Ravinandrasana et al. (2025) The first emergence of unprecedented global water scarcity in the Anthropocene

Identification

• Journal: Nature Communications
• Year: 2025
• Date: 2025-09-23
• Authors: Vecchia P Ravinandrasana, Christian L. E. Franzke
• DOI: 10.1038/s41467-025-63784-6

Research Groups

• Center for Climate Physics, Institute for Basic Science, Busan, Republic of Korea
• Department of Integrated Climate System Science, Pusan National University, Busan, Republic of Korea

Short Summary

This study estimates the Time of First Emergence (ToFE) of "Day Zero Drought" (DZD) events globally, attributing their timing and likelihood to human influence, and finds that many regions, including major reservoirs, face high DZD risk by the 2020s and 2030s, with urban populations particularly vulnerable at 1.5 °C warming.

Objective

• To estimate the Time of First Emergence (ToFE) of drought-driven water scarcity events, termed "Day Zero Drought" (DZD), which arise from hydrological compound extremes including prolonged rainfall deficits, reduced river flow, and increasing water consumption.
• To attribute the timing and likelihood of DZD events to human influence using a probabilistic framework and large ensemble climate simulations.
• To identify when and where anthropogenic DZD first emerges, how long these events will last, and how many people will be exposed to DZD.

Study Configuration

• Spatial Scale: Global, with analysis conducted at individual grid cells (approximately 1 degree by 1 degree for CESM2-LE, and 1.4 degrees by 1.4 degrees for CNRM).
• Temporal Scale: Historical and future projections from 1850 to 2100, with analyses focusing on decadal intervals and a 48-month timescale for drought and water scarcity indices.

Methodology and Data

• Models used:
  • Community Earth System Model version 2 Large Ensemble (CESM2-LE) (100 ensemble members, SSP3-7.0 scenario).
  • Centre National de Recherches Météorologiques (CNRM) model (10 ensemble members, SSP3-7.0 and SSP2-4.5 scenarios for sensitivity analysis).
• Data sources:
  • Climate projection data (CESM2-LE, CNRM) including temperature, precipitation, and river discharge.
  • ECMWF Reanalysis v5 (ERA5-Land) for observed meteorological and hydrological data (1979–2022).
  • Global Flood Awareness System (GloFAS) hydrological time series (GloFAS-ERA5).
  • Global Reservoir and Dams (GRanD) database (version 1.3) for reservoir location, capacity, and average discharge.
  • Global gridded monthly sectoral water consumption data (Khan et al., SSP3/SSP2, RCP6.0/RCP4.5 for 2010–2100, reconstructed for 1850–2009).
  • Global spatial population datasets (HYDE 3.3 for 1900–2009, NCAR SSP3 for 2010–2100).
  • Water stress indices: Standardized Precipitation-Evapotranspiration Index (SPEI48), Standardized River Flow Index (SRFI48), Standardized Water Scarcity Index (SWSI48), and Time for Reservoirs to Dry (TRD).
  • Bias correction: Univariate Quantile Delta Mapping (QDM) applied to drought indices using ERA5-Land as reference.
  • Attribution: Fraction of Attributable Risk (FAR) of joint probability to define ToFE (FAR ≥ 0.99).

Main Results

• Anthropogenic climate change is projected to trigger the onset and occurrence of DZD events, primarily due to severe and persistent compound multi-year hydrological droughts, reservoir depletion, and high-water demand.
• Approximately 14% of global reservoirs are at high risk of drying out due to DZD during their respective ToFE.
• About 35% of DZD-affected regions, including their reservoirs, are projected to face the emergence of water scarcity between 2020 and 2030 under the SSP3-7.0 scenario (CESM2-LE).
• Consistent DZD hotspots emerge across the Mediterranean, southern Africa, and parts of North America.
• Urban populations are particularly vulnerable, with 488 million people (322 million urban, 166 million rural) exposed to DZD at the 1.5 °C global warming level.
• Overall, over 753 million people (approximately 9% of the current global population) are projected to be exposed to DZD under the SSP3-7.0 scenario, with urban areas (467 million) more affected than rural areas (286 million).
• The length of time between successive DZD events (waiting time) is shorter than the duration of DZD in regions such as the Mediterranean, Southern Africa, Asia, and Australia, limiting recovery periods and exacerbating water scarcity risks.
• Almost 47% of global high-risk regions will experience extreme DZD events (where duration exceeds waiting time) with a frequency between 20% and 30%, while 12% of DZD-prone regions show a frequency above 50%.

Contributions

• This study provides the first global-scale assessment of the Time of First Emergence (ToFE) of "Day Zero Drought" (DZD) conditions and their characteristics, explicitly attributing these events to human influence.
• It introduces a comprehensive definition of DZD as an acute water scarcity event driven by compound multi-year hydrological drought, integrating prolonged rainfall deficits, reduced river flow, increased water consumption, and reservoir depletion (Time for Reservoirs to Dry, TRD).
• The research quantifies the spatial and temporal emergence of DZD hotspots, highlighting regions most vulnerable to unprecedented water scarcity under anthropogenic climate change.
• It assesses the duration and recurrence (waiting time) of DZD events, revealing that shorter recovery periods in some regions will exacerbate future water scarcity risks.
• The study quantifies population exposure to DZD, differentiating between urban and rural populations and linking this exposure to specific global warming levels, emphasizing the disproportionate impact on urban areas at 1.5 °C warming.

Funding

• Institute for Basic Science (IBS), Republic of Korea (IBS-R028-D1)
• National Research Fund of Korea (NRF-2022M3K1097082 and RS-2024-00416848)

Citation

@article{Ravinandrasana2025first,
  author = {Ravinandrasana, Vecchia P and Franzke, Christian L. E.},
  title = {The first emergence of unprecedented global water scarcity in the Anthropocene},
  journal = {Nature Communications},
  year = {2025},
  doi = {10.1038/s41467-025-63784-6},
  url = {https://doi.org/10.1038/s41467-025-63784-6}
}