AGU 2023

This year, the AGU 2023 Fall Meeting is taking place in San Francisco, just down the road from me. After the success of last year, Minah Yang and I are co-convening our AGU session on Machine Learning Subgrid-Scale Parameterizations for Earth System Modeling.

We have talks on Wednesday morning followed by a poster session on Thursday morning.

I will be presenting a poster Uncertainty Quantification of Machine Learning Subgrid-Scale Parameterizations for Gravity Waves

Sub-grid scale (SGS) parameterizations estimate effects of unresolved processes without modeling them directly and are often a large source of uncertainty in Earth system models (ESM). Advances in big data, machine learning/artificial intelligence (ML/AI) and hardware present new methods coupled with enhanced computational capabilities, which we can utilize for scientific applications. Specifically, we are interested in scientific ML/AI for replacing or adding to existing parameterizations.

We solicit contributions on various aspects of marrying data-driven methods to parameterizations from all weather and climate modeling communities. Topics may include:

Type of data: observations types, high-resolution model outputs, remote sensing products
Types of models and data processing: computer vision techniques, neural networks, filtering methods, data assimilation techniques for inverse problems,
Developing data-driven models that include physical knowledge to ensure stability/conservation laws.
Metrics to evaluate parameterization accuracy and model biases.
Online-coupling of data-driven parameterization in GCMs: computational challenges, generalization to new climate.