Data-driven parameterization of un(der)-resolved gravity wave momentum transport for atmospheric modeling

I was able to attend the SIAM Conference on the Mathematics of Planet Earth in person this year.

Abstract

Atmospheric gravity waves on scales of 102 to 105 meters play an important role in the global circulation by transporting momentum from the surface to the free atmosphere. The position and strength of the jet streams (and hence the track of winter storms) depends critically on these waves. This poses a problem to atmospheric modeling: large scale features that require a global model depend on fine scale phenomena that we simply can’t simulate on such a domain. This closure problem requires an estimate, or “parameterization” of the missing momentum fluxes based on the resolved scale flow and information about sub-grid scale wave sources (e.g., topography and convection).

I will detail novel data-driven strategies to overcome this challenge. The process involves two components. First, the missing waves must be estimated from high resolution simulations capable of accurately capturing all the relevant scales. Second, machine learning techniques are used to capture the missing momentum transport. We explore differing strategies, contrasting brute force approaches, where we let machine learning crunch the solutions, hoping to learn the statistics from the data alone, vs. hybrid approaches, where we couple machine learning with physics-based closures. The latter strategy involves an inverse problem: what are the optimal gravity wave sources, and machine learning: how to estimate them from available source indicators.