Elmotawakkil et al. (2025) Machine Learning and Remote Sensing for Soil Moisture Prediction

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

Identification

• Journal: Advances in geospatial technologies book series
• Year: 2025
• Date: 2025-10-17
• Authors: Abdessamad Elmotawakkil, Saad Jaldi, Mohammed Bouhassane, Adil Moumane, Nourddine Enneya
• DOI: 10.4018/979-8-3373-6608-1.ch007

Research Groups

Not explicitly mentioned in the provided text.

Short Summary

This study introduces an AI-driven framework to forecast soil moisture across five sites in Morocco's Draa Valley, demonstrating that tree-based models (Random Forest, XGBoost, CatBoost) significantly outperformed deep learning models with high accuracy.

Objective

• To introduce and evaluate an AI-driven framework for forecasting soil moisture across five sites in Morocco's Draa Valley to enhance irrigation scheduling, optimize water allocation, and support climate-resilient farming.

Study Configuration

• Spatial Scale: Five sites in Morocco's Draa Valley (Agdz, Tagounite, Tamegroute, Tansikht, Zagora).
• Temporal Scale: Dataset spanning from 2003 to 2024.

Methodology and Data

• Models used: Random Forest (RF), XGBoost, CatBoost, k-Nearest Neighbors (KNN), Long Short-Term Memory (LSTM), Temporal Convolutional Networks (TCN).
• Data sources: Historical climate records and remote sensing indicators, integrated using Google Earth Engine (GEE).

Main Results

• Tree-based models (Random Forest, XGBoost, CatBoost) clearly outperformed deep learning models in soil moisture prediction.
• The top-performing tree-based models achieved a Root Mean Square Error (RMSE) ranging from 2.89% to 9.11%.
• These models also demonstrated a Nash-Sutcliffe Efficiency (NSE) greater than 0.965.

Contributions

• Introduction of a novel AI-driven framework for accurate soil moisture prediction specifically tailored for arid regions facing water scarcity and climate variability.
• Empirical demonstration of the superior performance of tree-based machine learning models over deep learning architectures for this specific application.
• Offers a scalable and effective solution for precision agriculture, supporting optimized irrigation scheduling and water allocation in climate-resilient farming.

Funding

Not explicitly mentioned in the provided text.

Citation

@article{Elmotawakkil2025Machine,
  author = {Elmotawakkil, Abdessamad and Jaldi, Saad and Bouhassane, Mohammed and Moumane, Adil and Enneya, Nourddine},
  title = {Machine Learning and Remote Sensing for Soil Moisture Prediction},
  journal = {Advances in geospatial technologies book series},
  year = {2025},
  doi = {10.4018/979-8-3373-6608-1.ch007},
  url = {https://doi.org/10.4018/979-8-3373-6608-1.ch007}
}