Rainfall erosivity in semi-arid Mediterranean areas: a multi-decadal spatio-temporal analysis of the R-factor (1951–2022)

Soil loss is a major environmental concern because it can compromise the provision of ecosystem services on both local and global scales. Therefore, developing effective mitigation measures and soil protection strategies is essential and should be grounded in a solid understanding of the main factors and processes that trigger and predispose erosion. Water related soil erosion is increasing in temperate as well as tropical and subtropical areas, especially in hilly and mountainous environments where cultivation and land-use changes reduce vegetation cover. Regions in southern Italy are characterized by a morphological configuration that increases the vulnerability to water erosion, making them relevant case studies representative of the Mediterranean area. In this context, accurately estimating the R-factor, which quantifies the potential of precipitation to cause soil erosion, is essential for assessing erosion risk and supporting effective soil conservation planning.

This study aims to update estimates of the R-factor across three regions of southern Italy (i.e., Campania, Sardinia, and Sicily) and to analyze its spatio-temporal evolution over the past 72 years. The first part of the study uses high resolution (10-minute) database of recent rainfall observations (2002-2023) to derive point-scale reference values of the R-factor at 169 stations in Campania, 134 in Sardinia and 92 in Sicily. These benchmark values, calculated using the RUSLE-2 method for rainfall kinetic energy combined with an innovative approach for identifying erosive events, were then used to locally calibrate and validate a selection of simplified empirical models at the regional level. Such simplified models can estimate the R-factor based on coarser resolution (i.e. daily) rainfall data, enabling broader regional applicability while maintaining accuracy.

In the second part of the study, the calibrated models were forced with daily rainfall data arising from a spatially distributed database (spatial resolution: 10 km x 10 km) covering the three regions over the period 1951-2022. This allowed the generation of an updated R-factor map for the regions and enabled an unprecedented analysis of trends over the 72-year time series. Specifically, trend detection was performed using the non-parametric Modified Mann-Kendall test at a significance level of α = 0.05, while trend magnitude was estimated using Sen’s slope estimator. The R-factor was computed over moving time windows of 16 years with a 4-year lag, also considering a multi-model ensemble mean approach. Additionally, Moran’s autocorrelation indices were used to investigate the spatial distribution of trends.

The results of the study revealed an overall positive trend in the R-Factor across the study area, indicating an increase in erosive potential of rainfall over the past years in all examined regions, with a more pronounced intensification in inland areas. The observed trends in southern Italy, consistent with broader patterns observed across the Mediterranean region, highlight the need for proactive soil conservation measures aimed at planning resilient and sustainable management strategies for regional agroecosystems.