How does initial soil moisture influence the hydrological response? A case study from southern France

Abstract

The Cévennes–Vivarais region in southern France is prone to heavy rainfall that can lead to flash floods which are one of the most hazardous natural risks in Europe. The results of numerous studies show that besides rainfall and physical catchment characteristics the catchment's initial soil moisture also impacts the hydrological response to rain events. The aim of this paper is to analyze the relationship between catchment mean initial soil moisture θ̃ini and the hydrological response that is quantified using the event-based runoff coefficient ϕev in the two nested catchments of the Gazel (3.4 km2) and the Claduègne (43 km2). Thus, the objectives are twofold: (1) obtaining meaningful estimates of soil moisture at catchment scale from a dense network of in situ measurements and (2) using this estimate of θ̃ini to analyze its relation with ϕev calculated for many runoff events. A sampling setup including 45 permanently installed frequency domain reflectancy probes that continuously measure soil moisture at three depths is applied. Additionally, on-alert surface measurements at ≈10 locations in each one of 11 plots are conducted. Thus, catchment mean soil moisture can be confidently assessed with a standard error of the mean of ≤1.7 vol % over a wide range of soil moisture conditions. The ϕev is calculated from high-resolution discharge and precipitation data for several rain events with a cumulative precipitation Pcum ranging from less than 5 mm to more than 80 mm. Because of the high uncertainty of ϕev associated with the hydrograph separation method, ϕev is calculated with several methods, including graphical methods, digital filters and a tracer-based method. The results indicate that the hydrological response depends on θ̃ini: during dry conditions ϕev is consistently below 0.1, even for events with high and intense precipitation. Above a threshold of θ̃ini=34 vol % ϕev can reach values up to 0.99 but there is a high scatter. Some variability can be explained with a weak correlation of ϕev with Pcum and rain intensity, but a considerable part of the variability remains unexplained. It is concluded that threshold-based methods can be helpful to prevent overestimation of the hydrological response during dry catchment conditions. The impact of soil moisture on the hydrological response during wet catchment conditions, however, is still insufficiently understood and cannot be generalized based on the present results.