Abstract.The link between trends in circulation patterns and trends in the flood magnitude is studied for 122 mesoscale catchments in Germany for a period of 52 years . Flood trends, significant at the 10% level, are detected for a large number of catchments. The catchments are pooled into three regions, based on flood seasonality and flood trends. Field-significant increasing trends are found for winter in Regions West and East. For summer, increasing and decreasing flood trends are detected for Regions South and East, respectively. The temporal behaviour of three flood indicators of each region is compared to atmospheric indicators derived from circulation patterns. Significantly increasing frequency and persistence of flood-prone circulation patterns intensify the flood hazard during the winter season throughout Germany. Moreover, a trend towards a reduced diversity of circulation patterns is found causing fewer patterns with longer persistence to dominate the weather over Europe. This indicates changes in the dynamics of atmospheric circulations which directly influence the flood hazard. Longer persistence of circulation patterns which in general do not favour large precipitation amounts may lead to large runoff coefficients due to soil-moistening and hence cause floods.

Abstract. In data sparse mountainous regions it is difficult to derive areal precipitation estimates. In addition, their evaluation by cross validation can be misleading if the precipitation gauges are not in representative locations in the catchment. This study aims at the evaluation of precipitation estimates in data sparse mountainous catchments. In particular, it is first tested whether monthly precipitation fields from downscaled reanalysis data can be used for interpolating gauge observations. Secondly, precipitation estimates from this and other methods are evaluated by comparing simulated and observed discharge, which has the advantage that the data are evaluated at the catchment scale. This approach is extended here in order to differentiate between errors in the overall bias and the temporal dynamics, and by taking into account different sources of uncertainties. The study area includes six headwater catchments of the Karadarya Basin in Central Asia. Generally the precipitation estimate based on monthly precipitation fields from downscaled reanalysis data showed an acceptable performance, comparable to another interpolation method using monthly precipitation fields from multi-linear regression against topographical variables. Poor performance was observed in only one catchment, probably due to mountain ridges not resolved in the model orography of the regional climate model. Using two performance criteria for the evaluation by hydrological modelling allowed a more informed differentiation between the precipitation data and showed that the precipitation data sets mostly differed in their overall bias, while the performance with respect to the temporal dynamics was similar. Our precipitation estimates in these catchments are considerably higher than those from continental-or global-scale gridded data sets. The study demonstrates large uncertainties in areal precipitation estimates in these data sparse mountainous catchments. In such regions with only very few precipitation gauges but high spatial variability of precipitation, important information for evaluating precipitation estimates may be gained by hydrological modelling and a comparison to observed discharge.

In data sparse regions, as in many mountainous catchments, it is a challenge to generate suitable precipitation input fields for hydrological modelling, as station data do not provide enough information to derive areal precipitation estimates. This study presents a method using the spatial variation of precipitation from downscaled reanalysis data for the interpolation of gauge observations. The second aim of this study is the evaluation of different precipitation estimates by hydrological modelling. Study area is the Karadarya catchment in Central Asia (11 700 km²). ERA-40 reanalysis data are downscaled with the regional climate model Weather Research and Forecasting Model (WRF). Precipitation data from gauge observations are interpolated (i) using monthly accumulated WRF precipitation data, (ii) using monthly fields from multiple linear regression against topographical variables and (iii) with the inverse distance approach. These precipitation data sets are also compared to (iv) the direct use of the precipitation output from the WRF downscaled ERA-40 data and (v) precipitation from the APHRODITE data set. Our study suggests that using monthly fields from downscaled reanalysis data can be a good approach for the interpolation of station data in data sparse mountainous regions. Compared to mean annual precipitation from continental and global scale gridded data sets our precipitation estimates for the study area are considerably higher. The introduction of a calibrated precipitation bias factor for the comparison of different precipitation estimates by hydrological modelling allows for a more informed differentiation with regard to the temporal dynamics, on the one hand, and the overall bias, on the other hand. Uncertainty and sensitivity analyses suggest that our results are robust against uncertainties in the calibration parameters, other model parameters and inputs, and the selected calibration period

Temporal changes in daily precipitation observed at more than 2300 stations in Germany during the second half of the 20th century are analysed. Compared to other studies, this analysis is based on a very high spatial density of observation locations and complete areal coverage of Germany. Changes in four precipitation characteristics are investigated: (1) total amount of seasonal and monthly precipitation, (2) mean and 95%-quantile (q95) of daily precipitation, (3) transition probabilities to quantify wet and dry spells, and (4) precipitation amounts for a 7-day event with return period 100 years. For all parameters strikingly clear trend patterns in space and time (of the year) emerged. Stations with increasing and decreasing trends are never found in direct neighbourhood, but are well separated from each other. Changes are season-and even month-specific. These clear spatial and temporal patterns are an expression of the organisation of precipitation mechanisms over Germany. These findings add a note of caution in regard to trend analyses: Spatially and temporally aggregated trend studies might not disclose the complete range of changes and might miss important details. Interestingly, the variability of daily precipitation has changed in parallel with the mean behaviour: Those regions and seasons that show an increase in mean show also an increase in standard deviation, leading to a disproportional increase in heavy precipitation. In addition, there is a tendency towards higher persistence, in particular, longer wet spells in winter, spring and autumn, and longer dry spells in summer. If these trends continue, there will be an increasing potential for floods in winter and spring, and increasing problems for water availability in summer in regions that show signs of water stress today.

Abstract. In this paper, we present two micro rain radar-based approaches to discriminate between stratiform and convective precipitation. One is based on probability density functions (PDFs) in combination with a confidence function, and the other one is an artificial neural network (ANN) classification. Both methods use the maximum radar reflectivity per profile, the maximum of the observed mean Doppler velocity per profile and the maximum of the temporal standard deviation (±15 min) of the observed mean Doppler velocity per profile from a micro rain radar (MRR). Training and testing of the algorithms were performed using a 2-year data set from the Jülich Observatory for Cloud Evolution (JOYCE). Both methods agree well, giving similar results. However, the results of the ANN are more decisive since it is also able to distinguish an inconclusive class, in turn making the stratiform and convective classes more reliable.