A NEW SNOW LIGHT SCATTERING MODEL AND ITS APPLICATION IN SNOW PARAMETER RETRIEVAL FROM SATELLITE REMOTE SENSING

Light scattering models of snow are very important for remote sensing of snow. Many previous models have used unrealistic assumptions about the snow particle shape and microstructure. In this paper, a new model is proposed, wherein a bicontinuous medium is used to simulate the snow microstructure, and geometric optics theory is used in combination with the Monte Carlo method to simulate the scattering properties of snow. Then, using the radiative transfer equation, the snow reflectance, including polarized reflectance, can be simulated. Unlike other models that use Monte Carlo ray tracing, the new model is computationally efficient, and can be used for massive simulations and practical applications. The simulation results of the new model are compared with the ground measurements and simulation results of a traditional model based on Mie theory. Through validations and comparisons, the new model is shown to demonstrate a significantly improved capability in simulating the bidirectional reflectance of snow. The importance of grain shape and microstructure modeling in light scattering models of snow is confirmed by the simulation results' comparisons. Using the new model, snow surface grain size and pollution concentration can be retrieved using MODIS land surface reflectance, the retrieved snow grain size and snow surface area is validated using ground observations.