Rainfall generator for the Rhine basin: Description of simulations using gridded precipitation datasets and uncertainty analysis

The rainfall generator has been used to generate long synthetic series of daily precipitation and temperature for the Rhine basin [see e.g. Wójcik et al. (2000), Buishand and Brandsma (2001), and Beersma (2002)] using the nearest-neighbour resampling (NNR) technique. These simulations were driven by daily precipitation and temperature data for the period 1961 – 1995 from 34 stations across the Rhine basin. For hydrological applications, the simulated point precipitation and temperature data were converted to representative values of 134 HBV1 sub-basins, using a dataset which is nowadays known as CHR-OBS data [see Görgen et al. (2010)]. Recently, two additional gridded datasets (HYRAS and E-OBS) have become available. These two datasets fulfil the need for extended temperature and precipitation record lengths.

This report is set up as follows. First, the three datasets that were used in this research are highlighted in the next subsections. An intercomparison between the three precipitation datasets is presented in section 2, followed by a description of the nearest neighbour resampling technique in section 3. Results for the various simulation types are shown in section 4, after which results from an uncertainty analysis are presented in section 5. Finally conclusions are drawn in section 6.