Evaluation of the HUT modified snow emission model over lake ice using airborne passive microwave measurements

摘要: Abstract The algorithms designed to estimate snow water equivalent (SWE) using passive microwave measurements falter in lake-rich high-latitude environments due the emission properties of ice covered lakes on low frequency measurements. Microwave models have been used simulate brightness temperatures (Tbs) for snowpack characteristics terrestrial but cannot be applied because differing subsurface emissivities and scattering matrices present ice. This paper examines performance a modified version Helsinki University Technology (HUT) model that incorporates from lake sub-ice water. Inputs HUT include collected over brackish freshwater north Inuvik, Northwest Territories, Canada April 2008, consisting (depth, density, equivalent) (thickness type). Coincident airborne radiometer at resolution 80 × 100 m were as ground-truth evaluate simulations. results indicate media are simulated best when utilizing modeled effective grain size 1 mm RMS surface roughness ice/water interface compared measured flat Fresnel reflective input. Simulations 37 GHz (vertical polarization) produce Tbs, with Root Mean Square Error (RMSE) 6.2 K 7.9 K, well Bias Errors (MBEs) −8.4 K −8.8 K sites respectively. Freshwater simulations 6.9 19 GHz H exhibited RMSE (10.53 6.15 K respectively) MBE (−5.37 8.36 K did not accurately Tb variability (R = −0.15 0.01 respectively). Over water, 6.9 GHz had poor agreement while V (6.15 K), (−4.52 K) improved relative (R = 0.47). Salinity considerations reduced errors substantially, drop 51.48 K 57.18 K polarizations respectively, 26.2 K 31.6 K, although was simulated. With 37 GHz, exhibit potential track evolution, therefore SWE through winter season.