Special Section Guest Editorial: Intelligent Remote Sensing for Water Resources: Advances, Challenges, and Perspectives

Water, as a fundamental resource for life, food security, and ecosystem sustainability, is facing growing threats due to population growth, industrialization, land-use change, and climate variability.The increasing frequency and intensity of droughts, floods, and extreme weather events have intensified the global water crisis, challenging traditional hydrological monitoring and management systems.Remote sensing technologies, especially with the advent of high-resolution multispectral, hyperspectral, and radar satellites, have revolutionized the ability to monitor water bodies, hydrological states, and related environmental variables over vast and inaccessible regions.In recent years, the rapid advancement of artificial intelligence (AI), including deep learning, attention mechanisms, and semantic segmentation, has further transformed how remote sensing data is processed, interpreted, and applied.These intelligent techniques offer new levels of automation, accuracy, and insight, making them indispensable in modern water resources science and management.This special section aims to showcase the latest developments in satellite observation, intelligent algorithms, and data analysis approaches for water resources.Ten high-quality papers were selected through a rigorous peer-review process, highlighting novel techniques and significant applications from the following topics: ( 1) water body extraction and segmentation; (2) surface water mapping and classification; (3) flood monitoring and urban inundation detection; (4) water quality and environmental monitoring; (5) intelligent sensing for rainfall estimation.Xu et al. integrated synthetic aperture radar (SAR) backscatter intensity and polarimetric features to propose a dual-branch fusion network (DBFNet) for water body detection.Comparison results with five common deep learning models demonstrate the effectiveness of multi-feature fusion.Xiang et al. introduced multi-scale two-channel attention and multi-scale skip connections in a U-shaped network for pixel-level water segmentation, which enhances accurate tributary extraction in optical images.Wang et al. developed a new water index combination (WI2023) for Landsat-8 OLI data, which can eliminate the shadow between the water and non-water across diverse terrain types, providing optimal performance, especially in hilly regions.Koutsos et al. used Sentinel-1 SAR dual-polarization data (VV and VH), seven texture features, and a random forest (RF) classifier to map surface water in rice paddies and wetlands in Thessaloniki Plain, which achieved high accuracy even under complex visual conditions such as floating algae.Liang et al. utilized time-series C-band SAR data from Sentinel-1 and seasonal vegetation phenology to map wetland plant communities of Poyang Lake in 2019 by using RF classification, obtaining 88.5% classification accuracy with 0.821 Kappa coefficient.Yu et al. proposed an adaptive iterative thresholding segmentation method, guided by optical prior water body information to improve SAR-based flood mapping, which shows accuracy