|Title||Validation of the hydrological cycle of ERA-40|
|Series/Collection||ERA-40 Project Report Series|
|Authors||Hagemann, S, Arpe, K, Bengtsson, L|
|Event Series/Collection||ERA Report|
|Place of publication||Shinfield Park, Reading|
ECMWF has prepared a new 40 year reanalysis dataset (ERA40). Based on the observational data that were used, the whole ERA40 time period 1958-2001 can be divided into three parts: the satellite period 1989-2001, the pre-satellite period 1958-1972 and the transition period 1973-1988. These three periods correspond also to the three streams produced during the ERA40 production timeframe. The ERA40 dataset is expected to be a major dataset for climate research. Within the ERA40 project, the MPI had the task to perform a validation of the hydrological cycle. Here, mainly the 6 hour forecasts were considered. The validation shows that the ERA40 hydrological cycle has changed in several respects compared to the previous ERA15 re-analysis, whilst the hydrological cycle over land is generally improved. These improvements comprise the eliminated cold biases in winter, the reduced occurrence of negative P-E values, the removed dry bias in winter over Europe, and an improved representation of the snowpack. But the ERA40 hydrological cycle also has several deficiencies. The global water budget is not only unbalanced, but also P-E over the ocean is positive (and not negative as it should be) in the long term mean for the satellite and transition periods. This is related to an overestimation of precipitation over the ocean, especially in the tropics. The evapotranspiration over land is overestimated for many catchments, and, thus, the corresponding P-E is often underestimated. Using a simplified land surface scheme it was possible to derive improved values of evapotranspiration and runoff from ERA40 precipitation and 2m temperature that are consistent with the ERA40 data. The quality of the hydrological cycle differs between the periods as the biases in the hydrological cycle are strongly influenced by the different observing systems available in the three periods. Therefore, conclusions drawn for hydrological trends should be taken with great care.