Overview of soil organic carbon mapping using machine learning algorithms in Africa
This review provides an overview of the application of machine learning (ML) for Soil Organic Carbon (SOC) and Soil Organic Matter (SOM) mapping in African countries, highlighting its significance for agricultural productivity and environmental sustainability. In this review, studies conducted using...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
National Institute of Agronomic Research "INRA" Morocco
2024-12-01
|
| Series: | African and Mediterranean Agricultural Journal - Al Awamia |
| Online Access: | https://revues.imist.ma/index.php/Afrimed/article/view/49544 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This review provides an overview of the application of machine learning (ML) for Soil Organic Carbon (SOC) and Soil Organic Matter (SOM) mapping in African countries, highlighting its significance for agricultural productivity and environmental sustainability. In this review, studies conducted using geostatistical approaches without the use of ML algorithms, and those focusing on SOC stocks, were not considered. We identified 20 studies, with a notable emphasis on local research, including 5 studies in Morocco, 4 at the national scale, and 3 across the African continent. Random Forest (RF) model was the most frequently used and effective. Key predictors included spectral bands and vegetation indices, primarily derived from Landsat, elevation, and climatic variables from global or national databases. Despite promising results, challenges such as limited data availability and resistance to new methodologies persist. Opportunities for future research include focusing on small field studies, utilizing high-resolution remote sensing data, and exploring advanced ML techniques. This review aims to guide future research, foster collaboration, and optimize ML applications in SOC and SOM mapping for the advancement of African agriculture and sustainability. |
|---|---|
| ISSN: | 0572-2721 2658-9184 |