Climate change prediction over complex areas: spatial variability of uncertainties and predictions over the Pyrenees from a set of regional climate models

We used a set of six regional climate models (RCMs) from PRUDENCE project to analyse the uncertainties and direction and magnitude of the expected changes on precipitation and temperature (B2 and A2 scenarios) for the end of the 21st century in the Pyrenees, south of Europe. There have been few studies of climate change effects on this mountain range, though there can be noticeable impacts on the economy and ecology of the region and the surrounding lowlands. The analysis of the accuracy of the RCMs and the impacts of climate change over the region are addressed considering the mean values for the whole region, their spatial distribution patterns and the inter-model variability. Previously, the creation of distributed layers of temperature and precipitation from data provided by weather observatories was necessary to assess the ability of RCMs to reproduce the observed climate. Results show that mean biases between observed climate and control runs (1960–1990) are around 20% for precipitation and 1 °C for temperature. At annual basis, a mean decrease of 10.7 and 14.8% in precipitation, and an increase of 2.8 and 4 °C is expected in the next century in the area for A2 and B2 scenarios respectively. Effects of climate change will be more pronounced in the southern slopes of the range (Spanish Pyrenees), and lower in the coastland areas. Moreover, results on accuracy and expected changes are subject to a large spatial and seasonal variability as well, as the six RCMs present noticeable differences on accuracy and sensitivity to climate change forcings