MOS

MOS, or Model Output Statistics, is a method used in weather forecasting to improve accuracy by developing statistical relationships between numerical weather prediction model outputs and observed weather data. This technique, developed by Harry R. Glahn and D.A. Lowry, helps correct for weaknesses in numerical models, translating their forecasts into specific predictions for a particular location and time. For instance, MOS can refine wind speed forecasts, even if the numerical model itself has limitations in predicting wind occurrence or speed. This statistical approach ensures consistent and improvable results, unlike older subjective forecasting methods. Numerical weather prediction models, the basis for most modern forecasts, use physics-based equations to calculate atmospheric changes, a process that has become increasingly sophisticated with the advent of computers.123