The snowy winter of 1978 in Massachusetts allowed five opportunities to test the dependability of a CPS-9, 3.2-cm radar to determine snowfall rate and total snowfall accumulation at up to seven measuring sites within 50 km of the radar. Spaced at intervals of 0.5 -to 1-h, 166 snowfall-rate measurements yielded a correlation coefficient of 0.88. However, in correlating the total storm snowfall, the amount of radar-measured snowfall above a reference snowfall measurement site was made equal to the snowfall actually measured at this location. This calibration technique improved the storm snowfall correlation coefficient to 0.96.

Field experiments were conducted with the AFGL Mesonet and FPS-77 digital radar during snowstorms for evaluation of the use of radar to observe and predict snow. Simultaneous measurements of optical extinction coefficient, radar reflectivity, and snow depth were made, together with determinations of visual range. The measurements verify both the calibration of the visibility instruments and the relationship between visual range and extinction coefficient. A relationship found between extinction coefficient and snowfall rate compares well with relationships found by other investigators. Radar reflectivity is found to be related to extinction coefficient, with the correlation being improved by time averaging and by allowance for velocity of snowflakes from the radar beam to the ground. However, the specification errors are of the order of + or - 42 percent to + or - 65 percent, and with the small spatial and temporal variability of extinction coefficient in snow, the radar is of marginal use where observations are available from the airways network. (Author).