Achemrk et al. (2026) Tracking Quarter-Century Spatio-Temporal Soil Salinization Dynamics in Semi-Arid Landscapes Using Earth Observation and Machine Learning

⚠️ Warning: This summary was generated from the abstract only, as the full text was not available.

Identification

• Journal: Remote Sensing
• Year: 2026
• Date: 2026-02-26
• Authors: Aiman Achemrk, Jamal-Eddine Ouzemou, Ahmed Laamrani, Ali El Battay, Soufiane Hajaj, Sabir Oussaoui, Abdelghani Chehbouni
• DOI: 10.3390/rs18050687

Research Groups

Not explicitly mentioned in the provided text.

Short Summary

This study developed an integrated framework combining multi-temporal remote sensing and machine learning to model soil salinity dynamics in the Sehb El Masjoune (SEM) semi-arid region, Morocco. The framework revealed a nearly 10% expansion of moderately to highly saline areas from 2000 to 2025, primarily driven by recurrent droughts and inefficient drainage.

Objective

• To develop an integrated framework uniting multi-temporal Landsat imagery, hypsometric data, climatic indicators, and in situ soil electrical conductivity (ECe) measurements to model soil salinity dynamics using machine learning (ML) over the Sehb El Masjoune (SEM) semi-arid region, Morocco.

Study Configuration

• Spatial Scale: Sehb El Masjoune (SEM) semi-arid region, Morocco.
• Temporal Scale: 2000–2025 for multi-temporal Landsat imagery analysis; in situ soil samples collected in 2022, 2023, 2024, and 2025.

Methodology and Data

• Models used: Machine Learning (ML) algorithms, specifically Gradient-Boosted Trees (GBT), Support Vector Regression (SVR), and Random Forest (RF).
• Data sources: Multi-temporal Landsat imagery (2000–2025), hypsometric data, climatic indicators (e.g., Standardized Precipitation Index - SPI), and 233 in situ soil electrical conductivity (ECe) measurements collected between 2022 and 2025.

Main Results

• Support Vector Regression (SVR) achieved the highest predictive capability with a coefficient of determination (R2) of 0.76 and a Root Mean Square Error (RMSE) of 32.91 dS/m.
• SVR-based salinity maps showed a distinct spatial organization, with extremely saline soils (≥64 dS/m) concentrated in the central study area and a progressive decline toward adjacent agricultural lands (0–8 dS/m).
• From 2000 to 2025, moderately to highly saline areas (≥16 dS/m) expanded by nearly 10%, primarily driven by recurrent droughts and inefficient drainage.
• Hydroclimatic analysis confirmed that dry years (SPI ≤ −0.5) promoted net salinity build-up through the expansion and persistence of moderate-to-high salinity classes (≥16 dS/m).
• Wet years (SPI ≥ +0.5) favored temporary leaching and partial recovery, mainly within the low-to-moderate salinity range.

Contributions

• Developed a robust and scalable integrated remote sensing–machine learning framework for operational soil salinity monitoring.
• Provided valuable insights into the spatio-temporal dynamics of soil salinization over a 25-year period in a semi-arid Sabkha agroecosystem.
• Offered a practical tool for sustainable land-use planning in data-scarce regions facing similar salinization challenges.

Funding

Not explicitly mentioned in the provided text.

Citation

@article{Achemrk2026Tracking,
  author = {Achemrk, Aiman and Ouzemou, Jamal-Eddine and Laamrani, Ahmed and Battay, Ali El and Hajaj, Soufiane and Oussaoui, Sabir and Chehbouni, Abdelghani},
  title = {Tracking Quarter-Century Spatio-Temporal Soil Salinization Dynamics in Semi-Arid Landscapes Using Earth Observation and Machine Learning},
  journal = {Remote Sensing},
  year = {2026},
  doi = {10.3390/rs18050687},
  url = {https://doi.org/10.3390/rs18050687}
}