Both regional climate models and general circulation models include time-dependent partial differential equations and algebraic equations to represent the motion and physics of the atmosphere and surface properties in the climate system (such as wind, pressure, temperature, clouds, radiation, rainfall, and soil hydrology). Most regional climate models cover an area of 10⁵–10⁷ km² (10⁴–10⁶ mi²) with a typical resolution of 10–100 km (6–60 mi), compared to a few hundred kilometers resolution for most general circulation models, which cover the global atmosphere. Integration times for regional climate models range from a month to more than 10 years. Most regional climate models use the observed sea surface temperature or the mixed-layer ocean model for the ocean. The coupled climate model system, including the regional atmospheric model, ocean–sea ice model, and hydrology model, has also been successfully applied for studying some extreme weather events in Germany.

At the lateral boundaries, a regional climate model is driven by atmospheric wind, temperature, and humidity generated from a general circulation model output or global meteorological reanalysis, which is derived from the combination of the general circulation model output and observations. A regional climate model is, in principle, similar to the operational limited-area numerical weather forecasting models. But a regional climate model is integrated continuously over time intervals much longer than those used in numerical weather forecasting are. Consequently, lateral and surface boundary conditions become dominant in a regional climate model, while limited-area numerical weather forecasting is more dependent on the initial data.