(not sign)AI model harnesses physics to autocorrect remote sensing data.

A new artificial intelligence (AI) model developed by James Koch, a data scientist at Pacific Northwest National Laboratory (PNNL), uses physics-informed machine learning to address the problem of atmospheric corruption in remote sensing data. The model can remove the influence of the atmosphere from spectral data collected by remote sensors, allowing for more accurate analysis. Koch's framework is bidirectional, meaning it can also infer how a material on the ground would appear when imaged through a particular atmosphere. The AI technique has the potential to improve remote sensing capabilities in various applications, including target detection and coastal ecosystem health monitoring. [Extracted from the article]