Meromy, Leah H听1听;听Molotch, Noah P听2听;听Link, Timothy E听3听;听Fassnacht, Steven R听4听;听Herchmer, Eric听5听;听Roberts, Scott听6听;听Rice, Robert听7
1听University of Colorado, 麻豆影院, INSTAAR
2听University of Colorado, 麻豆影院, INSTAAR, Jet Propulsion Laboratory, California Institute of Technology
3听University of Idaho
4听Colorado State University, Fort Collins
5听University of Idaho
6听Marine Science Institute, University of California, Santa Barbara
7听University of California, Merced
The spatial distribution of snow water equivalent (SWE) is a key variable in many regional-scale land surface models and operational stream flow forecasting in the western United States. Currently, assimilation of snow sensor data into these models is performed without consideration of the spatial representativeness of the point data with respect to subgrid element SWE. In order to improve the understanding of the relationship between these point measurements and the surrounding grid element, we characterized the spatial distribution of snow depth and SWE within 1- and 16-km2 grids surrounding 15 snow stations (snow telemetry (SNOTEL) and California Snow Sensors) in California, Colorado, Wyoming, Idaho, and Oregon through the 2008 and 2009 snow seasons. Field observations of snowpack properties and binary regression tree models were used to find the correlation between SWE at the sensor site and the surrounding SWE in order to evaluate the effectiveness of the sensors in representing the surrounding area鈥檚 SWE. Field campaigns included detailed observations of the distribution of snowpack properties (depth, density, grain size, temperature) within 1-km2 around the snow sensor. Unlike previous works, we did not find consistent high biases in snow sensor depth values. Rather, we found biases that ranged from 73% overestimates to 33% underestimates. Relationships between snow sensor depth and grid element depth varied greatly and were inconsistent throughout the accumulation and melt seasons, and from site to site. Snow sensor bias behavior did follow expectations regarding physiographic relationships between snow sensor location characteristics and the mean characteristics of the surrounding grid element (e.g. elevation, solar radiation, and vegetation density). The scaling relationships derived from this research have the potential to dramatically improve the National Operational Hydrologic Remote Sensing Center (NOHRSC) SWE product. In this regard, the snow sensor bias information has indicated that many assimilations of point data into the NOHRSC model were unnecessary. Results and data will be disseminated to the broader community for use in regional scale land surface modeling studies and operational stream flow forecasting.