Super-resolution of sea surface temperature with convolutional neural network- and generative adversarial network-based methods. Issue 4 (23rd March 2022)

• Record Type: Journal Article
• Authors: Izumi, Tomoki; Amagasaki, Motoki; Ishida, Kei; Kiyama, Masato
• Is Part Of: Journal of water and climate change. Volume 13:Issue 4(2022)
• Page Start: 1673
• Page End: 1683
• Publication Date: 2022-03-23
• Subjects: convolutional neural network -- ESRGAN -- generative adversarial network -- RRDBNet -- single-image super-resolution; Water -- Periodicals; Hydrology -- Periodicals; Climatic changes -- Periodicals; Climatic changes; Hydrology; Water; Electronic journals; Periodicals; 333.9116
• DOI: 10.2166/wcc.2022.291
• Languages: English
• ISSNs: 2040-2244

Abstract

Abstract: In this paper, we perform the super-resolution of sea surface temperature data with the enhanced super-resolution generative adversarial network (ESRGAN), which is a deep neural network-based single-image super-resolution (SISR) method that uses a generative adversarial network (GAN). We generate high-quality super-resolution data with ESRGAN and with the super-resolution convolutional neural network (SRCNN) and residual-in-residual dense block network (RRDBNet) methods, which are based on convolutional neural networks (CNNs). The images generated with these methods are compared with high-resolution optimum interpolation sea surface temperature (OISST) data using root mean square error (RMSE), learned perceptual image patch similarity (LPIPS), and perceptual index (PI) evaluation methods. RRDBNet has a better RMSE than SRCNN and ESRGAN. However, CNN-based SISR methods do not provide a faithful representation of the ocean currents of OISST. ESRGAN has a better LPIPS and PI than CNN-based methods and can represent the complex distribution of ocean currents. HIGHLIGHTS: RRDBNet has a better RMSE than SRCNN and ESRGAN on super-resolution of sea surface temperature data. ESRGAN has a better LPIPS and PI than CNN-based methods and can represent the complex distribution of ocean currents. CNNs cannot interpolate the missing information, but GANs have better results for these parts.