Drought recovery of southern Andes natural catchments

Abstract

<p>Drought causes hydroclimatic stress on terrestrial ecosystems and we see that these effects can have a long duration even when the drought is alleviated. The impacts of a drought are commonly dependent on the severity and duration of the hydrological drought event. At the same time, we see that the recovery from a severe drought is also impacted by catchment characteristics and regional climatology. In this study, we focus on Chile which has frequently experienced multi-drought periods with severe impacts.</p> <p>In this study, we quantify the recovery of discharge, vegetation productivity (kNDVI), and soil moisture after hydrological droughts to quantify the drought termination (DT) and drought termination duration (DTD). We used the CAMELS-CL data set from 1988-2020 to study drought recovery in natural catchments. Using a composite analysis we obtain the average response of discharge, vegetation, and soil moisture after severe drought events for catchments throughout Chile. We estimate the impact of different explanatory variables and catchment properties from the CAMELS-CL data on DT and DTD using lasso regression for discharge, vegetation productivity, and soil moisture without selecting strongly correlated variables.</p> <p>Our study demonstrates that the drought recovery of discharge can be explained by local characteristics while these relationships are less pronounced for vegetation and soil moisture droughts. Longer recovery times were found in environments with less precipitation and higher temperatures, with mainly shrub land cover. Shorter recovery times, at higher latitudes with increasing precipitation and lower temperatures under higher vegetation cover. The explanatory variables for discharge DT and DTD are associated with precipitation, potential evapotranspiration, and baseflow, or by a combination of them with catchment characteristics related to storage and release (e.g. land use). To that end, this work can help to identify drought vulnerability in regions where observations are lacking and help to predict drought recovery periods.</p>