Creating synthetic night-time visible-light meteorological satellite images using the GAN method
This addresses a specific gap in meteorology support and forecasting, but is incremental as it applies existing GAN techniques with minor modifications to a new domain.
The paper tackles the lack of night-time visible-light meteorological satellite data by proposing a GAN-based method to generate synthetic images from infrared data and NWP products, showing effectiveness in creating realistic images.
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.