01 Revisit source apportionment analysis procedures for your second analysis

In the last stage, you should have identified additional high-quality source profile data that were needed to reflect local emissions more accurately. This may require field studies, emissions inventories, and information on industrial operations, traffic patterns, and seasonal variations.  By now, you also will have extended your database by incorporating a longer time series of air quality data. Using these refinements and other improvements will allow for a more comprehensive analysis that captures temporal trends and variations in pollutant concentrations, improving the robustness of your analysis approach. Ensure the dataset includes diverse meteorological conditions to enhance the reliability of the findings and you will be ready to make a plan to re-run the receptor model with refined source profiles and an expanded dataset. Prepare staff to compare the new results against previous findings, focusing on discrepancies that may highlight previously overlooked sources. Make sure your plan includes a validation step to compare improved model outcomes with independent data or additional monitoring efforts, ensuring that the enhanced model yields more accurate and actionable insights for air quality management. 

02 Extend chemical speciation database

A first step is to update your air quality monitoring database with new data for PM mass and specified elements, including inorganic elements, major ions, and carbonaceous aerosols, to extend the database's time series and improve its utility.  

Begin by systematically collecting new samples following standardized protocols to ensure compatibility with existing data. Ensure that the new samples are analyzed using consistent laboratory methods, calibrated instruments, and high-quality quality assurance/quality control (QA/QC) processes to maintain data integrity.

Once collected, integrate the new data into the existing database, clearly documenting sample dates, locations, methodologies, and analytical techniques used for transparency. Retain the same format and units as previous data entries for consistency.

To refine uncertainty estimates, reassess measurement methods and laboratory protocols that may have changed since prior datasets were created. Calculate updated uncertainties for each measurement based on the latest calibration standards, replicate analyses, and instrument precision. Use statistical methods, such as error propagation, to assess overall uncertainties when combining multiple datasets that may have different levels of uncertainty over different time periods. The paper by Liao et al (2015) combines multiple data sets (VOCs and PM2.5 and multiple time series (12-hour and 24-hour data) to enhance the interpretation of potential sources.  

Include this uncertainty information in the database alongside the new measurements. Finally, regularly review the entire data collection and processing workflow to identify opportunities for enhancing data quality and reducing uncertainty in future measurements. 

03 Add new species

Incorporating additional species and parameters into a mathematical receptor modeling exercise significantly enhances the robustness and accuracy of the analysis. By including more chemical species—such as ultrafine particles, organic carbon fractions, and specific trace elements or gases—you can capture a more comprehensive representation of the pollution mix and identify previously overlooked sources.  

Additionally, incorporating relevant meteorological parameters, like humidity and temperature, can improve model performance by accounting for atmospheric processes that influence pollutant dispersion and transformation. Ultimately, a more detailed dataset leads to better source apportionment results, informing targeted air quality management strategies and improving regulatory compliance. 

04 Update your source profile database

Conduct targeted field measurements using sampling techniques that capture emissions during typical operating conditions. Employ techniques such as Fourier Transform Infrared Spectroscopy (FTIR) for gas emissions and gravimetric methods for particulate matter to acquire robust chemical profiles.

Additionally, consider utilizing mobile air quality monitoring platforms to collect real-time data across various locations and times, identifying emission variations due to geographic and temporal factors.

Collaborate with academic institutions or environmental organizations with expertise in source profiling to enhance methodological rigor and credibility. Finally, ensure proper documentation of methodologies and conditions during sampling to facilitate transparency and reproducibility of the profiles for future receptor modeling studies. 

05 Repeat receptor modeling/geographic analysis

An initial receptor modeling analysis typically serves as a preliminary investigation, aiming to identify key sources of air pollution based on available data from a limited time period and selected parameters. This analysis often relies on existing source profiles, which may not fully represent local emissions' unique characteristics.

In contrast, a refined analysis builds upon the initial findings by extending the time-series data, allowing for a more comprehensive understanding of pollutant variability and trends. A longer time series captures seasonal and temporal changes in emissions, leading to more representative results.

Incorporating additional parameters—such as specific trace elements, ultrafine particles, and relevant meteorological conditions—improves the model's accuracy and reliability. These parameters provide insights into atmospheric processes and help distinguish between different sources more effectively.

Using new, local source profiles is another critical enhancement. By developing profiles from recent, localized data through field measurements or stakeholder engagement, the refined analysis can better represent the emissions from local activities, which is crucial for effective source apportionment.

Ultimately, the refined analysis yields more precise and actionable insights, supporting stronger regulatory decisions and targeted air quality management strategies tailored to local conditions and sources. 

