Abstract. Simulations with a hydrological model for the river Rhine for the present (1960–1989) and a projected future (2070–2099) climate are discussed. The hydrological model (RhineFlow) is driven by meteorological data from a 90-years (ensemble of three 30-years) simulation with the HadRM3H regional climate model for both present-day and future climate (A2 emission scenario). Simulation of present-day discharges is realistic provided that (1) the HadRM3H temperature and precipitation are corrected for biases, and (2) the potential evapotranspiration is derived from temperature only. Different methods are used to simulate discharges for the future climate: one is based on the direct model output of the future climate run (direct approach), while the other is based on perturbation of the present-day HadRM3H time series (delta approach). Both methods predict a similar response in the mean annual discharge, an increase of 30% in winter and a decrease of 40% in summer. However, predictions of extreme flows differ significantly, with increases of 10% in flows with a return period of 100 years in the direct approach and approximately 30% in the delta approach. A bootstrap method is used to estimate the uncertainties related to the sample size (number of years simulated) in predicting changes in extreme flows.

Weather radars give quantitative precipitation estimates over large areas with high spatial and temporal resolutions not achieved by conventional rain gauge networks. Therefore, the derivation and analysis of a radar-based precipitation ''climatology'' are highly relevant. For that purpose, radar reflectivity data were obtained from two C-band Doppler weather radars covering the land surface of the Netherlands ('3.55 3 10 4 km 2 ). From these reflectivities, 10 yr of radar rainfall depths were constructed for durations D of 1, 2, 4, 8, 12, and 24 h with a spatial resolution of 2.4 km and a data availability of approximately 80%. Different methods are compared for adjusting the bias in the radar precipitation depths. Using a dense manual gauge network, a vertical profile of reflectivity (VPR) and a spatial adjustment are applied separately to 24-h (0800-0800 UTC) unadjusted radar-based precipitation depths. Further, an automatic rain gauge network is employed to perform a mean-field bias adjustment to unadjusted 1-h rainfall depths. A new adjustment method is developed (referred to as MFBS) that combines the hourly mean-field bias adjustment and the daily spatial adjustment methods. The record of VPR gradients, obtained from the VPR adjustment, reveals a seasonal cycle that can be related to the type of precipitation. A verification with automatic (D # 24 h) and manual (D 5 24 h) rain gauge networks demonstrates that the adjustments remove the systematic underestimation of precipitation by radar. The MFBS adjustment gives the best verification results and reduces the residual (radar minus rain gauge depth) standard deviation considerably. The adjusted radar dataset is used to obtain exceedance probabilities, maximum rainfall depths, mean annual rainfall frequencies, and spatial correlations. Such a radar rainfall climatology is potentially valuable for the improvement of rainfall parameterization in weather and climate models and the design of hydraulic structures.

No abstract

No abstract

A stochastic weather generator has been developed to simulate long daily sequences of areal rainfall and station temperature for the Belgian and French subbasins of the River Meuse. The weather generator is based on the principle of nearestneighbour resampling. In this method rainfall and temperature data are sampled simultaneously from multiple historical records with replacement such that the temporal and spatial correlations are well preserved. Particular emphasis is given to the use of a small number of long station records in the resampling algorithm. The distribution of the 10-day winter maxima of basin-average rainfall is quite well reproduced. The generated sequences were used as input for hydrological simulations with the semidistributed HBV rainfall-runoff model. Though this model is capable of reproducing the flood peaks of December 1993 and January 1995, it tends to underestimate the less extreme daily peak discharges. This underestimation does not show up in the 10-day average discharges. The hydrological simulations with the generated daily rainfall and temperature data reproduce the distribution of the winter maxima of the 10-day average discharges well. Resampling based on long station records leads to lower rainfall and discharge extremes than resampling from the data over a shorter period for which areal rainfall was available.