Source apportionment of ambient PM10 and PM2.5 in Haikou, China

- The characteristics of ambient PM in Haikou were investigated.
- CMB model was used for source apportionment.
- Particulate matter was more affected by regional sources in winter than in spring through back trajectory analysis.

In order to identify the sources of PM10 and PM2.5 in Haikou, 60 ambient air samples were collected in winter and spring, respectively. Fifteen elements (Na, Mg, Al, Si, K, Ca, Ti, V, Cr, Mn, Fe, Ni, Cu, Zn and Pb), water-soluble ions (SO42 − and NO3−), and organic carbon (OC) and elemental carbon (EC) were analyzed. It was clear that the concentration of particulate matter was higher in winter than in spring. The value of PM2.5/PM10 was > 0.6. Moreover, the proportions of TC, ions, Na, Al, Si and Ca were more high in PM10 and PM2.5. The SOC concentration was estimated by the minimum OC/EC ratio method, and deducted from particulate matter compositions when running CMB model. According to the results of CMB model, the resuspended dust (17.5-35.0%), vehicle exhaust (14.9-23.6%) and secondary particulates (20.4-28.8%) were the major source categories of ambient particulate matter. Additionally, sea salt also had partial contribution (3-8%). And back trajectory analysis results showed that particulate matter was greatly affected by regional sources in winter, while less affected in spring. So particulate matter was not only affected by local sources, but also affected by sea salt and regional sources in coastal cities. Further research could focuses on establishing the actual secondary particles profiles and identifying the local and regional sources of PM at once by one model or analysis method.