Peter Berg

Dynamical Downscaling with Reinitializations: A Method to Generate Finescale Climate Datasets Suitable for Impact Studies

To retain the sequence of events of a regional climate model (RCM) simulation driven by a reanalysis, a method that has not been widely adopted uses an RCM with frequent reinitializations toward it ...

High-Resolution Climate Change Impact Analysis on Medium-Sized River Catchments in Germany: An Ensemble Assessment

AbstractThe impact of climate change on three small- to medium-sized river catchments (Ammer, Mulde, and Ruhr) in Germany is investigated for the near future (2021–50) following the Intergovernmental Panel on Climate Change (IPCC) Special Report on Emissions Scenarios (SRES) A1B scenario. A 10-member ensemble of hydrological model (HM) simulations, based on two high-resolution regional climate models (RCMs) driven by two global climate models (GCMs), with three realizations of ECHAM5 (E5) and one realization of the Canadian Centre for Climate Modelling and Analysis version 3 (CCCma3; C3) is established. All GCM simulations are downscaled by the RCM Community Land Model (CLM), and one realization of E5 is downscaled also with the RCM Weather Research and Forecasting Model (WRF). This concerted 7-km, high-resolution RCM ensemble provides a sound basis for runoff simulations of small catchments and is currently unique for Germany. The hydrology for each catchment is simulated in an overlapping scheme, with t...

Impact of RCM Spatial Resolution on the Reproduction of Local, Subdaily Precipitation

AbstractMany hydrological hazards are closely connected to local precipitation (extremes), especially in small and urban catchments. The use of regional climate model (RCM) data for small-scale hydrological climate change impact assessment has long been nearly unfeasible because of the low spatial resolution. The RCM resolution is, however, rapidly increasing, approaching the size of small catchments and thus potentially increasing the applicability of RCM data for this purpose. The objective of this study is to explore to what degree subhourly temporal precipitation statistics in an RCM converge to observed point statistics when gradually increasing the resolution from 50 to 6 km. This study uses precipitation simulated by RCA3 at seven locations in southern Sweden during 1995–2008. A positive impact of higher resolution was most clearly manifested in 10-yr intensity–duration–frequency (IDF) curves. At 50 km the intensities are underestimated by 50%–90%, but at 6 km they are nearly unbiased, when average...

Modelling Near Future Regional Climate Change for Germany and Africa

The scope of regional climate simulations carried out at the Institute for Meteorology and Climate Research (IMK) of Karlsruhe Institute of Technology (KIT) using the regional climate model (RCM) COSMO-CLM (CCLM) has been extended during the last years. From the focus first on Southwest Germany the area of interest has then been extended to the whole of Germany for the assessment of changes in flood risk for medium and small mountainous river catchments (CEDIM project: www.cedim.de), and within the frame of CORDEX (Coordinated Regional climate Downscaling Experiment, http://wcrp.ipsl.jussieu.fr/SF_RCD_CORDEX.html) also to the whole African continent. CORDEX aims to provide a framework to evaluate and benchmark RCMs and to design a set of experiments to produce climate projections for use in impact and adaption studies and as input to the IPCC 5th Assessment Report (AR5).

On the effects of constraining atmospheric circulation in a coupled atmosphere-ocean Arctic regional climate model

Impacts of spectral nudging on simulations of Arctic climate in coupled simulations have been investigated in a set of simulations with a regional climate model (RCM). The dominantly circumpolar circulation in the Arctic lead to weak constraints on the lateral boundary conditions (LBCs) for the RCM, which causes large internal variability with strong deviations from the driving model. When coupled to an ocean and sea ice model, this results in sea ice concentrations that deviate from the observed spatial distribution. Here, a method of spectral nudging is applied to the atmospheric model RCA4 in order to assess the potentials for improving results for the sea ice concentrations when coupled to the RCO ocean-sea ice model. The spectral nudging applied to reanalysis driven simulations significantly improves the generated sea ice regarding its temporal evolution, extent and inter-annual trends, compared to simulations with standard LBC nesting. The method is furthermore evaluated with driving data from two CMIP5 GCM simulations for current and future conditions. The GCM biases are similar to the RCA4 biases with ERA-Interim, however, the spectral nudging still improves the surface winds enough to show improvements in the simulated sea ice. For both GCM downscalings, the spectrally nudged version retains a larger sea ice extent in September further into the future. Depending on the sea ice formulation in the GCM, the temporal evolution of the regional sea ice model can deviate strongly.

Modelling Regional Climate Change in Germany

A series of regional climate simulations using the regional climate model COSMO-CLM (CCLM) have been carried out at the Institute for Meteorology and Climate Research (IMK) of Karlsruhe Institute of Technology (KIT) (Panitz et al. 2010). These simulations span the years 1971–2000 to represent the climate of the recent past and the years 2011–2040 to analyse the climate change in Southwest Germany during the next few decades. A second set of simulations covering all of Germany, have been carried out for the CEDIM-project “Hochwassergefahr durch Klimawandel”, to assess possible changes in flood hazard in the near future. Model validation, results for changes in temperature and precipitation statistics, and statistics on the computational resources used at the HLRS facitilites are presented.

Downscaling Climate Simulations for Use in Hydrological Modeling of Medium-Sized River Catchments

To assess a possible future change in flood and drought risks for medium and small-scale river catchments, one needs to have data of a higher spatial and temporal resolution than what is provided by the global climate models. The COSMO-CLM regional climate model has to this purpose been used to downscale a set of global climate simulations to a 7 km horizontal resolution. In order to assess some of the uncertainties involved in near future scenario simulations, several different global simulations are downscaled to produce an ensemble of high resolution data. This will then be used as input to hydrological catchment models to assess future changes in flood risk for three catchments in Germany, within the CEDIM-project “Flood hazard in a changing climate” (Hochwassergefahr durch Klimawandel).