Statistical Downscaling of Daily Precipitation over Sweden Using Gcm Output

Fredrik Wetterhall,András Bárdossy,Deliang Chen,Sven Halldin,Chong-yu Xu
DOI: https://doi.org/10.1007/s00704-008-0038-0
2008-01-01
Theoretical and Applied Climatology
Abstract:A classification of Swedish weather patterns (SWP) was developed by applying a multi-objective fuzzy-rule-based classification method (MOFRBC) to large-scale-circulation predictors in the context of statistical downscaling of daily precipitation at the station level. The predictor data was mean sea level pressure (MSLP) and geopotential heights at 850 (H850) and 700 hPa (H700) from the NCEP/NCAR reanalysis and from the HadAM3 GCM. The MOFRBC was used to evaluate effects of two future climate scenarios (A2 and B2) on precipitation patterns on two regions in south-central and northern Sweden. The precipitation series were generated with a stochastic, autoregressive model conditioned on SWP. H850 was found to be the optimum predictor for SWP, and SWP could be used instead of local classifications with little information lost. The results in the climate projection indicated an increase in maximum 5-day precipitation and precipitation amount on a wet day for the scenarios A2 and B2 for the period 2070–2100 compared to 1961–1990. The relative increase was largest in the northern region and could be attributed to an increase in the specific humidity rather than to changes in the circulation patterns.
What problem does this paper attempt to address?