Creating synthetic night-time visible-light meteorological satellite images using the GAN method

TL;DR

This paper introduces a deep learning approach using GANs with channel-wise attention to generate realistic night-time visible-light satellite images from infrared and weather prediction data, filling a critical data gap.

Contribution

The paper presents a novel GAN-based method with attention mechanisms to synthesize night-time visible satellite images from infrared and NWP data, improving realism and data availability.

Findings

Generated images are visually realistic during night.

Method effectively models nonlinear relationships between data types.

Results demonstrate improved image quality over baseline methods.

Abstract

Meteorology satellite visible light images is critical for meteorology support and forecast. However, there is no such kind of data during night time. To overcome this, we propose a method based on deep learning to create synthetic satellite visible light images during night. Specifically, to produce more realistic products, we train a Generative Adversarial Networks (GAN) model to generate visible light images given the corresponding satellite infrared images and numerical weather prediction(NWP) products. To better model the nonlinear relationship from infrared data and NWP products to visible light images, we propose to use the channel-wise attention mechanics, e.g., SEBlock to quantitative weight the input channels. The experiments based on the ECMWF NWP products and FY-4A meteorology satellite visible light and infrared channels date show that the proposed methods can be effective to create realistic synthetic satellite visible light images during night.