Tilahun et al. (2026) Topographic modulation of drought propagation establishes low-elevation hotspots and mid-elevation climatic refugia in Southern Africa

Identification

• Journal: Ecological Indicators
• Year: 2026
• Date: 2026-03-30
• Authors: Minyahel Tilahun, Solomon G. Tesfamichael, Takehiro Sasaki, Ayana Angassa
• DOI: 10.1016/j.ecolind.2026.114807

Research Groups

• Department of Geography, Environmental Management and Energy Studies, University of Johannesburg, Johannesburg, South Africa
• College of Agriculture and Natural Resources, Wolkite University, Wolkite 07, Ethiopia
• Graduate School of Environment and Information Sciences, Yokohama National University, Yokohama, Japan
• Institute for Multidisciplinary Sciences, Yokohama National University, Yokohama, Japan
• Department of Range and Forest Resources, Botswana University of Agriculture and Natural Resources, Gaborone, Botswana

Short Summary

This study quantifies the topographic modulation of drought propagation in Southern Africa, revealing that mid-elevation zones act as climatic refugia with slow drought propagation, while low and high elevations are vulnerable hotspots experiencing accelerated aridification, a pattern projected to intensify.

Objective

• To investigate how variations in topography regulate drought propagation across different elevation gradients, examining mechanisms contributing to lowland vulnerability versus mid-elevation buffering, and projecting their evolution over 30 years.
• To examine how characteristic timescales associated with drought intensification vary spatially across the region's climatic gradients and how these timescales evolve from meteorological to hydrological.
• To identify dominant drought timescales and spatial hotspots, informed by future projections, to enhance early-warning systems and guide targeted adaptation strategies.

Study Configuration

• Spatial Scale: Southern Africa (Angola, Botswana, Eswatini, Lesotho, Namibia, South Africa, and Zimbabwe), 0.5° (resampled to 1 km) spatial resolution.
• Temporal Scale: Historical analysis from 1950 to 2022; future projections from 2023 to 2052 (30 years). Drought timescales analyzed: 1, 3, 6, 9, 12, and 24 months.

Methodology and Data

• Models used: Standardized Precipitation Evapotranspiration Index (SPEI), run theory (for drought event identification), Mann-Kendall test (for trend analysis), Theil-Sen slope estimator (for trend magnitude), linear regression (for decadal trends), Kruskal-Wallis test (for significant differences across elevations), Dunn's test (post-hoc for specific differences), pixel-specific Theil-Sen trend (βhist) for future SPEI projection, and a deterministic seasonal adjustment term (εseas) and topographic modulation parameter (ωpars) in the projection model.
• Data sources: Global SPEI database (SPEIbase v.2.10), Shuttle Radar Topography Mission (SRTMGL1_003) for elevation, Climate Hazards Group Infrared Precipitation with Station data (CHIRPS Daily, v.2.0) for Mean Annual Precipitation (MAP), ERA5-Land (ERA5-L) for Mean Annual Temperature (MAT), MODIS Terra LST (MOD11A1.061), and Global Surface Water Dataset. Data acquisition and processing were conducted in Google Earth Engine (GEE).

Main Results

• The most severe long-term aridification (SPEI24 trends, r = 0.6) occurs at higher elevations (2000–2500 m).
• Mid-elevation zones (1000–1500 m) act as hydroclimatic buffers, exhibiting the slowest drought propagation rates and a decadal-scale shift toward shorter drought timescales, a pattern projected to intensify.
• Low-elevation zones are projected to accelerate toward more persistent drought conditions, with the fastest drought propagation observed in low-elevation and regional hotspots (e.g., Angola and Namibia).
• A significant post-2000 expansion in the dominance of short-term droughts (SPEI01, SPEI03) was identified regionally, particularly in the northern and eastern regions during the 2000s (Relative Contribution, RC = 0.8) and 2010s (RC = 0.6).
• Drought development is fundamentally controlled by the interaction of timescale and topography.
• Future projections (2023–2052) indicate that high-elevation areas (>2000 m) will experience the strongest absolute drying (−0.10%/month), with 67% of pixels showing significant trends, while low-elevations (<1000 m) show accelerated aridification (−0.15%/month, p < 0.01). Mid-elevations (1000–2000 m) show notable resilience (−0.02%/month, p > 0.1).
• Drought trends exhibit critical seasonal narrowing, with intensification concentrated at the rainy season's onset (October–December), posing risks to crop establishment.

Contributions

• Quantifies the critical mechanism of topographic modulation on drought propagation across elevation gradients in Southern Africa.
• Delivers a critical advancement in understanding hydroclimatic change through multi-scale drought characterization (1950–2022) and 30-year multi-timescale SPEI projection.
• Establishes a novel methodological standard for drought characterization and projection, revealing previously undocumented elevation-dependent drought propagation patterns.
• Identifies specific drought hotspots and the temporal reorganization of seasonal aridity, providing a mechanistic framework for predicting climate impacts across tropical regions worldwide.
• Offers urgently needed insights for climate adaptation policy and conservation planning by revealing fine-scale topographic controls and future drought trajectories, enabling precisely targeted strategies and averting maladaptation.
• Proposes a novel conceptual framework explaining the topographic partitioning of drought regimes, integrating orographic buffering, hotspot development, differential propagation, and topographic mediation of patterns and trajectories.

Funding

Not explicitly stated in the provided text.

Citation

@article{Tilahun2026Topographic,
  author = {Tilahun, Minyahel and Tesfamichael, Solomon G. and Sasaki, Takehiro and Angassa, Ayana},
  title = {Topographic modulation of drought propagation establishes low-elevation hotspots and mid-elevation climatic refugia in Southern Africa},
  journal = {Ecological Indicators},
  year = {2026},
  doi = {10.1016/j.ecolind.2026.114807},
  url = {https://doi.org/10.1016/j.ecolind.2026.114807}
}