Spatio-temporal patterns of ecologically-relevant river water quality extremes during hydrological droughts

Dry periods can negatively impact river ecosystems through deficits in streamflow (hydrological droughts) and deteriorations in water quality. Yet, literature on water quality extremes is in its infancy and the co-occurrence of hydrological droughts and water quality extremes remains poorly understood, especially for different hydro-climatic regimes. With hydrological models often struggling to capture extreme conditions, observation-based studies from multiple catchments are crucial to gain insights on regional-scale patterns. High-frequency observations - at least daily - are needed to properly characterize often short-lived extreme events, but such observations are rarely available in current large-sample water quality datasets, which mostly have a relatively low temporal resolution (e.g., monthly). Therefore, data fragmentation, both in space and time, has hampered our understanding of the co-occurrence of hydrological droughts and water quality extremes to date. Specifically, here we ask the questions: Where and when do ecologically-relevant water quality extremes occur during hydrological droughts? and What are the hydro-climatic factors influencing these events? To address these questions, we rely on a newly-assembled dataset of water quantity and quality observations (i.e., streamflow, water temperature, electrical conductivity, and dissolved oxygen) at a daily resolution for 43 catchments across Europe and the USA and the period 2005−2024. For this case study, we derive hydrological droughts and water quality extremes (namely, sustained periods of abnormally low flows, abnormally high water temperature and electrical conductivity, and abnormally low dissolved oxygen) using a percentile-based approach with seasonally and catchment varying thresholds. We then focus specifically on concurrent hydrological droughts and water quality extremes, and in particular on those events exceeding thresholds that can lead to stress for aquatic animals. We finally characterize these events using metrics describing their magnitude and timing, and perform regression analyses to shed light on our research questions. Preliminary results indicate that hydrological droughts co-occurring with water quality extremes exhibit notable spatio-temporal variability. They further suggest that ecologically-relevant thresholds are exceeded in terms of water temperature in particular, and especially in catchments with human influence and minimal snow contribution. This study will enhance our mechanistic understanding on the potential impacts of hydrological droughts and water quality extremes on river ecosystems, by providing relevant information for river management under drying and warming.