Validation of the hydrological cycle of ECMWF and NCEP reanalyses using the MPI hydrological discharge model

Abstract

To validate the hydrological cycle of the European Centre for Medium‐Range Weather Forecasts (ECMWF) and the National Centers for Environmental Prediction (NCEP) reanalyses in comparison with observed river discharge, a hydrological discharge model is used to compute the corresponding river discharge. The HD model requires daily time series of surface runoff and drainage from the soil as input fields. As it turned out that a direct application to the reanalyses was not possible, a simplified land surface scheme was developed to compute runoff and drainage fields from daily reanalysis values of total precipitation and 2 m temperature. These fields were then used as input to the global simulation of river discharge with the discharge model. Results show several shortcomings of the two reanalyses in representing the hydrological cycle at the land surface. The water balance is not closed, and the snowmelt is not incorporated in the runoff and drainage fields of either of the two reanalyses. In addition, the NCEP reanalysis overestimates summer precipitation and evapotranspiration for most parts of the Northern Hemisphere, while the ECMWF reanalysis underestimates 2 m temperatures in high latitudes during the winter and spring. In the monsoon region the hydrological cycle is well represented by both reanalyses, especially over India.