06 Identify a chemical transport model and prepare for modeling

Identifying a suitable chemical transport model (CTM) for regional-scale air quality modeling can complement you efforts to identify key sources of air pollution, but involves several key considerations, including the specific pollutants of interest, geographic scope, and computational resources. Popular models like the Community Multiscale Air Quality (CMAQ) model, the CAMx model, or the Weather Research and Forecasting model coupled with Chemistry (WRF-Chem) are strong candidates, as they are widely used and supported by extensive documentation and user communities.

Given the human resource burden of undertaking an AQ modeling program, it may make sense to establish a partnership with academic institutions or local scientific organizations that have aligned interests.  Begin by identifying potential collaborators who have expertise in air quality modeling or relevant fields. Approach universities or research centers that have conducted similar studies, leveraging their knowledge and resources. Articulate the goals of the partnership, emphasizing the mutual benefits—such as data sharing, methodological enhancements, and access to high-performance computing facilities.

During collaboration, clearly define roles and responsibilities, including data collection, model setup, and interpretation of results. Jointly work on integrating emission inventories, which are crucial for accurately simulating source impacts. Establish regular communication channels and provide opportunities for knowledge exchange, such as workshops or training sessions. By fostering cooperation, the partnership can enhance the quality and relevance of the regional air quality modeling, leading to better-informed management decisions and policies. 

07 Prepare emissions inputs

Format emissions inventory data for chemical transport modeling by compiling emissions data into a spreadsheet or database with columns for source categories, pollutants, geographic coordinates, and temporal resolution. Ensure that the data includes annual, monthly, daily, and hourly emissions, if available. Identify spatial surrogates to allocate emissions across different geographic areas. Common surrogates include population density, land use, traffic counts, and industrial activity. Match these surrogates with your inventory data to distribute emissions accurately across grid cells. 
Conduct emissions modeling to refine estimates, possibly using tools like the Environmental Protection Agency’s (EPA) MOVES model or the COPERT model for mobile sources or the Speciation Tool for particulate matter. Convert the formatted emissions inventory into gridded input files suitable for your chemical transport model, ensuring proper adherence to the model’s required format. 
Finally, validate the gridded emissions against observed air quality data to check for accuracy and ensure effective representation in the modeling framework. 

08  Prepare meteorological inputs

Prepare meteorological inputs for chemical transport modeling by selecting a suitable meteorological model, such as WRF (Weather Research and Forecasting), that can provide the necessary data resolution and coverage for your study area. 
Gather essential meteorological parameters, including wind speed and direction, temperature, humidity, precipitation, and surface pressure. These parameters should be available at multiple heights and over the same time frame as your emissions data. 
Generate or download high-resolution meteorological data. In general, finer spatial resolutions are preferable, ensuring that temporal resolution suits your modeling needs—typically hourly or sub-hourly data is ideal. If using WRF, configure it to simulate the desired meteorological conditions for the study period, and validate outputs against existing observational data to ensure accuracy. 
Finally, format the meteorological inputs according to the specifications of your chemical transport model, maintaining consistency in units and file structure. Accurate meteorological inputs are essential for reliable model performance and meaningful air quality assessments.  

09 Perform first Chemical Transport Model (CTM) run

Set up the modeling framework, ensuring proper configuration of the chosen chemical transport model (e.g., CMAQ, CAMx, WRF-Chem). Define the geographic domain, grid resolution, and specific modeling periods based on the objectives of the analysis. After configuration, initialize the model and run simulations. Monitor the model's convergence and ensure that it successfully completes without errors. Evaluate output data for key pollutants, focusing on average concentrations, spatial distributions, and temporal patterns. 
Post-simulation, analyze the results to identify high concentration areas and their correlations with known emission sources. Use receptor modeling techniques, such as source apportionment, to assess contributions from various categories (e.g., traffic, industrial emissions). Visualize the data through GIS tools to facilitate understanding of geographic patterns and relationships. 
Finally, compare the modeled results with observational data to validate accuracy, adjusting model parameters as necessary. This iterative process will help identify priority sources for further investigation and inform targeted air quality management strategies tailored to the jurisdiction's specific needs. 

10 Identify areas for improvement and make a plan for field campaigns

Using the results of the refined source apportionment analysis and your first regional CTM run, continue to focus on potential measurements, inventory activities or sampling that can be improved to further refine and enhance source identification. You wish to begin planning for a scientific research campaign to target key local sources. Work with key regional leaders to share the initial results and build community support for continued future analysis. As you master chemical transport modeling and prepare for Stage 4, consider whether you have the resources to undertake a CTM-based source apportionment analysis.