Abstract
Global precipitation monitoring is essential for understanding the earth’s water and energy cycle. Therefore, usage of satellite-based precipitation data is necessary where in situ data are rare. In addition, atmospheric-model-based reanalysis data feature global data coverage and offer a full catalog of atmospheric variables including precipitation. In this study, two model-based reanalysis products, the interim reanalysis by the European Centre for Medium-Range Weather Forecasts (ERA-Interim) and NASA’s Modern-Era Retrospective Analysis for Research and Applications (MERRA), as well as two satellite-based datasets obtained by the Global Precipitation Climatology Centre (GPCP) and Hamburg Ocean Atmosphere Parameters and Fluxes from Satellite Data (HOAPS) are evaluated. The evaluation is based on monthly precipitation in the tropical Pacific Ocean during the time period 1989–2005. Rain-gauge atoll station data provided by the Pacific Rainfall Database (PACRAIN) are used as ground-based reference. It is shown that the analyzed precipitation datasets offer temporal correlations ranging from 0.7 to 0.8 for absolute amounts and from 0.6 to 0.75 for monthly anomalies. Average monthly deviations are in the range of 20%–30%. GPCP offers the highest correlation and lowest monthly deviations with reference to PACRAIN station data. The HOAPS precipitation data perform in the range of the reanalysis precipitation datasets. In high native spatial resolution, HOAPS reveals deficiencies owing to its relatively sparse temporal coverage. This result emphasizes that temporal coverage is critical for controlling the performance of precipitation monitoring. Both reanalysis products show similar systematic behaviors in overestimating small and medium precipitation amounts and underestimating high amounts.
