The drought recovery spectrum: variable storage controls across a European climatic gradient

Hydrological drought recovery is a complex, non-linear process that frequently lags behind the return of seasonal precipitation. While drought onset and propagation are well-documented, the internal mechanisms governing recovery remain poorly understood. This research investigates the hypothesis that drought recovery is a scale-dependent process in which the effects of atmospheric variables, such as precipitation (P) and air temperature (T), are filtered through catchment storage compartments, specifically soil moisture (SM) and groundwater (GW), before resulting in the generation of stream runoff (Q). 

The study considers a multi-site climatic gradient to capture diverse hydrological behaviors: the Ressi catchment (Italy), which is a humid temperate pre-alpine climate with a mean annual P of 2119 mm and mean annual T  of 10.2 °C; the Can Vila catchment (Spain), which is a humid Mediterranean climate with mean annual  P of 918 mm and mean annual T of 10.3 °C; and the Weierbach catchment (Luxembourg), which is a temperate semi-oceanic climate with mean annual P of 898 mm and mean T of  8.7 °C. By leveraging high-resolution hydrometeorological data (P, T, SM, GW, and Q) spanning several years, the research moves beyond traditional linear analysis, employing wavelet analysis to identify time-frequency localizations and scale-dependent lag times in the relationships among the hydrometeorological variables considered.

Preliminary results show that hydrological recovery is not a simple function of P amount and distribution but is governed by internal storage behaviour. The Can Vila catchment functions as a threshold-based system in which intermittent Q depends on SM deficits. In this case, the soil acts as a collector, absorbing all initial P to satisfy SM deficits, resulting in no Q and a sudden “switch-like” recovery only once soil water storage is full. Conversely, “buffered” systems like the Weierbach catchment experience a lagged, multi-month recovery. In fact, the storage capacity acts as a long-term filter, providing resilience against short dry spells but requiring a prolonged period of consistent P to slowly recharge the system. Finally, “connected” systems such as the Ressi catchment demonstrate immediate recovery due to their short hydrological memory and constant vertical connectivity. In fact, considering the small size of the catchment’s storage, this leads to rapid fluctuations in SM, but also allows Q to reflect P almost instantaneously, with SM simply modulating the response of Q volume rather than delaying it. A more complete understanding of drought recovery dynamics will be gained through the upcoming analyses planned for GW.

The novelty of this work lies in the use of decadal wavelet analysis to examine where and how recovery from drought is influenced by local catchment characteristics. Considering that drought recovery is determined more by the internal conditions and dynamics than by meteorological factors alone, this study provides a framework for a more accurate understanding of drought recovery processes, highlighting the need for multi-compartmental monitoring to effectively manage water resources as climate variability increases across Europe.