Source Attribution

01 Make a plan for Stage 1

Source attribution involves determining what source categories contribute to observed air pollution. This can be accomplished using several different methods or approaches. These range from identifying key emitting sources within an emission inventory, to receptor modeling, based on chemically-speciated AQ monitoring data, to modeling air pollution using dispersion models (which only model the movement of pollutants via the wind) and/or chemical transport models (CTMs) that simulate both chemical transformation of emissions as well as meteorology. 

As a first step, familiarize yourself with some of these approaches and begin data collection that can help you implement a formal source apportionment in Stage 2. By gathering knowledge and collecting local data, you can start small and work toward building a program to eventually use all of these approaches. Review the following resources to get a basic understanding of receptor modeling approaches.

02 Understand sources that impact air quality in your jurisdiction

Before you embark on data collection or analysis efforts of your own, look at available data to see what others may have found  regarding the sources that affect your region of the globe, your country or even the local airshed of concern. Several global (or  regional) datasets exist, and academic studies have been carried out for short periods of time or for specific locations that  may closely represent your local conditions. You also might try conducting a Google Scholar search for “source  apportionment fine particles [your jurisdiction]” to see what studies have been conducted locally. 

03 Make a data collection plan

A good starting point for emission inventory approaches can be found in our inventory guidance and look at the “Key Categories” for your jurisdiction. To pursue “top-down” receptor model approaches, a local dataset of speciated AQ monitoring data is needed. Beyond the mass of fine particle pollution, a chemically-speciated dataset is typically derived from 24-hour integrated samples of fine particles collected every third or every sixth day on filters using manual samplers (e.g.  USEPA federal reference method, or FRM, samplers). 

As a starting point, jurisdictions can deploy one or two manual samplers  (co-located with one of your continuous monitors). You collect samples on quartz filters or – if you have capacity and  resources – you can collect two samples simultaneously, one on quartz filters and one on Teflon filters. If you can collect samples on both, it will enable a laboratory to measure a greater range of species (enhancing your ability to discern source  profiles).

Following the steps below, you should plan for filter preparation, sampling and archiving for at least one year, before arranging to have samples analyzed by a specialty analytical lab (private consultant or university lab). Learn more about filter preparation, transport, and collection protocols in the resources below. 

04 Set up a weighing room with an archival freezer

Even before you are ready to prepare and weigh filter samples, you will need to establish a central laboratory where this process can be carried out. This also needs capacity to store/archive the filters after collection and weighing in a freezer for eventual analysis. Learn more about the facilities that will be needed, establish this space and furnish the necessary equipment. 

05 Collect data per established study-design

Ensure you have a well-designed study by reading how others have conducted source apportionment studies using receptor models and ensure that your data collection methods will be fit for purpose. You need a minimum of 100 samples to achieve statistical robustness, so collecting samples at least every third day for one-year is a minimum number of samples for a single site. If you can collect at 2-3 sites and collect meteorological data at each site, you will have a more robust database to identify sources more accurately. 

06 Utilize other techniques to understand likely key sources

Local inventories, dispersion models, continuous monitoring data and trajectory approaches can all be useful for identifying potential sources of air pollution. Knowledge of local sources or likely categories will be useful when trying to identify likely “source profiles” in Stage 2 or 3 receptor model analyses. 

07 Improve data management

Develop a data management approach for 24-hour integrated samples including PM2.5 mass, but also individual chemical species that will be determined in the future when an analytical laboratory analyzes the samples. The data management approach needs to provide data redundancy and backup, but also should align with your continuous data management strategy developed under the air quality monitoring guidance (see Air quality monitoring guidance for Stage 1).  

08 Provide training and capacity-building to staff

While data collection is underway, your staff can participate in training programs to prepare them to get the samples analyzed  at a qualified university or private analytical laboratory and to analyze the results using receptor modeling approaches.  Beyond reading some of the academic literature, users should take training courses to learn how receptor modeling is both a  skill and an art. Learning to recognize patterns of species that indicate specific source profiles or ratios of pollutants that can  indicate the relationship between one source profile or another takes experience. The resources below can help you. 

09 Learn more about chemical transport modeling

While your data collection efforts are underway to prepare for a receptor modeling approach, you can also learn more about how monitoring and inventory activities will help to prepare a team to undertake regional air quality modeling in future stages. Understanding the principles of air quality modeling will help you ensure you have the right inputs when the time comes. 

10 Prepare staff and budget for Stage 2

While it is a challenge to collect filter samples correctly and accurately and to archive them so that they remain valid for analysis more than a year later, it is also quite challenging to prepare for managing chemical speciation data and undertake receptor modeling analysis. As you are able, begin to develop plans for contracting services of an analytical laboratory needed  to develop the chemical speciation data and learn all you can about conducting receptor modeling, including familiarizing  yourself with the source profiles that are common in your area.