An integrated WRF/HYSPLIT modeling approach for the assessment of PM2.5 source regions over the Mississippi Gulf Coast region

Anjaneyulu Yerramilli*, Venkata Bhaskar Rao Dodla, Venkata Srinivas Challa, La Toya Myles, William R. Pendergrass, Christoph A. Vogel, Hari Prasad Dasari, Francis Tuluri, Julius M. Baham, Robert L. Hughes, Chuck Patrick, John H. Young, Shelton J. Swanier, Mark G. Hardy

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

39 Scopus citations


Fine particulate matter (PM2.5) is majorly formed by precursor gases, such as sulfur dioxide (SO2) and nitrogen oxides (NOx), which are emitted largely from intense industrial operations and transportation activities. PM2.5 has been shown to affect respiratory health in humans. Evaluation of source regions and assessment of emission source contributions in the Gulf Coast region of the USA will be useful for the development of PM2.5 regulatory and mitigation strategies. In the present study, the Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model driven by the Weather Research & Forecasting (WRF) model is used to identify the emission source locations and transportation trends. Meteorological observations as well as PM2.5 sulfate and nitric acid concentrations were collected at two sites during the Mississippi Coastal Atmospheric Dispersion Study, a summer 2009 field experiment along the Mississippi Gulf Coast. Meteorological fields during the campaign were simulated using WRF with three nested domains of 36, 12, and 4 km horizontal resolutions and 43 vertical levels and validated with North American Mesoscale Analysis. The HYSPLIT model was integrated with meteorological fields derived from the WRF model to identify the source locations using backward trajectory analysis. The backward trajectories for a 24-h period were plotted at 1-h intervals starting from two observation locations to identify probable sources. The back trajectories distinctly indicated the sources to be in the direction between south and west, thus to have origin from local Mississippi, neighboring Louisiana state, and Gulf of Mexico. Out of the eight power plants located within the radius of 300 km of the two monitoring sites examined as sources, only Watson, Cajun, and Morrow power plants fall in the path of the derived back trajectories. Forward dispersions patterns computed using HYSPLIT were plotted from each of these source locations using the hourly mean emission concentrations as computed from past annual emission strength data to assess extent of their contribution. An assessment of the relative contributions from the eight sources reveal that only Cajun and Morrow power plants contribute to the observations at the Wiggins Airport to a certain extent while none of the eight power plants contribute to the observations at Harrison Central High School. As these observations represent a moderate event with daily average values of 5-8 μg m-3 for sulfate and 1-3 μg m-3 for HNO3 with differences between the two spatially varied sites, the local sources may also be significant contributors for the observed values of PM2.5.

Original languageEnglish (US)
Pages (from-to)401-412
Number of pages12
JournalAir Quality, Atmosphere and Health
Issue number4
StatePublished - Dec 2012

Bibliographical note

Funding Information:
Acknowledgment This study is carried out as part of the ongoing Atmospheric Dispersion Project (ADP) funded by the National Oceanic and Atmospheric Administration (NOAA) through the US Department of Commerce Grant #NA06OAR4600192. The HYSPLIT model is obtained from NOAA ARL.


  • PM
  • Pollutant dispersion
  • Source identification
  • WRF-HYSPLIT simulation

ASJC Scopus subject areas

  • Pollution
  • Atmospheric Science
  • Management, Monitoring, Policy and Law
  • Health, Toxicology and Mutagenesis


Dive into the research topics of 'An integrated WRF/HYSPLIT modeling approach for the assessment of PM2.5 source regions over the Mississippi Gulf Coast region'. Together they form a unique fingerprint.

Cite this