Topographic Effects on Statistical Characteristics of Heavy Rainfall and Mapping in the French Alps

Abstract

This paper uses detailed topographic characterization of relief for mapping statistical parameters of heavy rainfall in the French Alps. After determination of statistical parameters of heavy rainfall for time steps ranging from 1 to 24 h at rain gauging stations (10- and 100-yr rainfall), multivariate linear regression is used to identify the relationships between rainfall and morphometric parameters. Statistical characteristics of heavy rainfall events measured on short time steps (less than 3 h) are better linked to relief characteristics than are those on longer time steps. Furthermore, the rainfall parameters are shown to be closely linked to geographic features at a location in the Alps, such as distance to the Mediterranean Sea. More-local variables, such as altitude, slope, or azimuth, are less relevant. A mapping methodology based on linear relationships between rainfall parameters and topographic parameters is defined. This methodology takes into account the spatial structure of multivariate regression residuals. The performance of this method is compared with the simple interpolation method of kriging. For time steps shorter than 3 h, the information on relief improves the interpolation of heavy rainfall.