Decadal prediction of European soil moisture from 1961 to 2010 using a regional climate model

Abstract:

The German national research program on decadal climate prediction
(MiKlip) aims at the development of an operational decadal prediction
system. To explore the potential of decadal predictions a hindcast
ensemble from 1961 to 2010 has been generated by the MPI-ESM, the new
Earth system model of the Max Planck Institute for Meteorology. To
improve the decadal predictions on higher spatial resolutions we
downscaled the MPI-ESM simulations by the regional model COSMO-CLM
(CCLM) for Europe. In this study we will characterize and validate
the predictability of extreme states of soil moisture in Europe
simulated by the MPI-ESM and the value added by the CCLM. The water
amount stored in the soil is a crucial component of the climate system
and especially important for agriculture, and has an influence on
evaporation, groundwater and runoff. Thus, skillful prediction of soil
moisture in the order of years up to a decade could be used to
mitigate risk and benefit society. Since soil moisture observations
are rare and validation of model output is difficult, we will rather
investigate the effective drought index (EDI), which can be retrieved
solely from precipitation data. Therefore we show that the EDI is a
good estimator of the soil water content.