228
Analysis of tropical convective transport of trace gases and lightning NOx production during the TC4 mission using the GMI model

- Indicates paper has been withdrawn from meeting
- Indicates an Award Winner
Monday, 24 January 2011
Analysis of tropical convective transport of trace gases and lightning NOx production during the TC4 mission using the GMI model
Washington State Convention Center
Theodore V. Lyons III, University of Maryland, College Park, MD; and K. E. Pickering and D. Allen

Poster PDF (583.5 kB)

Tropical convection is an important distributor of trace gases, particularly from the lower troposphere into the tropical tropopause layer. These convective storms also produce NOx from lightning. Thus, both the dynamical and chemical modules of a model are vital to accurately portray the vertical profiles of trace gases. From a dynamical standpoint, we evaluate differences in areal coverage of convective activity and estimated cloud top height between versions of the NASA Global Modeling Initiative (GMI) chemical transport model driven by the GEOS-4 and GEOS-5 assimilated data and satellite and aircraft measurements of convection. We also appraise differences between the models within the convective column. It is shown that models produce convection with areal coverage far in excess of reality and that the convective mass flux and detrainment profiles differ considerably between the two versions of the model.

Our attention then shifts to the distribution and production of trace gases within the convective column. Using trace gas observations from the TC4 aircraft (DC-8 and WB-57), we analyze the efficacy of the GMI model in reproducing observed profiles. We consider ozone and carbon monoxide, which are nearly conserved on the time scale of convective cells. We find that in undisturbed regions, the models have good agreement with in situ measurements, but disturbed regions show less agreement which is consistent with likely errors in the model mass flux and detrainment profiles. NOx is also considered which is created in situ by lightning and redistributed throughout the column. We find significant differences in NOx partitioning between in situ observations and the model. We also compare the total column content and vertical profiles produced by the model to satellite datasets.