top of page



About the Data

 SPARTAN provides open access, pre-generated data sets from all of our sites. SPARTAN is a grass-roots network that started in 2012 and has grown to include 29 active sites (highlighted below in blue). The network will continue to grow in the coming years. 

 SPARTAN sites are intentionally colocated with AERONET Sunphotometers to provide the only global dataset that directly connects ground-based PM2.5 and satellite remote sensing. 

SPARTAN is a grass-roots network run by individual scientists who share their measurements as a public good. We request that you consult our citation policy to credit developers when including SPARTAN data in your publications.


SPARTAN collects measurements from two sampling instruments and provides data that is grouped into three categories: Filter-based, Nephelometer, and Time resolved PM2.5

SPARTAN data can be found at the link below, sorted by category or by site. 

SPARTAN data repository: By Data Product
SPARTAN data repository: By Site

SPARTAN known issues: Document

To see further details about all SPARTAN sites, and site specific data visualizations, navigate to the site page of interest below. 

Filter-based data is derived directly from measurements of particulate matter (PM2.5, PM10, and PMcoarse) samples collected on filters. Chemical speciation includes mass, water-soluble ions, equivalent black carbon, and trace elements. Reconstructed PM2.5, which includes major constituents such as ammonium sulfate, soil, and particle-bound water, is inferred from measured chemical species. For more details on filter-based parameters, see Sampling Parameters and Methods

Nephelometer data includes measurements of backscatter and total scatter at three wavelengths. PM2.5 and PM10 scatter are measured with a nephelometer that includes a cyclone inlet whereas TSP (total suspended particles) scatter does not.

Time resolved PM2.5 combines filter-based PM2.5 mass and total scatter to estimate PM2.5 at a finer resolution than available from filters alone. 

Anchor 1

SPARTAN data citation policy

This information is designed to help you appropriately credit the use of SPARTAN data in your publications. SPARTAN is a grass-roots network run by individual scientists who share their measurements as a public good. Thank you for making the effort to credit developers through co-authorship or citation. The success of SPARTAN depends on your generosity.


Each site has Principal Investigator(s) (PI) who is responsible for instrument operation. PI contact information can be found on the data pages for each SPARTAN site. We encourage you to contact the site PI when data for a particular site are used. Data from all sites are analyzed at Washington University in St. Louis for mass and composition. We also encourage you to contact Chris Oxford and Randall Martin at when SPARTAN data are used.


We recommend the following citations for use of SPARTAN data for referencing our sampling methods and data analysis techniques:

SPARTAN: a global network to evaluate and enhance satellite-based estimates of ground-level particulate matter for global health applications Snider, G. et al. : Atmos. Meas. Tech., 8, 505-521, 2015 , doi:10.5194/amt-8-505-2015.

Variation in Global Chemical Composition of PM2.5: Emerging Results from SPARTAN Snider, G., et al.: Atmos. Chem. Phys. 16, 9629-9653, 2016 doi:10.5194/acp-16-9629-2016.

​Large global variation in measured airborne metal concentrations driven by anthropogenic sources McNeill, J. et al. : Sci. Rep., 10, 21817, 2020 doi:


When publishing data from a few sites, please consider authorship for the site PI(s) and the following acknowledgement:

We thank the SPARTAN project for its effort in establishing and maintaining (site name(s)) sites. Please send a copy of the paper to Chris Oxford, Xuan Liu and Randall Martin at for a quick review prior to submission. Upon final publication, please send a copy of the paper to Chris Oxford at our SPARTAN email address, for posting on the SPARTAN website.


When publishing data from many sites:

If SPARTAN data are the main component of the paper than we encourage offering co-authorship to the following investigators: Chris Oxford, Xuan Liu, Yinon Rudich, Michael Brauer, and Randall V. Martin.

bottom of page