mesoscale numerical prediction

Mesoscale numerical prediction models are used in meteorology to forecast weather phenomena of intermediate size, typically ranging from 10 to 1,000 kilometers in horizontal extent. These models are finer-scale versions of regional numerical weather prediction models. They are crucial for predicting the evolution and behavior of mesoscale structures such as fronts, clusters of thunderstorms, sea breezes, and hurricane bands.12

Numerical weather prediction (NWP) models, in general, use physics-based equations to objectively calculate changes to the weather map. This approach, pioneered by Lewis F. Richardson, forms the basis of most modern weather forecasts. While human forecasters may interpret or modify the computer model outputs, the calculations of pressure, temperature, wind, and humidity for future times are the starting point. Mesoscale models allow for more detailed predictions of weather events that occur on a smaller scale than large cyclonic storms but are larger than individual clouds.2