Primary objective

To implement an innovative and effective integrated assessment tool for climate simulations based on dynamical constraints to produce optimal future climate projections from large model ensembles

Secondary objectives

To apply advanced Machine Learning to

1. characterize sources of systematic biases inherited in models by evaluating the observable governing processes rather than end metrics.

2. identify and elucidate emergent constraints for key climate projection metrics and pinpoint sources of multi-model spreads.

3. evaluate the representation of tropical Pacific variability and dynamic of ocean’s heat and carbon budgets, and their associated impacts for state-of-the-art Earth system models.

4. determine and attribute the transformation of internal variability and model uncertainty in response to long-term anthropogenic forcing.

5. support the climate change community by providing guidance for improving process in models and optimizing the monitoring network.