This work presents a characterization of the surface wind climatology over the Iberian Peninsula (IP). For this objective, an unprecedented observational database has been developed. The database covers a period of 6 years (2002–2007) and consists of hourly wind speed and wind direction data recorded at 514 automatic weather stations. The original observations underwent a quality control process to remove rough errors from the data set. In the first step, the annual and seasonal mean behaviour of the wind field are presented. This analysis shows the high spatial variability of the wind as a result of its interaction with the main orographic features of the IP. In order to simplify the characterization of the wind, a clustering procedure was applied to group the observational sites with similar temporal wind variability. A total of 20 regions are identified. These regions are strongly related to the main landforms of the IP. The wind behaviour of each region, characterized by the wind rose (WR), annual cycle (AC) and wind speed histogram, is explained as the response of each region to the main circulation types (CTs) affecting the IP. Results indicate that the seasonal variability of the synoptic scale is related with intra‐annual variability and modulated by local features in the WRs variability. The wind speed distribution not always fit to a unimodal Weibull distribution consequence of interactions at different atmospheric scales. This work contributes to a deeper understanding of the temporal and spatial variability of surface winds. Taken together, the wind database created, the methodology used and the conclusion extracted are a benchmark for future works based on the wind behaviour.

A set of four regional climate change projections over the Iberian Peninsula has been performed. Simulations were driven by two General Circulation Models (consisting of two versions of the same atmospheric model coupled to two different ocean models) under two different SRES scenario. The XXI century has been simulated following a full-transient approach with a climate version of the mesoscale model MM5. An Empirical Orthogonal Function analysis (EOF) is applied to the monthly mean series of daily maximum and minimum 2-metre temperature to extract the warming signal. The first EOF is able to capture the spatial structure of the warming. The obtained warming patterns are fairly dependent on the month, but hardly change with the tested scenarios and GCM versions. Their shapes are related to geographical parameters, such as distance to the sea and orography. The main differences among simulations mostly concern the temporal evolution of the warming. The temperature trend is stronger for maximum temperatures and depends on the scenario and the driving GCM. This asymmetry, as well as the different warming rates in summer and winter, leads to a continentalization of the climate over the IP. Zusammenfassung

No abstract

No abstract