Improved Air Quality Modeling in the NASA GEOS-5 Model Using a Multispecies Data Assimilation System of Tropospheric Constituents

The chemical mechanism from the GEOS-Chem global chemical transport model (CTM) has been implemented as a chemistry module into the NASA Goddard Earth Observing System (GEOS) data assimilation system (DAS). This coupled system provides a generic research platform for a wide range of air-quality problems, including surface ozone, ozone precursors, and carbon monoxide. It thus significantly enhances the capabilities of the GEOS DAS with respect to tropospheric chemistry and air quality applications. We show results from a one-year experiment that combines the online GEOS-Chem model forecasts with ozone observations from MLS and OMI, nitrogen dioxide from OMI, and carbon monoxide from MOPITT.

We find that assimilation of OMI and MLS ozone decreases the error in GEOS-Chem ozone forecasts by up to 50% in the upper troposphere and 25% near the surface compared to an assimilation-free control simulation. This result can be attributed to a combination of improved representation of stratosphere-troposphere exchange, lightning NOx emissions, and model background concentrations. The concurrent assimilation of NO₂ and CO helps identifying the relative importance of these factors. For instance, model-observation mismatches of NO₂ point to local deficiencies in model emissions of NO₂as being the cause of some of the observed ozone discrepancies in the free-running model. 

This new chemical data assimilation system leverages the expertise of NASA’s Global Modeling and Assimilation Office (GMAO) and the GEOS-Chem modeling community, making it uniquely positioned to benefit from advances in atmospheric chemistry modeling and data assimilation techniques. We discuss ongoing developments as well as challenges with respect to air quality modeling. This includes the increase in horizontal resolution to better capture fine-scale patterns that are critical for human health studies, the transition from 3D-Variational to ensemble-based 4D-Variational systems in the GMAO, and the expansion of the assimilation state vector to include emissions.