Sensed Snow Cover Data And A Spatially Distributed Snowmelt Model

摘要

Snow covered area (SCA) observations from the Landsat Enhanced Thematic Mapper (ETM+) were used in combination with a distributed snowmelt model to estimate snow water equivalent (SWE) in the headwaters of the Rio Grande basin (3,419 km~2) - a spatial scale that is an order of magnitude greater than previous reconstruction model applications. In this reconstruction approach, modeled snowmelt over each pixel is integrated over the time of ETM+ observed snow cover to estimate SWE. Considerable differences in the magnitude of SWE were simulated during the study. Basin-wide mean SWE was 2.6 times greater in April 2001 versus 2002. Despite these climatological differences, the model adequately recovered SWE at intensive study areas (ISAs); mean absolute SWE error was 23% relative to observed SWE. Reconstruction model SWE errors were within one standard deviation of the mean observed SWE over 37 and 55% of the four 16-km~2 intensive field campaign study sites in 2001 and 2002, respectively; a result comparable to previous works at much smaller scales. A key strength of the technique is that spatially distributed SWE estimates are not dependent upon ground-based observations of SWE. Moreover, the model was relatively insensitive to the location of forcing observations relative to commonly used statistical SWE interpolation models. Hence, the reconstruction technique is a viable approach for obtaining high-resolution SWE estimates at larger scales (e.g. >1000 km~2) and in locations where detailed hydrometeorological observations are scarce.