Using snow radiative transfer modeling to quantify surface melt over the Larsen C from passive microwave observations

Surface melt is an important factor for surface mass balance and stability of the Larsen C ice shelf. Since the 1990s, researchers have been tracking frequency of surface melt days by identifying anomalously high brightness temperatures. With recent improvements in both snow microwave radiative transfer modeling and polar firn modeling, we can start to quantify these spikes in brightness temperature in order to calculate snow liquid water content (LWC) using a dual frequency approach (18V and 36V). For proof of concept, we show that the Snow Microwave Radiative Transfer model (SMRT), when forced with firn stratigraphy output from the Community Firn Model, can accurately model brightness temperatures at these frequencies for non-melt days when compared to coincident Advanced Microwave Scanning Radiometer-2 (AMSR-2). Then, we present a sensitivity study relating brightness temperature to LWC to 1) establish continuity and uniqueness of the relationship between these two parameters, and 2) determine the maximum LWC to which brightness temperature at each frequency would be sensitive. Finally, using two knowns, brightness temperature in 18V and 36V, we can estimate two unknowns, LWC at two depths (~0.25 m, ~1 m). On melt days, we run SMRT over many iterations, testing combinations of realistic LWC at both layers. We compare the resulting modeled brightness temperatures to AMSR-2 observations, minimize the difference between the two, and select the associated two-layer LWC profile that results in a closest match between SMRT and AMSR-2. We repeat this methodology over areas with detectable surface melt to create a lower bound for surface melt across the Larsen C.