Document Type : Original Article
Authors
1 Assistant Professor, Department of Water Science and Engineering, Faculty of Agriculture and Natural Resources,Imam Khomeini International University, Qazvin, Iran
2 PhD in Irrigation and Drainage, Department of Water Science and Engineering, Faculty of Agriculture and Natural Resources, Imam Khomeini International University, Qazvin, Iran
3 Associate Professor, Department of Water Science and Engineering, Faculty of Agriculture and Natural Resources,Imam Khomeini International University, Qazvin, Iran
Abstract
Abstract
Accurate estimation of cultivated area is essential for agricultural planning and water resource management. The integration of remote sensing data and machine learning algorithms can enhance the accuracy of such estimations. This study evaluates the effectiveness of the Random Forest algorithm in estimating the dominant cultivated areas by integrating 10-meter resolution Sentinel-1 and Sentinel-2 data within the Qazvin Plain irrigation network. First, the Jeffries-Matusita separability test was conducted to assess the spectral discrimination of six land use classes (wheat-barley/maize, alfalfa, fallow, bare_land, and urban areas) during Winter and Spring cropping seasons. The results indicated high separability for some classes, while others exhibited spectral overlap. Subsequently, the combination of radar and optical data with the Random Forest algorithm significantly improved classification accuracy. In the Winter, the classification achieved a kappa coefficient of 0.99 and an overall accuracy of 99.69%, while in the spring cropping season, the kappa coefficient was 0.98, with an overall accuracy of 98.93%. An analysis of cultivated area trends for wheat-barley, maize, and alfalfa over the past decade revealed an increase in maize cultivation and a decline in alfalfa during the spring season. In the Winter, wheat and barley cultivation expanded, likely due to their relatively higher resilience to climatic fluctuations and economic incentives. The analysis of cultivated area changes within the irrigation network indicated a general shift towards increased maize cultivation, with a growing preference for spring cropping over Winter cropping.
Keywords
Main Subjects
)