Large-Scale River Mapping Using Contrastive Learning and Multi-Source Satellite Imagery

Zhihao Wei,Xiaowei Jia,Zhe Jiang,Yiqun Xie,Pengyu Liu,Kebin Jia

doi:10.3390/rs13152893

Abstract

River system is critical for the future sustainability of our planet but is always under the pressure of food, water and energy demands. Recent advances in machine learning bring a great potential for automatic river mapping using satellite imagery. Surface river mapping can provide accurate and timely water extent information that is highly valuable for solid policy and management decisions. However, accurate large-scale river mapping remains challenging given limited labels, spatial heterogeneity and noise in satellite imagery (e.g., clouds and aerosols). In this paper, we propose a new multi-source data-driven method for large-scale river mapping by combining multi-spectral imagery and synthetic aperture radar data. In particular, we build a multi-source data segmentation model, which uses contrastive learning to extract the common information between multiple data sources while also preserving distinct knowledge from each data source. Moreover, we create the first large-scale multi-source river imagery dataset based on Sentinel-1 and Sentinel-2 satellite data, along with 1013 handmade accurate river segmentation mask (which will be released to the public). In this dataset, our method has been shown to produce superior performance (F1-score is 91.53%) over multiple state-of-the-art segmentation algorithms. We also demonstrate the effectiveness of the proposed contrastive learning model in mapping river extent when we have limited and noisy data.

Highlights

River system plays a crucial role in global carbon circulation, as it delivers the carbonaceous matter within the global ecosystem and maintains the connection between ocean and land [1,2]
We have demonstrated the superiority of the proposed multi-source segmentation method over multiple widely used methods including water indicesbased method, traditional machine learning approaches and deep learning approaches
We compare with several baseline segmentation methods including a apopular popularwater water index-based approach, traditional machine learning approaches and index-based approach, traditional machine learning approaches and statestate-of-the-art deep learning approaches

Summary

Introduction

River system plays a crucial role in global carbon circulation, as it delivers the carbonaceous matter within the global ecosystem and maintains the connection between ocean and land [1,2]. Rivers are important in many countries due to the increasing demand to supply drinking water, irrigation and farming practices and power generation. Many existing methods utilize pre-defined water indices, such as Normalized Difference Water Index (NDWI) [5] and Modified Normalized Difference. Water Index (MNDWI) [6], which are computed from a subset of spectral bands. These indices are designed to enhance the water representation in contrast to other land covers based on the reflectance characteristics

Methods

Discussion

Conclusion