VALUE is an open European collaboration to intercompare downscaling approaches for climate change research, focusing on different validation aspects (marginal, temporal, extremes, spatial, process-based, etc.). Here we describe the participating methods and first results from the first experiment, using "perfect" reanalysis (and reanalysis-driven regional climate model (RCM)) predictors to assess the intrinsic performance of the methods for downscaling precipitation and temperatures over a set of 86 stations representative of the main climatic regions in Europe. This study constitutes the largest and most comprehensive to date intercomparison of statistical downscaling methods, covering the three common downscaling approaches (perfect prognosis, model output statistics-including bias correction-and weather generators) with a total of over 50 downscaling methods representative of the most common techniques. Overall, most of the downscaling methods greatly improve (reanalysis or RCM) raw model biases and no approach or technique seems to be superior in general, because there is a large method-to-method variability. The main factors most influencing the results are the seasonal calibration of the methods (e.g., using a moving window) and their stochastic nature. The particular predictors used also play an important role in cases where the comparison was possible, both for the validation results and for the strength of the predictor-predictand link, indicating the local variability explained. However, the present study cannot give a conclusive assessment of the skill of the methods to simulate regional future climates, and further experiments will be soon performed in the framework of the EURO-CORDEX initiative (where VALUE activities have merged and follow on). Finally, research transparency and reproducibility has been a major concern and substantive steps have been taken. In particular, the necessary data to run the experiments are provided at and data and validation results are available from the VALUE validation portal for further investigation: .

ABSTRACTClimate information provided by global or regional climate models (RCMs) are often too coarse and prone to substantial biases for local assessments or use in impact models. Hence, statistical downscaling becomes necessary. For the Swiss National Climate Change Initiative (CH2011), a delta‐change approach was used to provide daily climate scenarios at the local scale. Here, we analyse a Richardson‐type weather generator (WG) as an alternative method to downscale daily precipitation, minimum and maximum temperature. The WG is calibrated for 26 Swiss stations and the reference period is 1980–2009. It is perturbed with change factors derived from RCMs (ENSEMBLES) to represent the climate of 2070–2099 assuming the SRES A1B emission scenario. The WG can be run in multi‐site mode, making it especially attractive for impact‐modellers that rely on a realistic spatial structure in downscaled time‐series. The results from the WG are benchmarked against the original delta‐change approach that applies mean additive or multiplicative adjustments to the observations.According to both downscaling methods, the results reveal mean temperature increases and a precipitation decrease in summer, consistent with earlier studies. For the summer drying, the WG indicates primarily a decrease in wet‐day frequency and correspondingly an increase in mean dry spell length of between 18 and 40% at low‐elevation stations. By definition, these potential changes cannot be represented by a delta‐change approach. In winter, both methods project a shortening of the frost period (−30 to −60 days) and a decrease of snow days (−20 to −100%). The WG demonstrates though, that almost present‐day conditions in snow‐days could still occur in the future. As expected, both methods have difficulties in representing extremes. If users focus on changes in temporal sequences and need a large number of future realizations, it is recommended to use data from a WG instead of a delta‐change approach.

14 Many climate impact assessments require high-resolution precipitation time-series that have a 15 spatio-temporal correlation structure consistent with observations, for simulating either 16 current or future climate conditions. In this respect, weather generators (WGs) designed and 17 calibrated for multiple sites are an appealing statistical downscaling technique to 18 stochastically simulate multiple realizations of possible future time-series consistent with the 19 local precipitation characteristics and its expected future changes. In this study, we present the 20 implementation and validation of a multi-site daily precipitation generator following ideas of 21 Wilks (1998). The generator consists of several Richardson-type WGs run with spatially 22 correlated random number streams. We investigate the applicability of the generator for the 23 current climate by analysing systematic biases and stochastically generated variability and 24 assess the added value of a multi-site generator compared to multiple single-site WGs. Results 25 are presented for the Swiss hydrological catchment Thur in the Swiss Alpine region for 26 current climate condition. 27 The calibrated multi-site WG is skilful at individual sites in representing the annual cycle of 28 the precipitation statistics, such as mean wet day frequency and intensity as well as monthly 29 2 precipitation sums. It reproduces realistically the multi-day statistics such as the frequencies 1 of dry and wet spell lengths and precipitation sums over consecutive wet days. Substantial 2 added value is demonstrated in simulating daily areal precipitation sums in comparison to 3 multiple WGs that lack the spatial dependency in the stochastic process. Limitations are seen 4 in reproducing daily and multi-day extreme precipitation sums, observed variability from year 5 to year and in reproducing long dry spell lengths. Given the performance of the presented 6 generator, we conclude that it is a useful tool to generate precipitation series consistent with 7 the mean aspects of the current and future climate. 8

