Empirical runoff thresholds: results from catchments in Central Italy

Understanding how much runoff is generated from precipitation events is an increasingly relevant topic in the context of climate change, especially in the Mediterranean area, where the interactions among the landscape, topography, soil, precipitation variability and basin antecedent conditions make this process difficult to model [e.g., 1]. Several authors have focused on defining empirical thresholds beyond which runoff (R) originates for specific event rainfall depths (P) and the antecedent soil moisture (θ), for example, by using the Antecedent Soil moisture Index (ASI). These approaches yielded promising results at the plot or catchments scale [2,3,4,5]. However, soil moisture is not frequently monitored in catchments – or – when it is monitored via only a few point measurements it is not guaranteed it correctly represents the true catchment soil moisture conditions. In recent years, the increasing development of space-born platforms has provided soil moisture estimates at different spatial and temporal scales, offering a viable way to complement ground-based observations. The present study integrates ground-based data and satellite soil moisture estimates to empirically evaluate runoff thresholds in two scarcely anthropized, small to medium-sized and medium-low permeability basins in Umbria (Central Italy). In detail, the Tatarena Creek catchment has an area of 32 km2 and is mainly covered by flyschoid rocks; the Reschio Creek basin is 57 km2 and is characterised by layered sandstone interbedded with thin marls and pelitic levels. Corine Land Cover 2018 images show that the two basins have different land uses: the Tatarena is mainly characterised by agricultural areas (72%), and the Reschio basin shows broad- leaved forest as predominant land use (60%). The Copernicus Global Land Service SWI (Soil Water Index) 1km dataset is used to estimate the antecedent soil degree of saturation. As for the ASI index, the Antecedent SWI is introduced (ASWI, mm) by multiplying SWI values by the depth of investigation (i.e., D≈10 cm considering a T-value of 10 [6]). Event-scale runoff thresholds have been empirically defined by analysing R as a function of P + ASWI. Sixty-five runoff events were identified for the Tatarena Creek basin in the 2015- 2023. After the analysis, a runoff threshold of about 68 mm was obtained by the sum of the rainfall depth (P) and ASWI. Nineteen runoff events were identified for the Reschio Creek catchment during 2020-2023 (runoff threshold of about 75 mm). The preliminary results (Fig. 1) indicate a slightly higher P+ASWI threshold for the Reschio catchment than the Tatarena one. The approach presented here can be helpful to investigate other catchments, contributing to the knowledge of the rainfall-runoff process. In-depth studies are being conducted on basins with different lithological characteristics.