State-of-the-art in multispectral remote sensing for water body identification in agriculture

Abstract

Monitoring water in agricultural landscapes is essential for sustainable resource management, irrigation planning, and early flood detection. Recent advances in remote sensing, particularly the availability of high-resolution multispectral imagery from satellites and Unmanned Aerial Vehicles (UAVs), have greatly improved the capacity to detect and analyze water with higher spatial and temporal precision. This paper reviews 48 peer-reviewed studies published in the last six years, selected based on their focus on agricultural or non-urban areas, use of multispectral imagery, methodological detail, real-world validation, and scientific relevance. The analysis examines four main detection methodologies-spectral indices, decision trees, machine learning techniques, and neural networks-highlighting their respective strengths, limitations, and applicability to agricultural water monitoring. It also synthesizes information on platforms, sensors, and spectral bands most commonly used, as well as issues of data accessibility (free vs. commercial). The review consolidates the current state of the art in multispectral remote sensing for agricultural water detection and identifies key challenges and research opportunities to advance both methodological development and practical applications.