Spatial extension of extreme rainfall events: return period of isohyets area and influence of rain gauges network evolution

Those last 10 yrs, many cities in southern Europe have been affected by heavy rainfall events leading to severe runoffs. Thus, the assessment of rainfall risk needs a better knowledge of the climate hazards and particularly the rainfall hazards. A spatial stochastic analysis is proposed here, based on rainfall events selected during the 1958–1993 period in the South of France, using the French weather service daily rainfall database. The isohyet areas defined at various rainfall thresholds τ are estimated. Because of the sampling constraints, two different time steps have been considered: 24 and 48 h. A two parameters gamma distribution law has been fitted to each sample of isohyet areas, and the two parameters are related to the threshold τ. For those 36 yrs, the evolution of the network of rain gauges in space and time has been described. The empirical expression of the probability to observe a given rainy area with a given network configuration, using simulation, has been proposed. The bias introduced in sampling events has been estimated and it is shown that its influence on isohyet areas quantiles could be negligible, for studies based on daily rainfall gauges.