The representation of soil, vegetation, snow, mountains, and water bodies is an integral part of the numerical weather prediction system at ECMWF. Land can affect the weather, the magnitude of the weather effects, and the evolution of human activities.

The effects of land surface state anomalies persist for several days and so raise the importance of a correct initial condition and modelled evolution. A refined representation of land surface processes and their accurate initialisation hold potential for further improvements of weather prediction up to monthly range as indicated in predictability studies.

Research on land surface state therefore spans the areas of

  • parametrizations of biospheric and hydrological processes;
  • the analysis techniques for the assimilation of satellite and in-situ observations;
  • the land applications, including flood- and drought-monitoring and prediction, natural carbon fluxes, and fires.