|
Assessment of sugarcane carbon sequestration using UAV-based data |
|---|---|
| รหัสดีโอไอ | |
| Creator | Arthit Phuphaphud |
| Title | Assessment of sugarcane carbon sequestration using UAV-based data |
| Contributor | Premphawika Sawatwongchai, Kanin Kamma, Waraporn Sawangchom, Piyada Satjasang, Kittipong Laloon, Khwantri Saengprachatanarug, Jetsada Posom |
| Publisher | Faculty of Engineering, Khon Kaen University |
| Publication Year | 2569 |
| Journal Title | Agricultural and Biological Engineering |
| Journal Vol. | 3 |
| Journal No. | 2 |
| Page no. | 55-61 |
| Keyword | Sugarcane, Carbon sequestration, Unmanned aerial vehicle (UAV), Canopy height model, Machine learning, Allometric equation |
| URL Website | https://ph04.tci-thaijo.org/index.php/abe |
| Website title | ThaiJo |
| ISSN | 3056-932X (Online) |
| Abstract | Sugarcane is a major economic crop in Thailand and has a high potential for carbon sequestration through biomass accumulation, which can contribute to the reduction of greenhouse gas emissions from the agricultural sector. However, the assessment of carbon stock in sugarcane at the field scale remains challenging due to the reliance on field-based measurements that require considerable time, labor, and resources. These limitations restrict large-scale spatial assessment and long-term monitoring. Although standard equations for estimating carbon stock from sugarcane biomass are available, their application in combination with spatial data is still limited and lacks flexibility for practical use. This study aimed to develop a Machine Learning model for estimating carbon sequestration in sugarcane using data obtained from varietal test plots, with the Khon Kaen 3 cultivar selected as a case study. Unmanned aerial vehicle (UAV) imagery was integrated with empirical equations established by the Department of Agriculture, which were used as a reference for carbon calculation. UAV data were processed to generate Digital Surface Models (DSM) and Digital Terrain Models (DTM), from which the Canopy Height Model (CHM) was derived. Structural variables of sugarcane at the clump level were extracted from the CHM and used as input variables for Machine Learning model development. The results indicated that the maximum canopy height showed a strong positive relationship with sugarcane carbon stock. The Artificial Neural Network model provided the best performance, with a coefficient of determination (R²) of approximately 0.84 and a root mean square error (RMSE) of about 1.6 tons of carbon per rai. These findings demonstrate that the proposed approach effectively links UAV-derived structural information with carbon stock estimation and extends the applicability of standard carbon calculation equations to spatial-scale assessments. The approach can support carbon management and promote sustainable agricultural development in the future. |
