PREDICTION OF SNOWMELT RATES IN A DECIDUOUS FOREST

A model is presented for the prediction of the energy available for the melting of snow in a deciduous forest. The turbulent exchanges H and LE can be estimated using physically based equations, but are small at the forest floor, accounting typically for less than 10% of daily melt totals, largely because of the much reduced windspeed within the forest. Net allwave radiative energy flux density (Q*) alone is a good predictor of snowmelt rates. Physical modelling of Q* is not operationally possible, and the simple regression of Q* on solar energy flux (K ARROW DN ), a technique used with success in some circumstances, is shown to be inappropriate in a forest during snowmelt. A multivariate model using two commonly available variables explains over 3/4 of the variance of Q*.