Fusing Landsat-8, Sentinel-1, and Sentinel-2 Data for River Water Mapping Using Multidimensional Weighted Fusion Method

River water extent is critical for understanding river discharge or its hydrological conditions. Although numerous methods have been proposed to map river water from either optical or synthetic aperture radar (SAR) remotely sensed images, uncertainties still exist broadly. In this study, we developed an image fusion method that integrates Landsat-8, Sentinel-1, and Sentinel-2 images simultaneously for river water mapping with two major steps. Firstly, a posterior probability support vector machine (SVM) model was adopted to generate water probability maps from each individual image; and second, a multidimensional weighted fusion method (MDWFM) was developed to fuse these probability maps. Four reaches with different characteristics were selected as case study sites. High-resolution aerial images were acquired and used as the reference to evaluate our results. We found that the fusion process not only improves the quality of river water mapping but also excludes the cloud interference. The fused river water maps become more reliable after the conflicts from difference images being solved by the proposed MDWFM method that contains a proportional conflict redistribution rule. The weighted root mean square difference was reduced to 0.066, and the area under the ROC curve reached up to 0.984. The critical success index (CSI), kappa coefficient (KC), and F-measure reached up to 0.810, 0.836, and 0.895, respectively. These stable and accurate river extent mapping results obtained through fusing multiple images with high spatial resolution (SR) (10 m) and short revisit interval (0.4–4.4 days) are of great significance for enriching the data and methodology of hydrological studies.

I. Introduction

As an important component of surface water, river is a pivotal part of the hydrological cycle on Earth. It also plays a critical role in the transportation of materials and energy around the world [1]. At global scale, rivers transport approximately 236 Tg of particulate nutrients (organic carbon, nitrogen, and phosphorus) to the ocean per year [2], supporting numerous ecosystems. Therefore, continuously and effectively monitoring river hydrological conditions is of important practical significance for the sustainable development of human society and ecosystems. Traditional river hydrological monitoring is point-based or section-based, which is unable to reflect regional hydrological characteristics. As an advanced Earth observation manner, remote-sensing technology provides effective approaches for monitoring rivers continuously at large scales.