USING GEOSYNCHRONOUS SATELLITE IMAGERY TO ESTIMATE SUMMER-SEASON RAINFALL OVER THE GREAT-LAKES

The quality of Great Lakes hydrologic monitoring and forecasting depends on how accurately components of the water cycle are estimated Precipitation is the most important factor, but there is no way to directly measure the rain that falls over the lakes. This is a serious problem since the water area represents one-third of the drainage basin. Here, satellite-based rain estimates were combined with min gauge measurements to estimate summer lake rainfall. Hourly estimates were made from GOES imagery for the summers of 1988, 1989, and 1990, and accumulated over monthly and summer-seasons. Area-averaged gauge-measured rainfall from local rain gauge networks were used to adjust the satellite-based estimates. Resulting best estimates of mean-summer rainfall for Lakes Michigan, Superior, and Huron were 22.66, 23.83, and 18.80 cm, respectively. These mean-summer values were 1 to 5% different from those computed using the shoreline-gauge technique, which is the current operational method for estimating lake rainfall. Estimates for individual summers were 1 to 17% different from those computed using shoreline gauges. Considering the gauge measurements as the standard, the mean absolute error of monthly gauge-adjusted lake rainfall was 21%, whereas the mean monthly error was only 2%. These results suggest that gauge-adjusted satellite estimates of rainfall could provide a useful measure of summer lake rainfall that may improve input to Great Lakes water balance and hydrologic prediction models. Other analyses were used to quantify the effects of the lakes on summer rainfall. Generally, they revealed lake-induced rain minima over and downwind of the lakes.