The use of precipitation and snow survey data in water supply forecasting

The use of precipitation measurements and snow survey data singly and in combinations for forecasting is discussed and applied to the North Fork Clearwater River in Idaho. An average error of 8.7% was obtained using precipitation data alone, and 4.6% when combined with snow survey data. Forecasts issued by the Weather Bureau in January, February and March will use only precipitation data but April and May forecasts will also utilize snow survey data where applicable.In discussion Mr. Clarence Pederson indicates his approval of the combination of variables in the formula. However, he warns against using a precipitation period that extends beyond the date of the forecast. Using estimated or average figures invariably leads to trouble. Also reliable accuracy checks cannot be made by inserting the values of the variables used in developing the formula. If every 3rd or 5th data is omitted, and then a reliability check is made using these omitted, the check will be much more significant.