Forecasting during the Lake-ICE/SNOWBANDS Field Experiments

Despite improvements in numerical weather prediction models, statistical models, forecast decision trees, and forecasting rules of thumb, human interpretation of meteorological information for a particular forecast situation can still yield a forecast that is superior to ones based solely on automated output. While such time-intensive activities may not be cost effective for routine operational forecasts, they may be crucial for the success of costly field experiments. The Lake-Induced Convection Experiment (Lake-ICE) and the Snowband Dynamics Experiment (SNOWBANDS) were conducted over the Great Lakes region during the 1997/98 winter. Project forecasters consisted of members of the academic as well as the operational forecast communities. The forecasters relied on traditional operationally available data as well as project-tailored information from special project soundings and locally run mesoscale models. The forecasting activities during Lake-ICE/SNOWBANDS are a prime example of how the man-machine mix of the forecast process can contribute significantly to forecast improvements over what is available from raw model output or even using traditional operational forecast techniques.