Spatiotemporal Data Augmentation of MODIS-Landsat Water Bodies Using Adversarial Networks

With increasing demands for precise water resource management, there is a growing need for advanced techniques in mapping water bodies. The currently deployed satellites provide complementary data that are either of high spatial or high temporal resolutions. As a result, there is a clear trade-off between space and time when considering a single data source. For the efficient monitoring of multiple environmental resources, various Earth science applications need data at high spatial and temporal resolutions. To address this need, many data fusion methods have been described in the literature, that rely on combining data snapshots from multiple sources. Traditional methods face limitations due to sensitivity to atmospheric disturbances and other environmental factors, resulting in noise, outliers, and missing data. This paper introduces Hydrological Generative Adversarial Network (Hydro-GAN), a novel machine learning-based method that utilizes modified GANs to enhance boundary accuracy when mapping low-resolution MODIS data to high-resolution Landsat-8 images. We propose a new non-saturating loss function for the Hydro-GAN generator, which maximizes the log of discriminator probabilities to promote stable updates and aid convergence. By focusing on reducing squared differences between real and synthetic images, our approach enhances training stability and overall performance. We specifically focus on mapping water bodies using MODIS and Landsat-8 imagery due to their relevance in water resource management tasks. Our experimental results demonstrate the effectiveness of Hydro-GAN in generating high-resolution water body maps, outperforming traditional methods in terms of boundary accuracy and overall quality. This study addresses the imperative challenges of water resource management, including coastal zone oversight, detecting sea border shifts due to rising waters, and erosion tracking. Satellite data currently offers a choice between high spatial detail with infrequent updates or lower spatial detail with more frequent updates, presenting a trade-off between data precision and frequency. To efficiently monitor environmental resources like water bodies, we require data with both high spatial detail and frequent updates. To meet this need, we introduce the Hydrological Generative Adversarial Network, a novel machine learning tool that enhances data clarity, particularly in outlining water bodies. In testing, we employed images from the Moderate Resolution Imaging Spectroradiometer satellite, providing less detailed images, and the Land Remote-Sensing Satellite, offering highly detailed imagery. In essence, this study enhances water resource management by effectively combining data from multiple sources, even in adverse conditions, potentially advancing environmental protection and management efforts. Remote sensing data augmentation for improving the accuracy of environmental assessments can be achieved using adversarial networks High spatiotemporal resolution of water bodies data enhances the precision of their areal forecasting Shape and areal accuracies play an important role for efficient spatiotemporal data interpolation