Ernst et al. (2025) Historical and Future Drought Intensification in the Pantanal Wetland: Evidence from Multi-Source Weather Data and CMIP6 Multi-Model Projections

Identification

• Journal: Environments
• Year: 2025
• Date: 2025-11-02
• Authors: Jakob Ernst, Milica Stojanovic, Rogert Sorí
• DOI: 10.3390/environments12110413

Research Groups

• Centro de Investigación Mariña, Environmental Physics Laboratory (EPhysLab), Universidade de Vigo, Ourense, Spain

Short Summary

This study assessed historical (1980-2024) and future (2026-2100) drought intensification in the Pantanal wetland using multi-source weather data and bias-corrected CMIP6 multi-model projections. It revealed a significant historical drying trend and projected intensified, severe multi-year droughts under future climate change scenarios, particularly when considering evaporative demand.

Objective

• To explore historical (1980–2024) and probable future (2026–2100) drought conditions in the Pantanal wetland, employing bias-corrected CMIP6 multi-model data under SSP2-4.5 and SSP5-8.5 scenarios, using both the Standardised Precipitation Index (SPI) and Standardised Precipitation–Evapotranspiration Index (SPEI).

Study Configuration

• Spatial Scale: The Pantanal wetland, spanning parts of Brazil, Bolivia, and Paraguay.
• Temporal Scale: Historical period: 1980–2024. Future projections: 2026–2100. Reference periods for drought indices: 1980–2014, 1980–2100, and 2041–2100.

Methodology and Data

• Models used:
  • Drought Indices: Standardised Precipitation Index (SPI) and Standardised Precipitation–Evapotranspiration Index (SPEI) at 3-month and 12-month timescales.
  • Climate Models: Multi-model mean (MMM) from five Coupled Model Intercomparison Project Phase 6 (CMIP6) models (CanESM5, EC-Earth3, GFDL-ESM4, MPI-ESM1-2-HR, NorESM2-MM).
  • Potential Evapotranspiration (PET) Calculation: Modified Hargreaves method.
  • Bias Correction: Quantile mapping (QM) approach applied to MMM data.
  • Drought Risk Assessment: Thermal–counterfactual (CF) approach with temperature anomaly bins and risk ratios (RRb).
• Data sources:
  • Historical/Reference Data: Multi-Source Weather data (MSWX-PAST) for 1980–2024, providing monthly total precipitation (Tp), mean, maximum, and minimum 10 m air temperature (Tas, Tmax, Tmin) at 0.1° resolution, derived from bias-correcting ERA5 reanalysis.
  • Future Projections: NEX-GDDP-CMIP6 dataset (global downscaled climate scenarios from CMIP6 General Circulation Models) for 2015–2100 under Shared Socioeconomic Pathways (SSPs) SSP2-4.5 and SSP5-8.5.

Main Results

• Historical Trends (1980–2024): A statistically significant drying trend was observed, with negative slopes ranging from −0.39 to −0.42 units per decade for SPI3 and SPEI3, culminating in extreme droughts in 2019/2020 and 2023/2024. The SPEI showed a more pronounced decline than the SPI, highlighting the role of evaporative demand.
• Future Projections (2026–2100):
  • Precipitation and Temperature: Precipitation is projected to decrease by more than 15 mm month⁻¹ under SSP2-4.5 and more than 25 mm month⁻¹ under SSP5-8.5 by 2100. Mean air temperature is projected to increase by over 3 °C under SSP2-4.5 and more than 6 °C under SSP5-8.5.
  • Drought Severity: Both scenarios project increased frequency, duration, and severity of droughts. The SPEI consistently indicates stronger drying than the SPI, emphasizing the growing role of temperature-driven evaporative demand.
  • SSP2-4.5 Scenario: Projects more variable but still intensifying dry spells. Risk ratios for extreme droughts (τ ≤ −1.65) approach ~2.5 for SPI3 and ~3.4 for SPEI3 relative to the 0.5–1.5 °C warming bin. Three extreme droughts are anticipated around 2090.
  • SSP5-8.5 Scenario: Projects persistent, severe multi-year droughts, becoming prevalent from 2060 onwards and surpassing extreme thresholds. Risk ratios for extreme droughts (τ ≤ −1.65) climb to ~7.5 for SPI3 and ~7 for SPEI3, implying a dramatic increase in occurrence probability. A long-lasting extreme drought is forecasted for the mid-2070s.
• Reference Period Impact: The choice of reference period significantly influences drought assessment, with a fixed historical baseline (1980–2014) revealing a clearer shift towards moderate-to-extreme droughts compared to future baselines.

Contributions

• Provides comprehensive, high-resolution, and multi-scalar drought projections for the Pantanal using the latest CMIP6 models and Shared Socioeconomic Pathways (SSPs), addressing gaps from previous studies that relied on older models or limited indices/timescales.
• Utilizes both SPI and SPEI across multiple timescales and reference periods, offering a nuanced interpretation of drought severity and frequency relative to historical and projected climates, thereby improving robustness for impact and adaptation assessments.
• Employs a bias-corrected multi-model mean (MMM_BC) approach, which is shown to provide a more realistic representation of recent drought extremes in the Pantanal compared to previous ensemble-based methods.
• Quantifies how drought risk scales with warming through a temperature-binned attribution framework, explicitly demonstrating the progressive amplification of drought risk, particularly for severe and extreme events, and confirming the critical role of temperature in atmospheric evaporative demand.

Funding

• Xunta de Galicia (project Excelencia-ED431F-2024/03, ED431C2021/44, Postdoctoral Grant No. ED481D −2024/017)
• Ministerio de Ciencia, Innovación y Universidades, Spain (MICIU/AEI, grant RYC2021-034044-I)
• European Union Next Generation EU/PRTR
• Computing resources provided by CESGA (Centro de Supercomputación de Galicia) and RES (Red Española de Super-computación)

Citation

@article{Ernst2025Historical,
  author = {Ernst, Jakob and Stojanovic, Milica and Sorí, Rogert},
  title = {Historical and Future Drought Intensification in the Pantanal Wetland: Evidence from Multi-Source Weather Data and CMIP6 Multi-Model Projections},
  journal = {Environments},
  year = {2025},
  doi = {10.3390/environments12110413},
  url = {https://doi.org/10.3390/environments12110413}
}