Monitoring networks and meteo-marine forecast

Lorenzo Papa

Hydrographic Institute of the Italian Navy and University of Genova-DIFI (I)

Abstract

In Winter the Mediterranean is frequently affected by depressions associated with a branch of the westerly jet stream located about 35° N. At intervals, winds from North Africa and the Atlantic converge over the Mediterranean bringing together air of markedly different temperature and initiating frontogenesis. Atlantic depressions may enter the western basin or, more frequently, cyclogenesis occurs in the Gulf of Genoa or in the Tyrrhenian Sea in the lee of the Alps. Nearly 70% of all Mediterranean depressions are of the "Genoa type". In summer the Mediterranean basin is dominated by the expanded Azores anticyclone in the west and air-mass contrasts are much reduced compared with winter. As nonlinear dynamical systems, the atmospheric motions in this area are not perfectly predictable in a deterministic sense and the pressure at the sea level can be seen as a superposition of periodic, almost periodic, transient and random processes. The frequency-domain analysis represents data series in terms of contributions occurring at different time scales, or characteristics frequencies. In this note we present experimental monitoring networks and a numerical model that might be viewed as a process or algorithm, that can predict the future behaviour of the pressure field in the Ligurian Sea or generally in the north-western Mediterranean. The new technique operates in the frequency-domain of pressure data measured at the sea level in Genoa, Asti and Cuneo. This model operates without information from the fluid-dynamical numerical weather prediction (NWP) models that have become the mainstay of weather forecasting for lead times ranging from one to a few days in advance. An important application of this technique to weather forecasting in the north-western Mediterranean basin is in conjunction with NWP models information, since it seems a powerful guidance product to aid weather forecasters in this area.