2 results
Develop atmospheric inversions of methane and nitrous oxide to constrain greenhouse-gas sources and sinks at high latitudes. The project combines CarboScope transport and inversion modelling with Arctic and Canadian observations, isotopic measurements and satellite data to investigate carbon-climate feedbacks and permafrost regions. The appointment lasts…
Develop NeuralGCM+, an open-source climate model combining machine learning with physical principles to improve projections of unseen future climates. The project builds on NeuralGCM, couples atmospheric and ML-based ocean models, and investigates which processes should remain governed by physical equations and which can be learned from historical…
End of results.