Deepthi et al. (2026) Chaos Theory-Based Downscaling for Future Rainfall Projections from Climate Models

Identification

• Journal: Water Resources Management
• Year: 2026
• Date: 2026-02-01
• Authors: B. Deepthi, Bellie Sivakumar
• DOI: 10.1007/s11269-025-04382-5

Research Groups

• Department of Civil Engineering, Indian Institute of Technology Bombay, Powai, Mumbai, India
• Kerala State Council for Science, Technology and Environment, Thiruvananthapuram, India

Short Summary

This study applies a novel chaos theory-based downscaling method to project future monthly rainfall across India (0.25° × 0.25° resolution) using five CMIP6 General Circulation Models (GCMs) under four Shared Socioeconomic Pathway (SSP) scenarios (2015-2099), revealing significant regional shifts in both southwest and northeast monsoon rainfall patterns.

Objective

• To apply a chaos theory-based local approximation approach to downscale future rainfall projections from CMIP6 GCMs for India under various SSP scenarios and analyze the projected changes in southwest and northeast monsoon rainfall patterns across different future timeframes.

Study Configuration

• Spatial Scale: Mainland India, consisting of 4948 grids at a resolution of 0.25° × 0.25°. The coarse-resolution GCM outputs range from 0.9375° to 2.8° in latitude and 1° to 2.8° in longitude.
• Temporal Scale:
  • Historical period: 1961–2014 (used for model parameterization and validation).
  • Future period: 2015–2099, analyzed across three timeframes:
    • Near-future: 2020–2039
    • Mid-future: 2040–2069
    • Far-future: 2070–2099
  • Monthly rainfall data.

Methodology and Data

• Models used:
  • Downscaling method: Chaos theory-based local approximation approach.
  • General Circulation Models (GCMs): Five CMIP6 GCMs (CMCC-ESM2, CESM2-WACCM, EC-Earth3-Veg, CMCC-CM2-SR5, and BCC-CSM2-MR).
• Data sources:
  • Observed precipitation: Daily gridded rainfall data (0.25° × 0.25°) from the India Meteorological Department (IMD) for 1961–2014.
  • GCM data: Outputs from five CMIP6 GCMs accessed via the Earth System Grid Federation (ESGF) portal.
  • Predictor variables: Eight climate variables (eastward wind, northward wind, relative humidity, specific humidity, surface temperature, air temperature, sea level pressure, and geopotential height) at 100 Pa pressure level.
  • Scenarios: Four Shared Socioeconomic Pathways (SSPs): SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5.

Main Results

• Southwest Monsoon (June–September):
  • Near-future (2020–2039): Projected notable decreases in rainfall across eastern India and the Western Ghats, particularly under high-emission scenarios (SSP2-4.5 and SSP3-7.0). Some regions in western India are expected to experience increases.
  • Mid-future (2040–2069): Decreases are projected to intensify and expand across eastern India and the Western Ghats, with SSP3-7.0 and SSP5-8.5 showing the largest reductions. The percentage of grids with a projected decrease exceeding 50 mm/month increases for some models compared to the near-future.
  • Far-future (2070–2099): These patterns are projected to persist, with even greater reductions in eastern India and the Western Ghats. Modest increases are projected under SSP1-2.6 for most models.
• Northeast Monsoon (October–December):
  • Near-future (2020–2039): Increases are projected over northern, northeastern, and Western Ghats regions across all scenarios, especially under SSP3-7.0 and SSP5-8.5. Southeast India (e.g., Tamil Nadu) might experience declines.
  • Mid-future (2040–2069): Increases are projected to expand further, particularly for SSP3-7.0 and SSP5-8.5.
  • Far-future (2070–2099): The percentage of grids with projected increases in northeast monsoon rainfall exceeding 50 mm/month is higher than in the mid-future across all SSP scenarios and GCMs, especially under high-emission scenarios.
• Overall: The study indicates a potential shift in seasonal water availability, with India expected to witness a greater decrease in southwest monsoon rainfall in the near-future, while changes in northeast monsoon rainfall are comparatively modest.

Contributions

• This study represents the first application of the chaos theory-based local approximation downscaling approach for future rainfall projections from GCMs.
• It provides high-resolution (0.25° × 0.25°) future rainfall projections for India using an ensemble of five CMIP6 GCMs under four SSP scenarios.
• It offers detailed analysis of projected changes in both southwest and northeast monsoon rainfall across three distinct future timeframes (near-, mid-, and far-future).
• The findings highlight critical implications for regional water resources assessment and management strategies in India, particularly concerning potential shifts in seasonal water availability.

Funding

This research received no specific grant from funding agencies in public, commercial, or not-for-profit sectors.

Citation

@article{Deepthi2026Chaos,
  author = {Deepthi, B. and Sivakumar, Bellie},
  title = {Chaos Theory-Based Downscaling for Future Rainfall Projections from Climate Models},
  journal = {Water Resources Management},
  year = {2026},
  doi = {10.1007/s11269-025-04382-5},
  url = {https://doi.org/10.1007/s11269-025-04382-5}
}