Current work includes the development of ProxyVis global multi-satellite composite imagery; experimenting with different AI algorithms to improve the ProxyVis imagery, including generating higher quality image over the land and at high latitudes; and testing ProxyVis as a replacement for visible imagery in other multispectral algorithms. In addition, the AI version aims to better represent cold cloud tops. While the low-level oceanic clouds are overall well-represented in ProxyVis, the cold cloud tops have little brightness contrast, which could present problems for machine learning algorithms and interpretability of multi-spectral algorithms that might be able to use ProxyVis as a night-time replacement for visible imagery for quantitative applications.
This presentation will provide an overview of the four existing ProxyVis algorithms, discuss their differences and similarities, and provide guidance for their suggested uses. Furthermore, the presentation will provide examples of ProxyVis use at different forecast offices, and contrast those examples with the widely used Nighttime Microphysics RGB product. Specifically, examples will detail the tracking of low-level clouds in a variety of scenarios. Finally, we will discuss progress and development of new AI-based versions of ProxyVis.
Disclaimer: The scientific results and conclusions, as well as any views or opinions expressed herein, are those of the author(s) and do not necessarily reflect those of NOAA or the Department of Commerce.

