ENERGY BALANCE SNOWMELT MODELING FOR DATA-POOR BASINS

Few investigations have been made intomodeling snowmeltin data-poor basins; hence, a degree-day method iswidely used and is routinely justified under the auspices that energy-balance models require too many input data. To test this claim, we investigated the utility of merely adopting a full energy approach to model snowmelt. This study first developed so-called “full energy balance snow model” to simulate snowmelt at four sites located in Japan and the USA. The results showed very good agreement between observed and predicted snow water equivalent, R2>0.95. We duplicated the simulations using the approximated version of the model that requires only air temperature and wind speed as input data. Although the original model corresponds better, the performance of its simplified version can be evaluated as good, R2>0.9. These results provide significant information for the development of appropriate approximations in energy balance snowmelt modeling.