Snow is an important part of the water cycle, with over 1 billion people dependent on melt for their water supply. My research aims are to develop algorithms to improve global snow mass and soil moisture estimates from satellite data at microwave wavelengths. We will use data assimilation techniques to blend physically-based models with other remote sensing data, such as temperature and reflectance data at near-infrared wavelengths. Scattering of electromagnetic radiation is hugely sensitive to the size of the snow crystals, so we need to know this well in order to retrieve snow mass from the satellite data. Near-infrared reflectance, and physically-based models of the snow will give us an idea of how large the snow crystals are. Once this system has been developed, we will then apply it to a 30+ year dataset of observations to examine whether the snow mass has changed with the climate.