Predicting snow course snow water equilavent

A procedure for predicting future snow course snow water equivalent from current data and the statistical characteristics of monthly historical data is presented. The method depends on the autocorrelation of snow water equivalent between months and includes a prediction error term which is determined for exceedance probability level of 10 through 90 percent. The algorithm was developed and tested using 44 years of January through April snow course data from a site in the Cascade Mountains of Oregon. Testing of the procedure produced acceptable results for high, average, and low years and for all exceedance levels. Results can be used to modify streamflow forecasts based on the probabilities of additional snow accumulation.