A Journey to the Edge of our Planet: How Data Assimilation Solve Problems in the Upper Atmosphere

The Earth’s upper atmosphere is a critical juncture between the Earth and space. Both space and terrestrial weathers impact its conditions, which in turn affect a number of modern technologies and infrastructures. Data assimilation and computational statistics play a key role in integrating Earth and geospace observations into Earth and geospace systems models. To make the problem tractable, most data assimilation approaches resort to scalable methods derived from well-established Gaussian process-based formulations of the Bayesian estimation problem. Such approaches are effective in describing the mean behavior of the systems but fall short of representing realistic variability. In this talk, the pros and cons of standard data assimilation approaches, along with a non-Gaussian multi-resolution inferential framework that overcomes the current methodological limitations will be discussed in the context of applications to the Earth’s upper atmosphere.