Bryan Duncan (PI)
University of Maryland Baltimore County
duncan@code916.gsfc.nasa.gov

A Global Modeling Initiative (GMI) Study of the Sensitivity of Transport to Meteorological Fields

A credible model simulation of the transport and transformation of trace gases and aerosol in the atmosphere requires a realistic representation of small (sub-grid) to large (resolved) scale meteorological phenomena.  We propose to evaluate the credibility of circulation produced by several meteorological fields that drive transport in the Global Modeling Initiative (GMI) chemical tracer model (CTM).  We will assess the model’s simulation of large-scale circulation, such as stratosphere-troposphere exchange (STE) and tropospheric inter-hemispheric exchange (IHE), as well as smaller scale motions, including convection and pollution plume transport.  The GMI CTM is uniquely suited for this study as it can use meteorological input from various general circulation models (GCM) and data assimilation systems (DAS).  The proposed work is divided into two sections.  In the first section, we will evaluate the performance of the GMI CTM with transport driven by meteorological fields from several sources:  1) GEOS-4-DAS fields [Goddard Earth Observing System (GEOS) DAS, version 4] from the Goddard Modeling and Assimilation Office (GMAO), 2) GEOS-4 DAS forecast fields (GEOS-4-Forecast), 3) Oslo/European Centre DAS forecast fields (Oslo/EC-Forecast) from Oslo University and U. California, Irvine, and 4) GEOS-4 Atmospheric GCM fields (GEOS-4-AGCM).  It is known that DAS fields can cause unrealistic transport, such as excessive STE, in a CTM due to the non-physical forcing introduced by data insertion.  A CTM with transport driven by GCM-derived fields, on the other hand, will not, in general, produce constituent fields that can be directly compared with observations.  The CTM simulations proposed here will evaluate differences between simulations using DAS and forecast meteorological fields.  The use of forecast fields should partially damp the physical effects of data insertion, producing more realistic transport while maintaining information from the assimilation procedure.  We will perform a series of CTM simulations for January through April 2001, which include the time of the TRACE-P field campaign, evaluating the runs with measurements from the campaign and observations taken from space.  Second, we propose to assess the credibility of GMI CTM simulations using Oslo/EC-Forecast and GEOS-4-Forecast fields as applied to the impact of boreal forest fires in 2002 on United States air quality.  (If the GEOS-4-Forecast fields are not available, we will use GEOS-4-DAS fields instead.)  To evaluate the fidelity of transport in the two simulations, we will compare CTM output with in situ observations, such as ground-level ozone (O3) from the AIRNow database, and remotely sensed data (e.g., carbon monoxide (CO) from the MOPITT instrument).  We will concentrate on the summer of 2002, a time with a moderate number of Canadian forest fires, but a significant impact on United States air quality.  The simulations will use aerosol and trace gas emissions developed specifically for 2002.  We will constrain reactive nitrogen (NOx) emissions from the fires using observations of tropospheric NO2 from the SCIAMACHY instrument.  In addition, we will constrain the boreal fire injection heights of pollution through comparison of model output from sensitivity simulations with observations of CO (MOPITT, AIRS) and aerosol (MODIS).  We will investigate the fire’s effect on i) summertime surface O3 and CO, including the number of violations of the National Ambient Air Quality Standard (NAAQS) for O3, and ii) the radiative budget by black and organic carbon aerosol.

+ Back to Participant Listing