A New Approach to Measuring Precipitation over Snow Cover

One of the great challenges for truly global precipitation measurement is the remote sensing of precipitation over snow cover. Due to the physical limitation in the current retrieval methodology, satellite-based measurements of precipitation over snow-covered areas are unreliable and largely unavailable. In this presentation, a new satellite-based approach to the estimation of precipitation over snow cover is proposed and tested. The method is based on the principle that precipitation can be inferred by the changes in the water content of the snowpack. During the EOS era operational remote sensing of snow water equivalent is available, with similar spatial and temporal resolutions of the precipitation-sensing passive microwave sensors. With these satellite-based snow water equivalent measurements, daily precipitation amounts can be derived. We tested the method for the Northern Hemisphere for three snow-accumulation seasons, with AMSR-E snow water equivalent data, and compared with existing datasets, including CPC gauge analysis and GPCP. The new precipitation estimates captured natural- and realistic-looking storm events over largely under-instrumented regions. The spatial distribution appeared more reasonable than existing global datasets over many boreal inland areas. The results indicate this approach is feasible and promising. Besides the capability to estimate precipitation over snow cover, this new approach has the following additional advantages over the conventional methods: 1. The relationship between precipitation and the observed variable (i.e., SWE) is more direct than the conventional methods, which have to rely on scattering signals from hydrometeors (passive microwave) or cloud top brightness temperatures (infrared) to infer precipitation; 2. Temporal sampling error is small. The method will not miss any precipitation amount even if there are no instantaneous satellite overpasses during the precipitation event. The memory of the snowpack stores the precipitation information continuously; and 3. Biases in SWE do not propagate. Since precipitation is the time-derivative of SWE, constant biases or slow-varying systematic errors in SWE will not have any significant impact on the precipitation estimates. Its unique strengths and the prospects of improved snow water equivalent retrievals make the new approach a promising alternative to the conventional precipitation remote sensing methods over snow cover.