Spatial distribution and trend analysis of extreme rainfall time series in Apulia region (Italy)

Roseto R., Dellino P., Schena P., Capolongo D., Spatial distribution and trend analysis of extreme rainfall time series in Apulia region ( Italy). (IT ISSN 0391-9838, 2023). Extreme rainfall events have been extensively studied at different spatial scales over the Mediterranean and Italian territory. Different interactions between atmospheric disturbances and morphological features have been found, mainly depending on precipitation duration and geo-graphical position. While climate change is shown to be responsible for the increase of frequency and intensity of extreme rainfall, the analysis of temporal trends at the national scale does not show significant variations and calls for more detailed studies. The main purposes of this study were both to investigate how the main geographical features of Apulia region controlled the spatial distribution of the extreme precipitations of short duration and to carry on a time-series analysis. The parameters of the Depth-Duration-Frequency (DDF) curves were correlated to elevation, latitude, longitude, distance from the sea and aspect by means of both Simple Linear Regression and Multiple Linear Regression. A wide range of duration were considered, spanning from 1 hour to 5 days. The Mann-Kendall non-parametric test was used as to demonstrate the significance of trend analysis. Multiple Linear Regression shows to overperform the Simple Linear Regression results, revealing that elevation has the strongest correlation with the DDF parameters, followed by longitude and distance from the sea. A slight increase of the sub-daily precipitation and a decrease of the multi- daily precipitation results from trend analysis, whereas no significant differences emerged during the climate normal 1991-2020. A set of maps are also presented as to point out the occurrence of local anomalies related either to the spatial variation of the DDF parameters and to the temporal trend direction. The outcomes of this study reveal the growing importance of this type of analysis, as well as the differences resulting from performing Simple and Multiple Linear Regression over time series related to a wide range of rainfall duration.