Numerical modeling as support to the study of precipitation over the Madeira Island (Portugal)

Abstract

The Madeira is a small mountainous island (~740 km2) located in the North Atlantic Ocean (32°75'N and 17°00'W), and after the disaster occurred on 20 February 2010, when extreme precipitation induced flash floods and landslides in some spots of the island, efforts have been directed toward the understanding of the precipitation affecting the island. The occurrence of extreme precipitation in three seasons in the last years have raised questions mainly about the atmospheric conditions that may lead to the development of intense precipitation with high impact at the surface. Our goals in this work are to identify and describe the large- and meso- scale features associated to four periods of significant accumulated precipitation during the autumn 2012. The precipitation recorded by the Madeira meteorological network is analyzed, as well as satellite images in order to describe the characteristics of the precipitation systems that reached the Madeira in each period. The synoptic scale conditions are described thank to the 6-hourly ARPEGE operational analyses. The four periods were associated with different large-scale precipitating systems. The mesoscale environment and precipitating systems have been then studied thank to very-high resolution numerical simulations using the MESO-NH non-hydrostatic model. The numerical experiments were carried out using as initial and boundary conditions the ARPEGE analyses. The model was configured in two two-way nested domains: the larger domain with a grid spacing of 2.5 km and 600x500 points, and the inner domain, with 600x600 grid points and an horizontal resolution of 0.5 km. The simulations compared well with the rain gauge observations over the island. The simulation shows a strong impact of the mountainous island on the spatial distribution and volume of precipitation. The 500 m resolution simulation evidences the relationship between the local topography and precipitation.