Using continual pretraining with a geospatial foundation model
- Geoffrey Dawson
- Chris Dearden
- et al.
- 2024
- ML4EO 2024
Geoffrey is a research scientist focusing on climate and sustainability. His interests are in applying machine learning to various climate focused problems, with a particular focus on remote sensing and physical modelling.
Prior to joining IBM Geoffrey was a researcher at the Bristol Glaciology Center (University of Bristol), where his research focused on developing new techniques to measure changes to the Antarctic ice sheet and Arctic sea ice. For example he developed a new technique to measure summer sea ice thickness, based on deep learning. This is the first time summer sea ice thickness has been able to be measured at the pan-arctic scale, and this new data will hopefully our ability to forecast sea ice extent and improve our understanding of how the Arctic is changing due to climate change.