İpek et al. (2025) Integrating Spatio-Probabilistic mapping and remote sensing for comprehensive drought risk assessment

Identification

• Journal: Modeling Earth Systems and Environment
• Year: 2025
• Date: 2025-11-25
• Authors: Ahmet Faruk İpek, Ercan Kahya
• DOI: 10.1007/s40808-025-02696-2

Research Groups

• Civil Engineering Faculty, Civil Engineering Department, Hydraulics and Water Resources Engineering PhD Program, Istanbul Technical University, İstanbul, Turkey
• Civil Engineering Faculty, Civil Engineering Department, Istanbul Technical University, İstanbul, Turkey

Short Summary

This study developed a Spatio-Probabilistic Drought Mapping (SPDM) framework by integrating multiple drought indices with remote sensing and land cover analysis to assess drought dynamics and environmental impacts in the Küçük Menderes Water Basin. The research identified western regions as high-risk areas and quantified severe impacts on vegetation, agriculture, and forest fires during major drought episodes.

Objective

• To analyze drought dynamics and environmental impacts in the Küçük Menderes Water Basin using an integrated approach combining multiple drought indices (SPI, SPEI, PDSI) with vegetation (NDVI) and land cover (CORINE) analysis.
• To introduce and apply a Spatio-Probabilistic Drought Mapping (SPDM) framework that consolidates all drought severity levels into unified probability surfaces for regional-scale risk assessment.
• To examine drought severity, duration, spatial distribution, and trends, and assess its impacts on agriculture, forest fires, livestock, and land cover changes within the basin.
• To evaluate the basin’s overall sensitivity to drought and analyze the correlations among SPI, SPEI, and PDSI to provide a comprehensive understanding of drought dynamics.

Study Configuration

• Spatial Scale: Küçük Menderes Water Basin, western Turkey (Aegean Region), with a total area of 7029.31 square kilometers. Analysis focused on 11 meteorological stations within and near the basin.
• Temporal Scale: Meteorological data from 1990–2020 (30 years); NDVI analysis for August 2000, 2005, 2010, 2015, and 2020; CORINE Land Cover data for 1990 and 2018; agricultural production and livestock data from 1991–2018; forest fire data from 1991–2021.

Methodology and Data

• Models used:
  • Standardized Precipitation Index (SPI)
  • Standardized Precipitation Evapotranspiration Index (SPEI)
  • Palmer Drought Severity Index (PDSI)
  • Spatio-Probabilistic Drought Mapping (SPDM) framework (integrating drought event categorization, Inverse Distance Weighting (IDW) interpolation, and Natural Breaks (Jenks) optimization)
  • Normalized Difference Vegetation Index (NDVI) calculation and image differencing for change detection
  • CORINE Land Cover Classification System
• Data sources:
  • Meteorological data (monthly precipitation and temperature) from 11 observation stations (1990–2020) provided by the Turkish State Meteorological Service.
  • Clear-sky Moderate Resolution Imaging Spectroradiometer (MODIS) satellite data (August 2000, 2005, 2010, 2015, 2020) for NDVI maps, obtained from NASA’s MODIS data portal.
  • CORINE Land Cover Classification System vector data (1990 and 2018) from the Joint Research Centre (JRC) of the European Commission.
  • Agricultural production and livestock data (1991–2018) from the Izmir Provincial Directorate of Agriculture and Forestry.
  • Forest fire statistics (number of fires and burned area, 1991–2021) from the General Directorate of Forestry, Forest Fire Fighting Department for Izmir province.

Main Results

• The western regions of the Küçük Menderes Basin (Kuşadası, Çeşme-Karaburun, and İzmir-Körfez) were identified as the most drought-prone areas.
• Major drought episodes occurred in 1992–1994, 2000–2001, 2007–2008, 2017, and 2018, with the 1992–1994 and 2007–2008 periods being particularly severe and widespread.
• Average drought occurrence probabilities over 30 years were 16.39% for SPI, 18.9% for SPEI, and 15.12% for PDSI, indicating approximately 5 drought years within the 30-year period.
• NDVI analysis (2000–2020) revealed a negative change in vegetation cover across 297.28 square kilometers (4.33% of the basin), primarily impacting grasslands, pastures, and agricultural areas, while 70.33 square kilometers (1.02%) showed positive changes, indicating an increase in forested areas.
• CORINE Land Cover data (1990–2018) showed a 92% increase in artificial/urban areas, an 8.36% decrease in agricultural areas, and a 15% reduction in wetlands. Conversely, water bodies increased by 62% due to new dam construction.
• Drought events led to significant agricultural losses, with crop yield reductions reaching up to 50% for cherries and approximately 25% for wheat and 37% for mandarins during severe drought periods (e.g., 1992, 2007–2008).
• A strong correlation was found between prolonged droughts and increased forest fire damage, particularly evident in 1994, 2000, and 2008.
• While complex due to policy variables, extreme droughts (1992–1994, 2007–2008) coincided with initial decreases in livestock numbers.
• Correlation analysis showed strong relationships between SPI and SPEI (0.80–1.00), and moderate to strong correlations between SPI and PDSI (0.60–0.79), with SPEI and PDSI showing stronger correlations in mid-range periods.

Contributions

• Introduction of the novel Spatio-Probabilistic Drought Mapping (SPDM) framework, which unifies multiple drought severity levels into integrated spatial risk surfaces, providing a simpler and more intuitive approach for regional-scale drought risk assessment.
• Development of a comprehensive, integrated assessment methodology combining multiple drought indices (SPI, SPEI, PDSI) with remote sensing (NDVI) and land cover change analysis (CORINE) to characterize drought dynamics and impacts.
• Generation of spatially explicit, high-resolution drought risk maps for the Küçük Menderes Basin, clearly delineating vulnerable areas.
• Quantification of the multi-faceted environmental and socio-economic impacts of drought on vegetation, land cover, agricultural production, and forest fires at a basin scale.
• Provides a robust framework for understanding drought-induced environmental changes and supports the development of proactive climate adaptation and sustainable resource management strategies for water-stressed regions.

Funding

This study did not receive any external funding or financial support.

Citation

@article{İpek2025Integrating,
  author = {İpek, Ahmet Faruk and Kahya, Ercan},
  title = {Integrating Spatio-Probabilistic mapping and remote sensing for comprehensive drought risk assessment},
  journal = {Modeling Earth Systems and Environment},
  year = {2025},
  doi = {10.1007/s40808-025-02696-2},
  url = {https://doi.org/10.1007/s40808-025-02696-2}
}