Sseguya et al. (2024) Drought Quantification in Africa Using Remote Sensing, Gaussian Kernel, and Machine Learning

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

Identification

• Journal: Water
• Year: 2024
• Date: 2024-09-18
• Authors: Fred Sseguya, Kyung Soo Jun
• DOI: 10.3390/w16182656

Research Groups

[Not specified]

Short Summary

This study employs remote sensing data and machine learning to refine meteorological, agricultural, and hydrological drought indices across Africa, identifying the Classification and Regression Tree (CART) model as the most accurate for drought prediction.

Objective

• To improve the precision and classification of drought indices (RDI, SMADI, and SDI) using Gaussian kernel denoising and machine learning models to overcome the limitations of limited ground data.

Study Configuration

• Spatial Scale: Africa (with specific focus on the Sahel, North Africa, and southwestern Africa)
• Temporal Scale: 2001–2020

Methodology and Data

• Models used: Gaussian kernel convolution, Classification and Regression Tree (CART), Support Vector Machine (SVM), and Random Forest (RF).
• Data sources: Remote sensing data.

Main Results

• Model Performance: CART achieved the highest accuracy with Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) values between 0 and 0.4; RF ranked second, and SVM was the least reliable (values < 0.7).
• Drought Prevalence: Meteorological drought affected 30% of Africa, agricultural drought affected 22%, and hydrological drought affected 21%.
• Regional Impact: Persistent drought conditions were identified in the Sahel, North Africa, and southwestern Africa.

Contributions

• Developed a refined approach to drought monitoring by combining Gaussian kernel-based denoising of multi-band composite images with machine learning to enhance the accuracy of the Reconnaissance Drought Index (RDI), Soil Moisture Agricultural Drought Index (SMADI), and Streamflow Drought Index (SDI).

Funding

[Not specified]

Citation

@article{Sseguya2024Drought,
  author = {Sseguya, Fred and Jun, Kyung Soo},
  title = {Drought Quantification in Africa Using Remote Sensing, Gaussian Kernel, and Machine Learning},
  journal = {Water},
  year = {2024},
  doi = {10.3390/w16182656},
  url = {https://doi.org/10.3390/w16182656}
}