Event-based dynamics of the soil water content at Alpine sites (Tyrol, Austria)

Abstract

The relevance of the soil moisture content (SMC) for runoff processes and thus for flood prediction is rather undisputed; however, there are still gaps in our knowledge of the precipitation – soil moisture reaction pattern. A new event-based approach, using statistical methods to analyse monitoring data, is presented and applied for soil moisture data of three test sites, situated in different parts of the Tyrolean Alps (Austria).The analyses revealed a significant influence of the initial soil moisture on the remaining soil storage capacity. A comparison of the value pairs initial soil moisture and soil moisture difference (SMCmax caused by rain events minus initial SMC) resulted in funnel-shaped distributions, whereby the upper boundary of the data should roughly match the field capacity of the soils. For purposes of flood prediction, precipitation sums and initial conditions are the relevant variables, while precipitation duration, intensities, and antecedent precipitation have significantly less influence on SMC. However, the functions describing the relationship between precipitation and SMC showed poor model fits despite their significance.The linkage of soil moisture content between different depths, which, inter alia, are important when using remote sensing data for spatial SMC estimations, showed that the soil moisture conditions in the upper layer do not necessarily represent the water storage state of the deeper layers. The relation strongly depends on the kind of transition between the horizons, underlining the great importance of proper spatial information on soil characteristics.Approaches to estimate the initial SMC at the beginning of heavy rainfall events on the basis of antecedent precipitation sums are not very promising due to the weak relationships between these parameters.