Fine-scale atmospheric modelling and predictability

Abstract
numerical modelling of small-scale atmospheric features requires an accurate description of the large- scale atmosphere, explicit representation of appropriate physical mechanisms, and assimilation of relevant meteorological data. This allows analysis and forecasting of important weather phenomena such as thunderstorms, floods, fog, orographically-induced winds, or urban weather and pollution. Depending on the situation, local weather may be more or less predictable than large-scale weather, and short-range forecasts may be harder to produce than longer-range ones.