Improve microwave snow emissivity model using QCA/DMRT

An accurate microwave snow emissivity model is essential to accurate remote sensing of atmospheric and land surface parameters. Accurate land emissivity information will increase the assimilation of satellite microwave data in numerical weather prediction (NWP) models. The current Land Emissvity Model at Microwave frequencies (LEMM) by Weng et al. (2001) was used in the US Joint Center for Satellite Data Assimilation (JCSDA) community radiative transfer model (CRTM) which is now being used in several NWP models. This model works well over the most of snow-free land types and also works over snow at microwave frequencies less than 30 GHz. However, the snow emissivity at higher microwave frequencies is often underestimated from this model, especially for newly formed snow. The information of snow emissivity at frequencies above 89 GHz play an important role in microwave remote sensing and satellite data assimilation systems since frequency channels at 150 and 183 GHz are covered by Advanced Microwave Sounding Unit-B (AMSU-B), Microwave Humidity Sounder (MHS), Special Sensor Microwave Imager and Sounder (SSMIS), and Global Precipitation Measurement (GPM) Microwave Imager (GPMI). To improve the capability of simulating snow emissivity at high frequencies, a multi-layer Dense Media Radiative Transfer (DMRT) theory based on the Quasicrystalline Approximation (QCA) will be applied to develop a fast but accurate multi-layer snow emissivity model applicable for remote sensing and data assimilation communities. The QCA/DMRT model is a physical model, which takes into account the collective scattering effects of the particles by including the wave interactions among the particles. QCA/DMRT model has been validated against ground-, aircraft-, and space-based observations over limited area. It has also been validated over a wide frequency range at 18.7GHz, 36.5GHz and 89 GHz. The NASA Land Information System (LIS) and NOAA Land -Surface Model (LSM) outputs will be used as input parameters to QCA/DMRT model. An algorithm will be developed to estimate snow grain size and density required in the emissivity models using NASA LIS and NOAA LSM snow types. To speed the computation, parameterized algorithm and look up table of snow optical properties based on QCA/DMRT simulations will also be developed. The improved microwave emissivity model will benefit the usage of JPSS and GOES-R satellite data and products in weather forecast.