To support politicians drawing decisions on national or regional level with respect to climatological issues the downscaling of global climate predictions to the regional scale is essential. This downscaling is important for forecasts on all time scales and, thus, also on the decadal time scale. The assessment of the reliability of the forecasts and the probability of the occurrence of the calculated future climatic states are further important issues for the decision makers.
The main goal of the project LACEPS (within Module C of the MiKlip programme) is the development of a limited-area ensemble prediction system for medium-range climate on the regional scale using the COSMO-CLM regional climate model. This ensemble prediction system aims to span the complete space of all potential climate states responding to the incomplete knowledge and uncertainties of the model’s input data and the model’s inherent uncertainties. From the resulting ensemble probability density functions for each climate variable may be derived. For an optimum use of computing resources the number of ensemble members needs to be reduced to a robust subset of those approximating the probability density function of all forecasts best.
For the diagnostics of the LACEPS forecast quality it is planned to envisage the air temperature in 2 m height above ground and the total precipitation at the surface. These parameters are selected because they both have a high impact on economy, agriculture, policy, and society as well as on the daily life of the citizens.