Kreri et al. (2026) Remote sensing assessment of vegetation and moisture dynamics in semi-arid regions

Identification

• Journal: Scientific Reports
• Year: 2026
• Date: 2026-01-28
• Authors: Sarah Kreri, Nezha Farhi, Ahmed Bennia, Abdessamed Derdour, Lahsen Wahib Kébir, Khalid Alharbi, Amanuel Kumsa Bojer, Ahmed Arafat
• DOI: 10.1038/s41598-026-37781-8

Research Groups

• Centre des Techniques Spatiales, Agence Spatiale Algérienne, Arzew, Algeria
• Laboratory of Environmental and Energy Systems (LSEE), University of Tindouf, Algeria
• Artificial Intelligence Laboratory for Mechanical and Civil Structures and Soil, University of Naama, Algeria
• Laboratory for the Sustainable Management of Natural Resources in Arid and Semi-Arid Zones, University of Naama, Algeria
• Department of Geography, College of Engineering and Architecture, Umm Al-Qura University, Makkah, Saudi Arabia
• Ethiopian Artificial Intelligence Institute, Addis Ababa, Ethiopia
• Department of Civil Engineering, College of Engineering, Taif University, Taif, Saudi Arabia

Short Summary

This study assessed spatio-temporal changes in land use and vegetation cover in the Oued Louza watershed, Algeria, from 1987 to 2020 using remote sensing and GIS. It revealed a dramatic decline in vegetation cover (from 42% to 10%) and significant urban expansion (by 27%), driven by both climatic variability and anthropogenic pressures.

Objective

• To examine the spatio-temporal changes in vegetation cover and wetland areas within the Oued Louza watershed (Sidi Bel Abbès, Algeria) from 1987 to 2020.
• To assess the impact of climatic variability (precipitation, relative humidity) and anthropogenic factors on these observed changes.
• To evaluate the effectiveness of integrating multiple remote sensing indices (NDVI, SAVI, NDWI, TWI) with a majority voting technique for enhancing land cover classification accuracy in a semi-arid region.

Study Configuration

• Spatial Scale: Oued Louza watershed, Sidi Bel Abbès province, Algeria, covering an area of 746 square kilometers.
• Temporal Scale: 33-year period, from 1987 to 2020.

Methodology and Data

• Models used:
  • Normalized Difference Vegetation Index (NDVI)
  • Soil Adjusted Vegetation Index (SAVI)
  • Normalized Difference Water Index (NDWI)
  • Topographic Wetness Index (TWI)
  • Maximum Likelihood Classification (MLC)
  • Majority Voting technique (for combining spectral indices)
  • Kappa coefficient (for classification validation)
  • Linear regression (for climatic correlation analysis)
• Data sources:
  • Satellite imagery: Landsat 5 Thematic Mapper (TM) (1987) and Landsat 8 Operational Land Imager (OLI) (2020) from USGS Earth Explorer (30 m spatial resolution).
  • Digital Elevation Model (DEM): Shuttle Radar Topography Mission (SRTM) (30 m spatial resolution) from USGS database.
  • Climatic data: NASA POWER database (annual precipitation, mean temperature, relative humidity) for the Sidi Bel Abbès region (1987-2020).
  • Validation data: Google Earth imagery and Copernicus Sentinel-2 data (for 2020), historical topographic maps (for 1987).
  • Software: ArcGIS 10.7/10.8, ENVI 5.1, ERDAS Imagine, Excel, Origin 2020.

Main Results

• Vegetation cover in the Oued Louza watershed decreased from 42% in 1987 to 10% in 2020, representing a 32% reduction.
• Urban areas expanded significantly from 1% in 1987 to 27% in 2020, an increase of 26%.
• Natural areas increased from 36% to 46%, while forest cover slightly decreased from 21% to 17%.
• Climatic analysis revealed a 22.7% decrease in precipitation and a 6.6% decrease in relative humidity between 1987 and 2020, alongside a 4.41% increase in temperature.
• A positive correlation (R² = 0.6132, r = 0.78) was observed between annual precipitation and relative humidity.
• NDWI maps indicated a widespread drying trend, with maximum values decreasing from 0.455 in 1987 to 0.16 in 2020.
• High TWI values correlated with humid zones and denser vegetation, while low TWI values corresponded to drier areas with rapid runoff.
• Land use/land cover classification achieved high accuracy, with Kappa coefficients of 0.84 for 1987 and 0.91 for 2020, validating the multi-index majority voting approach.

Contributions

• This study is the first to integrate NDVI, SAVI, NDWI, and TWI indices into a majority-voting model for land-cover classification in the Oued Louza watershed, demonstrating its effectiveness in improving classification accuracy.
• It provides crucial insights into the complex interplay of climatic variability and anthropogenic pressures driving significant land use and vegetation cover changes in a semi-arid region of Algeria.
• The research highlights the robust potential of remote sensing and GIS technologies for comprehensive environmental monitoring and assessment in data-scarce semi-arid environments.
• The findings offer essential information for developing sustainable land and water resource management strategies in the Oued Louza watershed and similar vulnerable regions.

Funding

• Deanship of Graduate Studies and Scientific Research, Taif University.

Citation

@article{Kreri2026Remote,
  author = {Kreri, Sarah and Farhi, Nezha and Bennia, Ahmed and Derdour, Abdessamed and Kébir, Lahsen Wahib and Alharbi, Khalid and Bojer, Amanuel Kumsa and Arafat, Ahmed},
  title = {Remote sensing assessment of vegetation and moisture dynamics in semi-arid regions},
  journal = {Scientific Reports},
  year = {2026},
  doi = {10.1038/s41598-026-37781-8},
  url = {https://doi.org/10.1038/s41598-026-37781-8}
}