Impacts of climate change on agriculture have been predominantly analyzed by using biophysical and crop specific model applications. Vulnerability assessments which identify the vulnerability of regions with their farming systems are urgently required, because agricultural adaptations to climate change are related to regional specifics, and therefore research has to consider the regional level. Therefore sector- and system-specific approaches have to be developed. This paper presents the methodology of a vulnerability assessment for organic farming systems in the Brandenburg Region, which considers regional-specific climatic impact, as well as the regional-specific adaptive capacity. In this region, the cultivation and management of legume-grass swards have a key position, especially the climate change impact on legume symbiotic nitrogen fixation and nitrogen mineralization. Adaptation strategies of crop production systems include reduced soil tillage, which plays an important role also in organic farming systems (reducing soil erosion, improving water infiltration, reducing evaporation and improving soil structure, control of N-dynamics) are developed and tested by means of an action research approach.
A statistical downscaling method has been developed to produce highly resolved precipitation data from regional climate model (RCM) output, using the model CLM (2 runs, scenario A1B). The procedure is based on the analogue method with the predictors precipitation (daily sums on CLM grid points) and objective weather types (DWD). Analogue days of the time period 2001-2009 are searched using corrected and adjusted data of radar Essen and DWD measurements of objective weather types. The radar data is used to produce high-resolution precipitation data sets (1km², 5min) with realistic spatial and temporal correlations for three catchments in North Rhine-Westphalia. Results in
the reference period (1961 - 1990) are examined using extreme value statistics and compared to corrected station data. Data sets of the near and the far future (2021-2050, 2071-2100) are analysed with respect to future trends, and uncertainties of the downscaling procedure are discussed.
Rainfall statistics are composed based on data gained by precipitation measurements and from climate models. These statistics are carried out for both periods in the past and the future. When analysing the time series, different trends can be seen in the measured data of the past and the model data for future periods. Influences on the statistically determined precipitation amounts caused by changes can be neglected for past periods. However, significant increases of the statistical precipitation amounts can be observed for the future. Here a pragmatic approach is presented, showing how to consider possible increases in the statistical precipitation amounts – due to the climate change signal – in the dimensioning of water management systems.
The precipitation data of the Regional Climate Model CLM are used for the water management impact models within the dynaklim networking and research project. For this purpose, it is necessary to apply a bias correction to the CLM
precipitation data. First, the bias assessed for varying temporal resolutions and precipitation characteristics is described. Subsequently, a method for the bias correction is introduced. The developed methodology is a modified form of the socalled
quantile mapping. The focus lies on the corrections of the dry days and the heavy rainfall events. They are considered separately, deviating from other quantile mapping procedures.