GeoViT: versatile vision transformer architecture for geospatial image analysis

Greenhouse gases (GHGs) are pivotal drivers of climate change, necessitating precise quantification and source identification to foster mitigation strategies. We introduce GeoViT, a compact vision transformer model adept in processing satellite imagery for multimodal segmentation, classification, and regression tasks targeting CO <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> and NO <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> emissions. Using GeoViT, we achieve superior accuracy in estimating power generation rates, fuel type, plume coverage for CO <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> , and high-resolution NO <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> concentration mapping as compared to previous state-of-the-art models while significantly reducing model size. GeoViT demonstrates the efficacy of vision transformer architectures in harnessing satellite imagery data for enhanced GHG emission insights, which would be instrumental in advancing climate change monitoring and emission regulation efforts globally.