Decreasing the computational cost of high-resolution, analytical inversions

Analytical inversions of satellite observations of atmospheric composition can improve emissions estimates and quantify error but are computationally expensive at high resolution. We propose two methods to decrease this cost. In an inversion of GOSAT satellite methane observations, the methods reproduce high-resolution results at a quarter of the cost. The reduced-dimension method creates a multiscale grid. The reduced-rank method solves the inversion where information content is highest.

Citation

Nesser, H., D.J. Jacob, J.D. Maasakkers, T.R. Scarpelli, M.P. Sulprizio, Y. Zhang, and C.H. Rycroft, Reduced-cost construction of Jacobian matrices for high- resolution inversions of satellite observations of atmospheric composition, Atm. Meas. Tech., 14, 5521–5534, https://doi.org/10.5194/amt-14-5521-2021, 2021. [Link]

In progress: Understanding the sensitivity of regional inverse analyses to boundary conditions

Regional inverse studies use chemical transport models (CTMs) and observations of methane concentrations to improve methane emission estimates at the high resolution needed to identify emission sources. However, the optimized emission estimates are sensitive to the methane concentrations used by the CTM at the boundary of the domain. We are developing a theoretical and numerical framework to quantify, predict, and correct the influence of these biases on the improved emissions estimates.