Anthropogenic climate change has reduced drought recovery probabilities across the western US

Effective water resource management necessitates understanding drought recovery timelines. Prolonged droughts severely impact ecosystems, water supplies, and societal infrastructure. Resource managers and the public alike need to know the probability of drought recovery (PDR) to inform crucial decisions, such as implementing or easing water restrictions. The 2022 California drought, for instance, highlighted this need, with the state facing its third consecutive dry year and nearing severe water allocation cuts before record-breaking precipitation in 2023 led to drought recovery. Drought recovery is a complex process depending on drought severity, subsequent meteorological conditions (precipitation and evaporative demand), and the amount of surplus moisture needed to overcome prior deficits. Elevated temperatures post-drought can exacerbate water deficits by increasing evaporative demand, while above-average precipitation generally facilitates recovery. Current efforts to predict drought duration and mitigation strategies often rely on historical analogs, assuming hydrometeorological stationarity. However, this assumption is increasingly questionable due to anthropogenic climate change, which has been shown to intensify droughts and alter their onset and severity. This research directly addresses whether and how climate change affects observed PDR, providing a crucial understanding for improved water resource management strategies.

Prior research has extensively documented the impacts of climate change on drought frequency, intensity, and duration, particularly in the southwestern US. Studies have shown increased atmospheric evaporative demand in the region's hot, dry seasons, leading to more frequent occurrences of dry and hot conditions, and accelerating drought onset ('flash droughts'). While some observed reductions in precipitation may be due to natural variability, climate change has significantly increased the severity and frequency of droughts. Models have projected further increases in both drought frequency and duration under future climate scenarios. However, the impact of climate change on observed drought recovery probabilities (PDR) following severe drought events, remained unexplored until this study.

This study employs a multi-pronged approach to assess the impact of climate change on drought recovery in the western US. The researchers utilize monthly self-calibrating Palmer Drought Severity Index (scPDSI) data, a widely used drought indicator based on precipitation and reference evapotranspiration (ETo). Data are analyzed for five large river basins: the Upper and Lower Colorado Basins, the Great Basin, the Pacific Northwest, and California. Three datasets were analyzed: 1) observational scPDSI data (1901-2023); 2) experimental data initializing each month with severe drought and applying observational data to model recovery; and 3) scPDSI data from 23 Coupled Model Intercomparison Project Phase 6 (CMIP6) climate models. The observational and experimental datasets also include counterfactual estimates excluding anthropogenic climate trends, allowing isolation of climate change impacts. Drought recovery is defined as non-drought conditions (scPDSI ≥ 30th percentile) in each of the 24 months following severe drought (scPDSI ≤ 10th percentile). PDR is calculated for different time periods, climate scenarios, and seasons. Statistical significance is assessed using McNemar's test (for observed vs. counterfactual data), t-tests (for historical vs. recent data), and bootstrap confidence intervals (for model-based data). Seasonal variations and the time of emergence (ToE) of climate change signals are also explored. The analysis determines the contribution of changes in precipitation (P) and ETo to changes in PDR.

The study's key findings reveal a significant decrease in drought recovery probabilities across the western US in recent decades (2000-2021) compared to the historical period (1901-1980). Observational data shows a 25-50% reduction in PDR 18 months after severe drought in California and the southwestern basins (UCB, LCB, GB), with at least one-third of this decline attributable to anthropogenic climate change. Counterfactual simulations, removing anthropogenic trends, support this finding. Analysis of climate model simulations (CMIP6) show qualitatively similar results, with statistically significant reductions in PDR for the Lower Colorado Basin, Pacific Northwest, and California. These model-based findings indicate that it now takes 1–4 months longer on average to recover from severe drought in these regions, with the greatest increases in recovery time seen for coastal areas and the Pacific Northwest. The primary driver of this reduced PDR is the increase in evaporative demand (ETo) during non-winter months, as revealed by the model simulations. While some coastal regions showed slightly increased winter precipitation resulting in faster winter recovery, this effect is far outweighed by the summer ETo driven decrease in recovery probability. Time series analysis using moving 21-year windows revealed significant multidecadal variability in observed PDR, highlighting the challenge of detecting trends solely from observational data. The climate model results, however, consistently demonstrate a recent decline in PDR across all regions.

These findings demonstrate a clear link between anthropogenic climate change and the decreased probability of drought recovery in the western US. The observed decrease in PDR translates into significantly longer drought durations, impacting water resources, ecosystems, and society. The increased evaporative demand driven by elevated temperatures, particularly in non-winter months, is the main culprit. While natural variability contributes to drought conditions, the study isolates a significant portion of the observed PDR reduction attributable to anthropogenic forcing. The discrepancy between the observational and model results for some regions reflects the challenges in detecting climate change signals amidst high natural variability. The study's multiple lines of evidence approach helps mitigate this uncertainty. This study adds vital knowledge to the field, highlighting the need for water management strategies that account for longer and more severe drought events. While some coastal regions see faster winter recoveries due to increased winter precipitation, overall the model forecasts show a continuation of the trends leading to longer recovery times.

This research strongly suggests that anthropogenic climate change has significantly reduced drought recovery probabilities in the western US, primarily through increased evaporative demand during non-winter months. This leads to longer drought durations and has critical implications for water management and drought preparedness. The multi-method approach used enhances the robustness of the findings, although further investigation could explore additional drought indicators and refine understanding of interactions between climate change and natural variability. Future research should focus on developing improved drought forecasting models incorporating these findings and exploring more effective adaptation strategies to mitigate the impacts of increasingly prolonged and severe droughts.

The study acknowledges limitations inherent in using observational data, including potential biases and limited sample sizes, particularly for rarer events such as multi-year droughts. While the use of climate models helps address some of these limitations, uncertainties remain regarding model representation of climate processes, particularly regarding land-surface feedbacks in arid regions. Furthermore, scPDSI, while widely used, might not capture the complexities of all drought types and doesn’t fully reflect the impact of reduced snowpack on the region's hydrology. Future studies should consider integrating other drought metrics, improving model parameterizations, and investigating specific impacts on various sectors.