An evaluation of statistical synoptic models of rainfall in Spain

• Main Author: Spellman, Greg
• Other Authors: McClatchey, J.
• Published: University of Northampton 2003
• Subjects: 551.577246; QC980 Climatology and weather
• Online Access: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.402520

This study investigates the control of atmospheric circulation patterns on rainfall incidence in Spain. The main objective of the research is to evaluate a range of statistical synoptic approaches with the aim of identifying the scheme that best models circulation to association. Spatial patterns of rainfall in Spain are first investigated using Principal Components Analysis and Cluster Analysis. Distinct precipitation affinity groups emerge that display covariant rainfall behaviour and reflect differences in latitude, the influence of topography and distance from the synoptic feature responsible for rainfall. The method allows seasonal redefinition of boundaries and the investigation of the effect of climate change. In total 24 synoptic models are investigated. The best performing models (a daily weather type model and a monthly airflow index model) use standardized data and the 500hPa contour surface. Some of the problems associated with non-stationarity are attempted by modifying models using kinematic information. Adjustments to the models (inclusion of frontal information and stochastic modelling) can improve results on a sub-regional scale. Effective models are then used to empirically downscale from General Circulation Model (GCM) scenarios obtained from the Canadian Centre for Climate Modelling and Analysis. The downscaling procedure is of limited use due to errors in GCM output but results suggest strongly increasing anticyclonicity in the Iberian area and a decrease in rainfall in many areas. There are uncertainties associated with regional scale climate change estimation using current empirical methods, nevertheless as GCM output inevitably becomes more accurate the scope for detailed regional assessment will improve