Improving chemical transport models for atmospheric simulations relies on future developments of mathematical methods and parallelization methods. Better mathematical methods allow simulations to more accurately model realistic processes and/or to run in a shorter amount of time. Parellization methods allow simulations to run in much shorter amounts of time, therefore allowing scientists to use more accurate or more detailed simulations (higher resolution grids, smaller time steps). The state-of-the-science GEOS-Chem model is modified to use the Kinetic Pre-Processor, giving users access to an array of highly efficient numerical integration methods and to a wide variety of user options. Perl parsers are developed to interface GEOS-Chem with KPP in addition to modifications to KPP allowing KPP integrators to interface with GEOS-Chem. A variety of different numerical integrators are tested on GEOS-Chem, demonstrating that KPP provided chemical integrators produce more accurate solutions in a given amount of time than the original GEOS-Chem chemical integrator. The STEM chemical transport model provides a large scale end-to-end application to experiment with running chemical integration methods and transport methods on GPUs. GPUs provide high computational power at a fairly cheap cost. The CUDA programming environment simplifies the GPU development process by providing access to powerful functions to execute parallel code. This work demonstrates the accleration of a large scale end-to-end application on GPUs showing significant speedups. This is achieved by implementing all relevant kernels on the GPU using CUDA. Nevertheless, further improvements to GPUs are needed to allow these applications to fully exploit the power of GPUs. / Master of Science

Air quality in the Mexico City Metropolitan Area (MCMA) is the subject of many studies due to concerns from high emissions and their adverse effects on public health and the environment. In this study, a high resolution simulation is performed with the Community Multi-scale Air Quality modeling system (CMAQ) using meteorology generated by the Weather Research Forecasting system (WRF). The boundary conditions for CMAQ are provided by the Goddard Earth Observing System-CHEMistry model (GEOS-Chem). The simulation period was March 2-7, 2006. Hourly species concentrations of O3, NOx, CO, SO2, PM10, and PM2.5 for the period were provided by the Automatic Air Quality Monitoring Network (labeled as RAMA). Preliminary evaluation showed GEOS-Chem and CMAQ being in good agreement with their predicted concentrations. In comparison with the base case boundary conditions, the GEOS-Chem case performs better and predicts closer to the observed values of O3, NOx, PM10, PM2.5, and SO2. Particle trajectory analysis was performed using the HYbrid Single-Particle Lagrangian Integrated Trajectory model (HYSPLIT) to ascertain the major sources of SO2 emitters and their impact on the MCMA.

Isoprene emissions affect human health, air quality, and the oxidative capacity of the atmosphere. Globally anthropogenic non-methane volatile organic compounds (NMVOC) emissions are lower than that of isoprene, but local hotspots are hazardous to human health and air quality. In Africa the tropics are a large source of isoprene, while Nigeria appears as a large contributor to regional anthropogenic NMVOC emissions. I make extensive use of space-based formaldehyde (HCHO) observations from the Ozone Monitoring Instrument (OMI) and the chemical transport model (CTM) GEOS-Chem to estimate and examine seasonality of isoprene emissions across Africa, and identify sources and air quality consequences of anthropogenic NMVOC emissions in Nigeria. / Earth and Planetary Sciences

Abstract : This is a study concerning comparisons between the Dubovik Aerosol optical depth (AOD) retrievals from AEROCAN (ARONET) stations and AOD estimates from simulations provided by a chemical transport model (GEOS-Chem : Goddard Earth Observing System Chemistry). The AOD products associated with the Dubovik product are divided into total, fine and coarse mode components. The retrieval period is from January 2009 to January 2013 for 5 Arctic stations (Barrow, Alaska; Resolute Bay, Nunavut; 0PAL and PEARL (Eureka), Nunavut; and Thule, Greenland). We also employed AOD retrievals from 10 other mid-latitude Canadian stations for comparisons with the Arctic stations. The results of our investigation were submitted to Atmosphere-Ocean. To briefly summarize those results, the model generally but not always tended to underestimate the (monthly) averaged AOD and its components. We found that the subdivision into fine and coarse mode components could provide unique signatures of particular events (Asian dust) and that the means of characterizing the statistics (log-normal frequency distributions versus normal distributions) was an attribute that was common to both the retrievals and the model. / Résumé : Cette étude compare des épaisseurs optiques d’aérosols (AOD) à 5 stations arctiques d’AEROCAN (AERONET), obtenues d’une part à l’aide de l’algorithme d'inversion de Dubovik appliqué à des mesures in situ, et d’autre part du modèle de transport chimique (GEOS-Chem : Goddard Earth Observing Système Chemistry). Les produits d’AOD associés à l’algorithme d’inversion sont divisés en composantes totales, fines et grossières. Pour chacune des 5 stations (Barrow, Alaska, Resolute Bay, au Nunavut, 0PAL et PEARL (Eureka), Nunavut, et Thulé, au Groenland), la période de récupération est de janvier 2009 à janvier 2013. Nous avons également utilisé les mesureurs d’AOD de dix autres stations canadiennes de latitudes moyennes, à des fins de comparaison. Les résultats de l’étude ont été soumis à la revue Atmosphere-Ocean. Pour résumer brièvement ces résultats, le modèle a généralement, mais pas toujours, eu tendance à sous-estimer l'AOD moyenne et de ses composantes. Nous avons constaté que la subdivision en composantes fine et grossière pourrait fournir des signatures uniques d'événements particuliers (poussière asiatique) et que les moyens de caractériser des statistiques (les distributions de fréquence log-normale versus les distributions normales) était un attribut qui était commun aux deux les mesureurs et le modèle.

The Models-3/Community Multi-scale Air Quality modeling system (CMAQ), coupled with Goddard Institute for Space Studies (GISS) atmospheric General Circulation Model (GCM), fifth Generation Mesoscale Model system (MM5), and Goddard Earth Observing System-CHEMistry (GEOS-Chem), was used to simulate atmospheric concentration of ozone and particulate matter over the continental United States 12-km and 36-km (CONUS) domains at year 2000 and year 2050. In the study, GISS GCM model outputs interfaced with MM5 were utilized to supply the current and future meteorological conditions for CMAQ. The conventional CMAQ profile initial and boundary conditions were replaced by time-varied and layer-varied GEOS-Chem outputs. The future emission concentrations were estimated using year 2000 based emissions with emission projections suggested by the IPCC A1B scenario. Multi-scenario statistical analyses were performed to investigate the effects of climate change and change of anthropogenic emissions toward 2050. The composite effects of these changes were broken down into individual effects and analyzed on three distinct regions (i.e., Midwest, Northeast and Southeast). The results of CMAQ hourly and 8-hour average concentrations indicate the maximum ozone concentration in the Midwest is increased slightly from year 2000 to year 2050, as a result of increasing average and maximum temperatures by 2 to 3 degrees Kelvin. In converse, there is an observed reduction of surface ozone concentration in the Southeast caused by the decrease in solar radiation. For the emission reduction scenario, the decline of anthropogenic emissions causes reductions of both ozone and PM2.5 for all regions. The emission reduction has compensated the effect of increasing temperature. The overall change on the maximum daily 8-hr ozone and average PM2.5 concentrations in year 2050 were estimated to be 10% and 40% less than the values in year 2000, respectively. The modeling results indicates the effect of emissions reduction has greater impact than the effect of climate change.