Intelligent Remote Sensing Canal System Detection and Irrigation Water Use Estimation: A Case Study in the Transboundary Mekong River Basin

Human activities significantly impact global water resource availability through alterations in terrestrial water cycle processes, with agricultural irrigation being a primary driver. Accurately quantifying irrigation water use is essential for understanding regional water resource dynamics, optimizing water resource allocation, and improving agricultural productivity. However, high-quality data on irrigation canal networks is often lacking at regional scales, hindering the precise delineate of river sources for irrigation. To address this, this study developed a large-scale canal system detection method using artificial intelligence (AI) techniques and large-scale satellite remote sensing images. The method enabled the identification of canal networks, clarified the irrigation intake points, and facilitated the calculation of irrigation water volumes supplied by various mainstream and tributaries in the basin. The Mekong River Basin, where riparian states heavily rely on tributaries for irrigation and face difficulties in acquiring canal data, is selected as the study area. The results show that the developed Convolutional Neural Network (CNN)-based method successfully detected 291 irrigation canals sourced from mainstream and tributaries of the Mekong River, with 43% of the main canals drawing directly from the mainstream and the remainder from tributaries. Spatial analysis reveals a higher canal density in the south compared to the north of the basin. Additionally, irrigation water use is markedly higher during the dry season from November to the following April, accounting for 69% of annual irrigation consumption, peaking in January and reaching a minimum in September. This research has the potential to address critical data gaps in irrigation in the Mekong River Basin, enhance the understanding of agricultural irrigation water use, and provide essential insights for effective water resource management and sustainable agricultural development.