Estimating snow water equivalent in a Sierra Nevada watershed via spaceborne radiance data assimilation

Abstract

This paper demonstrates improved retrieval of snow water equivalent (SWE) from spaceborne passive microwave measurements for the sparsely forested Upper Kern watershed (511 km2) in the southern Sierra Nevada (USA). This is accomplished by assimilating AMSR‐E 36.5 GHz measurements into model predictions of SWE at 90 m spatial resolution using the Ensemble Batch Smoother (EnBS) data assimilation framework. For each water year (WY) from 2003 to 2008, SWE was estimated for the accumulation season (1 October to 1 April) with the assimilation of the measurements in the dry snow season (1 December to 28 February). The EnBS SWE estimation was validated against snow courses and snow pillows. On average, the EnBS accumulation season SWE RMSE was 77.4 mm (13.1%, relative to peak accumulation), despite deep snow (average peak SWE of 545 mm). The prior model estimate without assimilation had an accumulation season average RMSE of 119.7 mm. After assimilation, the overall bias of the accumulation season SWE estimates was reduced by 84.2%, and the RMSE reduced by 35.4%. The assimilation also reduced the bias and the RMSE of the 1 April SWE estimates by 80.9% and 45.4%, respectively. The EnBS is expected to work well above tree line and for dry snow.

Estimating Snow Mass in North America Through Assimilation of Advanced Microwave Scanning Radiometer Brightness Temperature Observations Using the Catchment Land Surface Model and Support Vector Machines

“…In terms of space‐borne radiance data assimilation, Andreadis and Lettenmaier () showed that assimilating SSM/I observations into a one‐layer model had no impact on snow depth estimates but that assimilating these observations into a multilayer snow model can dramatically improve the model predictability. Similarly encouraging performance in snow mass estimation was also witnessed during the snow accumulation season via assimilation of AMSR‐E Tb observations in Siberia (Che et al, ) and in the Sierra Nevada (Li et al, ). In terms of continental‐scale estimates, Kwon et al () assimilated vertically polarized AMSR‐E Tb observations at 18.7 and 36.5 GHz over North America from December 2002 to February 2003 and found modest improvements in snow depth across regions of tundra (as defined in Sturm et al, ) and bare soil cover (as defined in the Community Land Model Version 4 for land without vegetation).…”

Ensemble-based assimilation of fractional snow-covered area satellite retrievals to estimate the snow distribution at Arctic sites

“…However, GRACE TWS retrievals feature a very coarse resolution (around 100 km) so that they would only be useful in conjunction with fSCA retrievals for very large scale applications. On the other hand, the use of higher-resolution PM SWE retrievals (order 25 km) in the assimilation has shown particular promise (e.g., De Lannoy et al, 2012;Li et al, 2017). At the same time, PM SWE retrievals are not accurate in complex topography and forested areas, nor for wet and deep snowpacks (Foster et al, 2005), which might limit the applicability of such multisensor assimilation approaches.…”

“…This was extended to a real multisensor experiment by jointly assimilating PM SWE and MODIS fSCA retrievals (De Lannoy et al, 2012). Li et al (2017) used the ES to assimilate PM SWE retrievals and estimate the SWE distribution, markedly outperforming the OL. Of late, particle filter (PF; see Van Leeuwen, 2009) schemes have been gaining popularity in snow DA studies (Charrois et al, 2016;Magnusson et al, 2017).…”

The Utility of Infrequent Snow Depth Images for Deriving Continuous Space‐Time Estimates of Seasonal Snow Water Equivalent

“…Despite the importance of snow (in particular its water equivalent [SWE]), it remains relatively poorly characterized, especially in mountainous regions where both high values of SWE and its spatial variability simultaneously exist. In a recent survey of the state of hydrologic remote sensing, Lettenmaier et al () concluded that SWE is the variable “that is most in need of new strategic thinking from the hydrologic community.” They note that while efforts to develop estimates of SWE using passive microwave sensors go back almost four decades, large‐scale applications of those methods are hampered by extreme complications in forested and topographically complex environments with deep snow (Cai et al, ), or for snowpacks containing substantial liquid water (Li, Durand, & Margulis, ). The need for snow estimates is echoed in the recently published National Academies Decadal Review of Earth Science and Applications from Space (i.e., “Decadal Survey”; National Academies of Sciences, Engineering, and Medicine, ), which classified among its “most important” level of objectives to “Quantify rates of snow accumulation, snowmelt, ice melt, and sublimation from snow and ice worldwide at scales driven by topographic variability.”…”