Multi-Temporal Passive and Active Remote Sensing for Agricultural Mapping and Acreage Estimation in Context of Small Farm Holds in Ethiopia

Tesfamariam Engida Mengesha,Lulseged Tamene Desta,Paolo Gamba,Getachew Tesfaye Ayehu
DOI: https://doi.org/10.3390/land13030335
IF: 3.905
2024-03-07
Land
Abstract:In most developing countries, smallholder farms are the ultimate source of income and produce a significant portion of overall crop production for the major crops. Accurate crop distribution mapping and acreage estimation play a major role in optimizing crop production and resource allocation. In this study, we aim to develop a spatio–temporal, multi-spectral, and multi-polarimetric LULC mapping approach to assess crop distribution mapping and acreage estimation for the Oromia Region in Ethiopia. The study was conducted by integrating data from the optical and radar sensors of sentinel products. Supervised machine learning algorithms such as Support Vector Machine, Random Forest, Classification and Regression Trees, and Gradient Boost were used to classify the study area into five first-class common land use types (built-up, agriculture, vegetation, bare land, and water). Training and validation data were collected from ground and high-resolution images and split in a 70:30 ratio. The accuracy of the classification was evaluated using different metrics such as overall accuracy, kappa coefficient, figure of metric, and F-score. The results indicate that the SVM classifier demonstrates higher accuracy compared to other algorithms, with an overall accuracy for Sentinel-2-only data and the integration of optical with microwave data of 90% and 94% and a kappa value of 0.85 and 0.91, respectively. Accordingly, the integration of Sentinel-1 and Sentinel-2 data resulted in higher overall accuracy compared to the use of Sentinel-2 data alone. The findings demonstrate the remarkable potential of multi-source remotely sensed data in agricultural acreage estimation in small farm holdings. These preliminary findings highlight the potential of using multi-source active and passive remote sensing data for agricultural area mapping and acreage estimation.
environmental studies
What problem does this paper attempt to address?
### Problems the Paper Aims to Solve This paper aims to address the issue of mapping and estimating the area of agricultural land in small-scale farms in Ethiopia. Specifically: 1. **Background Issues**: - In many developing countries, small-scale farmers are the main source of food and produce a large amount of crops. - Accurate crop distribution mapping and area estimation are key to optimizing crop production and resource allocation. - Small-scale farms in Ethiopia account for 96% of the total arable land, but there is currently a lack of regularly updated high-resolution crop distribution maps. 2. **Research Objectives**: - Utilize multi-temporal passive and active remote sensing data (Sentinel-1 and Sentinel-2 data) for land use and cover classification. - Develop a spatio-temporal, multispectral, and multi-polarization land use classification method to assess crop distribution mapping and area estimation in the Oromia region of Ethiopia. - Use supervised machine learning algorithms (Support Vector Machine, Random Forest, Classification and Regression Trees, and Gradient Boosting) to classify the study area. - Validate the classification accuracy of different algorithms under the integration of optical and radar data. 3. **Specific Issues**: - How to improve the accuracy of crop distribution mapping for small-scale farms? - How to use multi-source remote sensing data (optical and radar) to improve the accuracy of agricultural land area estimation? - How do different machine learning algorithms perform in handling complex agricultural landscapes? Through this research, the paper demonstrates the potential of multi-source remote sensing data in estimating the area of agricultural land in small-scale farms and validates the effectiveness of different machine learning algorithms.