Inversion of snow parameters from passive microwave remote sensing measurements by a neural network trained with a multiple scattering model

Simultaneous inversion of the three parameters was performed which included mean-grain size of ice particles in snow, snow density, and snow temperatures from five brightness temperatures. Good results for the inversion of parameters were obtained using the neural network based on the simulated data computed from the dense media radiative transfer equation that takes into account the effects of multiple scattering.