Chemical-transport model calculations driven by assimilated data

Output from NASA's uniform grid (GEOS DAS; Bloom et al., 1996) and stretched-grid (GEOS SG-DAS; Rabinovitz et al., 1999) data assimilation systems are being used at the University of Maryland to drive uniform grid [Allen et al., 1996] and stretched- grid [Allen et al., 1999] versions of the University of Maryland/Goddard Chemical-Transport Model (SG-CTM).

Stretched-grid CTM simulations are useful for looking at the effect of small-scale mixing processes on the larger or global scale chemical balance. An example is mixing in a strong convective storm or after injection from a local source. A relatively high resolution is necessary to simulate the dilution that occurs in the mixing region. A reasonable estimate of dilution is needed because the net O3 production rate varies nonlinearly with the NOx concentration [Chatfield and Delany, 1990; Liu et al., 1987]. Another example is stratosphere-troposphere exchange, a process which is driven by the large-scale circulation; however, the mixing associated with it occurs at grid scales too small to be resolved explicitly by global models. Stretched-grid simulations are also useful for interpreting measurements taken over limited areas such as those from field studies or aircraft missions. In addition, stretched-grid simulations are useful when high resolution chemical emission data are available over only a portion of the globe.

The first application of the SG-CTM to airborne field project case studies was for the SONEX (the SASS (Subsonic Assessment) Ozone and Nitrogen Oxides Experiment) campaign, in which considerable deep convection and STE events were encountered. The relative importance of various odd nitrogen (NOy) sources including lightning, aircraft, and surface emissions on upper tropospheric total odd nitrogen was estimated for the SONEX period using the three-dimensional SG-CTM. For this simulation, the stretched-grid was chosen so that its maximum resolution was located over eastern North America and the North Atlantic; a region that included most of the SONEX flight paths. The SONEX period (October-November 1997) was simulated by driving the SG-CTM with assimilated data from the GEOS DAS. A new algorithm that parameterized the flash rate in terms of upper tropospheric convective mass flux was used to estimate the lightning flash rates needed to calculate NOy emission by lightning.

Model-calculated upper tropospheric NOy and NOy measurements from the NASA DC-8 aircraft were compared. Spatial variations in NOy were well captured especially with the stretched-grid run. This experiment was especially useful for identifying the cause(s) of several observed NOy peaks. Peaks due to lightning and stratospheric NOy were captured on numerous occasions; although comparison with measurements indicates that the magnitude of stratospheric peaks was often overestimated. The lightning algorithm reproduces the temporally and spatially averaged total flash rate accurately. However, the location and strength of individual convective (transport and lightning) events is occasionally missed and the use of observed lightning emissions significantly improves the simulation on a few occasions. The following figure shows our model results compared with observations as a function of time throughout one of the SONEX flights. Aircraft (lightning) emissions contributed 15% ( 22%) of the upper tropospheric NOy averaged along SONEX flight paths within the North Atlantic Flight Corridor (NAFC) with the contribution by aircraft (lightning) exceeding 40% (75%) during portions of some flights.

These simulations have improved our understanding of convection, lightning NOx emissions and the NOy budget over the NAFC; however, the dynamical fields used to drive this model were obtained from a coarser uniform grid model. Most future calculations with the SG-CTM will use driving fields from the GEOS SG-DAS. Research topics include:

1) A study of the effects of STE associated with very deep convection on chemical budgets in the upper troposphere and lower stratosphere.

2) A study of the effects of STE associated with synoptic-scale events (e.g., tropopause folds, downward transport behind cold fronts, etc.) on chemical budgets in the UTLS.

3) A study of the above processes using the very fine downscaling of transport and vertical mixing that can be achieved through use of a SG-CTM driven by data from the GEOS SG-GCM or GEOS SG- DAS.

4) A study of the physical and chemical effects of regional vertical redistribution processes (e.g., investigate the effects of perturbed profiles of trace gases and aerosols on HOx production, photochemical ozone production, and the resulting radiative forcing of climate). Examine what effects these changes of radiative forcing have on regional dynamics.

REFERENCES: Allen, D. J., R. B. Rood, A. M. Thompson, and R. D. Hudson, Three-dimensional radon-222 calculations using assimilated meteorological data and a convective mixing algorithm, J. Geophys. Res., 101, 6871-6881, 1996.

Allen, D. J., K. E. Pickering, A. M. Thompson, G. Stenchikov, and Y. Kondo, A three-dimensional total odd nitrogen (NOy) simulation during SONEX using a stretched-grid chemical transport model, J. Geophys. Res., submitted, 1999.

Bloom, S. C., L. L. Takacs, A. M. DaSilva, and D. Ledvine, Data assimilation using incremental analysis updates, Mon. Weather Rev., 124, 1256-1271, 1996.

Chatfield, R. B., and A. C. Delany, Convection links biomass burning to increased tropical ozone: However, models will tend to overpredict O3, J. Geophys. Res., 95, 18,473-18,488, 1990.

Fox-Rabinovitz, M. S., L. L. Takacs, D. P. Dee, L. Rukhovets, J. Guo, and R. Govindaraju, A variable resolution stretched grid data assimilation system: Regional applications, Mon. Weather Rev., To be submitted, 1999c.

Liu, S. C., M. Trainer, F. C. Fehsenfeld, D. D. Parrish, E. J. Williams, D. W. Fahey, G. Hübler, and P. C. Murphy, Ozone production in the rural troposphere and the implications for regional and global ozone distributions, J. Geophys. Res., 92, 4191-4207, 1987.