Many climate impact assessments require high-resolution precipitation time series that have a spatio-temporal correlation structure consistent with observations, for simulating either current or future climate conditions. In this respect, weather generators (WGs) designed and calibrated for multiple sites are an appealing statistical downscaling technique to stochastically simulate multiple realisations of possible future time series consistent with the local precipitation characteristics and their expected future changes. In this study, we present the implementation and validation of a multi-site daily precipitation generator re-built after the methodology described in Wilks (1998). The generator consists of several Richardson-type WGs run with spatially correlated random number streams. This study aims at investigating the capabilities, the added value and the limitations of the precipitation generator for a typical Alpine river catchment in the Swiss Alpine region under current climate. The calibrated multi-site WG is skilful at individual sites in representing the annual cycle of the precipitation statistics, such as mean wet day frequency and intensity as well as monthly precipitation sums. It reproduces realistically the multi-day statistics such as the frequencies of dry and wet spell lengths and precipitation sums over consecutive wet days. Substantial added value is demonstrated in simulating daily areal precipitation sums in comparison to multiple WGs that lack the spatial dependency in the stochastic process. Limitations are seen in reproducing daily and multi-day extreme precipitation sums, observed variability from year to year and in reproducing long dry spell lengths. Given the performance of the presented generator, we conclude that it is a useful tool to generate precipitation series consistent with the mean climatic aspects and likely helpful to be used as a downscaling technique for climate change scenarios.

ABSTRACTFundamental changes in the hydrological cycle are to be expected in a future warmer climate. For Switzerland, recent climate change assessments based on the ENSEMBLES regional climate models project for the A1B emission scenario summer mean precipitation to significantly decrease by the end of this century, whereas winter mean precipitation tend to rise in Southern Switzerland. From an end‐user perspective, projected changes in seasonal means are often insufficient to adequately address the multifaceted challenges of climate change adaptation. In this study, we investigate the projected changes in seasonal precipitation by considering changes in frequency and intensity, precipitation type (convective vs stratiform) and temporal structure (wet and dry spells) over Switzerland. As proxies for rain‐type changes, we rely on the parameterized convective and large‐scale precipitation components simulated by the models. The study reveals that the projected summer drying over Switzerland at the end of the century is mainly driven by a widespread reduction in the number of precipitation days. Thereby, the drying evolves altitude‐specific: over low‐land regions it is associated with a decrease in both convective and large‐scale precipitation. Over elevated regions it is primarily associated with a decline in large‐scale precipitation only, whereas convective precipitation remains at current levels. As a consequence, almost all the models project an increase in convective fraction at elevated altitudes. The decrease in the number of wet days during summer is accompanied by decreases (increases) in the number of multi‐day wet (dry) spells. This future shift in multi‐day episodes also lowers down the likelihood of short dry spell occurrence in all of the models. The models further project a higher mean precipitation intensity in spring and autumn north of the Alps, whereas a similar tendency is expected for the winter season over most of Switzerland.

are excellent suggestions for improving the current generator. Some of the suggestions require rather fundamental changes beyond the scope of this paper. Therefore, these particular issues must be tackled more thoroughly in future work. However, as suggested by the reviewer, we will discuss these aspects in the text of the revised manuscript version and put the existing approach in this context.