MODELLING EXTREME PRECIPITATION IN HAZARDOUS MOUNTAINOUS AREAS. CONTRIBUTION TO LANDSCAPE PLANNING AND ENVIRONMENTAL MANAGEMENT

Extreme precipitation episodes are very common in Mediterranean area and can lead to serious and catastrophic environmental hazards. They have special incidence during autumn months, September, October and November (SON) with important impacts on society, leading frequently to significant economic losses and mortality. These events have special impact in mountainous areas where steep slopes enhance the effects of extreme precipitation. In mountainous areas rain gauge stations are sparse and normally in lower amount. Due to these reasons it is very important to map with higher accuracy the distribution of extreme precipitation. Also, in mountainous environments precipitation patterns can change in small distances that make the prediction more difficult, but also more important. A better prediction of areas with higher values of extreme precipitation will contribute to a better land use planning and avoid the effects of flash floods, landslides and soil erosion recognized as environmental problems. The aim of this paper is testing several well-known interpolation methods, Inverse Distance Weight (IDW) with weighs of 1, 2, 3, 4 and 5, Local Polynomial (LP) with order 1 and 2, Radial Basis Methods (RBS), particularly Spline With Tension (SPT) and Thin Plate Spline (TPS), and Kriging techniques, Ordinary Kriging (OK) and Ordinary CoKriging (COK) in order to identify the less-biased method to interpolate extreme precipitation calculated from the 95(th) percentile (P95) of SON precipitation in a mountainous area located in Portugal. The results show that extreme precipitation increases with the altitude and there are important differences between stations located at higher and lower altitudes. This relation is observed in the omni-directional semi-variograms calculated where we identified two major P95 areas coincident with higher elevations. The first one occurred at 12.19 km and the second at 23.57 km. The higher values of P95 are identified at Southeast and Northeast. In contrast, the lower P95 values are identified at Northwest due to lower altitudes and in the Northeast corner as a consequence of rain shadow effect. Prediction with precision of precipitation patterns in mountainous areas is difficult due to lack of data and the complex effect of topography in rainfall, however, it is of major importance in order to identify vulnerable areas. The findings observed in this study are a fundamental contribution to landscape planning and environmental management in areas with higher occurrence and vulnerability to extreme precipitation.