No abstract

Atmospheric reanalyses covering the European region are mainly available as part of relatively coarse global reanalyses. The aim of this article is to present the development and evaluation of a next generation regional reanalysis for the European CORDEX EUR-11 domain with a horizontal grid spacing of approximately 6 km. In this context, a reanalysis is understood to be an assimilation of heterogeneous observations with a physical model such as a numerical weather prediction (NWP) model. The reanalysis system presented here is based on the NWP model COSMO by the German Meteorological Service (Deutscher Wetterdienst) using a continuous nudging scheme. In order to assess the added value of data assimilation, a dynamical downscaling experiment has been conducted, i.e. an identical model set-up but without data assimilation. Both systems have been evaluated for a 1 year test period, employing standard measures such as analysis increments, biases, or log-odds ratios, as well as tests for distributional characteristics. An important aspect is the evaluation from different perspectives and with independent measurements such as satellite infrared brightness temperatures using forward operators, integrated water vapour from GPS stations, and ceilometer cloud cover. It can be shown that the reanalysis better resolves local extreme events; this is basically an effect of the higher spatio-temporal resolution, as known from dynamical downscaling approaches. However, an important criterion for regional reanalyses is the coherence with independent observations of high temporal and spatial resolution, resulting in significant improvement over dynamical downscaling. The system is intended to become operational within a year, continuously reprocessing and evaluating longer time periods. The reanalysis data are planned to become available to the research community within a year.

Precipitation forecasts from mesoscale numerical weather prediction (NWP)

No abstract

A statistical downscaling approach for extremes using censored quantile regression is presented. Conditional quantiles of station data (e.g., daily precipitation sums) in Germany are estimated by means of the large-scale circulation as represented by the NCEP reanalysis data. It is shown that a mixed discrete–continuous response variable, such as a daily precipitation sum, can be statistically modeled by a censored variable. Furthermore, a conditional quantile skill score is formulated to assess the relative gain of a quantile forecast compared with a reference forecast. Just like multiple regression for expectation values, quantile regression provides a tool to formulate a model output statistics system for extremal quantiles.