A local regression approach to analyze the orographic effect on the spatial variability of sub-daily rainfall annual maxima

In this work we investigate the spatial variability of sub-daily rainfall extremes over Italy considering the influence of local orographic effects. We consider the average annual maxima computed from the recently-released Improved Italian - Rainfall Extreme Dataset (I-2-RED) in about 3800 time series with at least 10 years of data (1916-2020 period) and we analyze the orographic effects through a local regression approach which gathers stations in a grid cell-centered area of 1 km resolution. Several constraints are considered to tackle problems determined by the low data density of some areas and by the extrapolation at low/high elevations. Different criteria for selecting the local sample are examined. This work confirms with increased detail previous findings, such as a generally positive gradient of the 24 h average annual maxima and the evidence of negative gradients in large mountainous areas for the 1 h maxima. The use of a local regression approach allows to identify the areas showing the reverse orographic effect, providing material for future investigations on the physical explanation of this evidence. Moreover, the reconstructed maps will allow to apply more accurate approaches in works related to the spatial variability of other rainfall statistics, such as the quantiles required for hydrologic design.

Automated summary

In this study, the spatial variability of sub-daily rainfall extremes over Italy is investigated taking into account the influence of local orographic effects. The study uses the Improved Italian - Rainfall Extreme Dataset (I-2-RED) to analyze the orographic effects through a local regression approach. Different criteria for selecting the local sample are examined. The results confirm previous findings and provide material for future investigations on the physical explanation of reverse orographic effects.