Interpreting Satellite Remote Sensing Observations using a Chemical Transport Model: Implications for Processes affecting Tropospheric NOx and Ozone
Abstract
Nitrogen oxides (NOx = NO + NO2) and ozone (O3) play important roles in the troposphere that impact air quality and climate. This thesis presents three projects that demonstrate how satellite observations interpreted with a chemical transport model can provide insight into emission, chemistry, and transport processes that affect NOx and O3 concentrations in the troposphere.
Satellite observations from three instruments (MLS, OSIRIS, ACE-FTS) are used to reveal coherent patterns of low ozone events (< 20 ppbv) in the tropical upper troposphere. Modeling studies using the GEOS-Chem chemical transport model indicate that these events result from deep convective processes that rapidly transport ozone-depleted air from the marine boundary layer. The satellite observations indicate spatial shifts in the frequency of low ozone events that arise from changes in convection associated with the El Niño and Madden-Julian Oscillations.
A comparison between nitric acid (HNO3) columns from the IASI satellite instrument and those simulated with GEOS-Chem reveal a model underestimation over Southeast Asia. Sensitivity studies indicate that this bias is likely driven by nonlinear chemistry effects during the lightning NOx parameterization. We tested a subgrid lightning plume parameterization and found that an ozone production efficiency of 15 mol/mol in lightning plumes over Southeast Asia with an additional 0.5 Tg N would reduce the regional nitric acid bias from 92% to 6%.
The GEOS-Chem adjoint model is used as a benchmark to evaluate mass balance methods for inverse modeling of NOx emissions from synthetic NO2 columns. We find that error in mass balance inversions can be reduced by a factor of two by using an iterative process that utilizes finite difference to linearize the model around its a priori state. The iterative finite difference mass balance and adjoint-based 4D-Var methods produce similar top-down inventories when inverting hourly synthetic observations, reducing the a priori error by a factor of 3-4. Inversions of synthetic satellite observations from low Earth and geostationary orbits also indicate that the finite difference mass balance and adjoint-based 4D-Var inversions produce similar results, reducing a priori error by a factor of 3.