The impact of non-spherical particle shape on snow parameters retrieved from satellite data

In current snow remote sensing retrieval algorithms, snow is assumed to consist of spherical particles, which makes it easy to compute inherent optical properties (IOPs) from snow grain size and impurity concentrations using a Mie code. Since snow grains are not spherical in shape, they should be represented by a non-spherical particle model in order to accurately describe radiative transfer in snow. This approach would allow us to test the impact of non-sphericity on snow remote sensing retrievals. By assuming ellipsoidal particle shapes a ray-tracing technique was used to calculate the snow IOPs, and a coupled snow-atmosphere radiative transfer model was used to simulate radiances that would be measured by a satellite sensor. A snow-atmosphere retrieval algorithm, developed for the GCOM-C1/SGLI sensor, provides simultaneous retrieval of aerosol and snow parameters from a 2-layer snow model. The retrieval parameters include aerosol optical depth, snow impurity concentration, and snow grain size in the two snow layers. MODIS data obtained over the Arctic, will be used to compare snow parameters retrieved from both the spherical and non-spherical particle models. The results will be used to quantity the error incurred by assuming a spherical particle shape in snow remote sensing retrievals.