GHOST: A globally harmonised dataset of surface atmospheric composition measurements

GHOST: Globally Harmonised Observations in Space and Time, represents one of the biggest collection of harmonised measurements of atmospheric composition at the surface. In total, 7,275,148,646 measurements from 1970-2023, of 227 different components, from 38 reporting networks, are compiled, parsed, and standardised. Components processed include gaseous species, total and speciated particulate matter, and aerosol optical properties.

The main goal of GHOST is to provide a dataset that can serve as a basis for the reproducibility of model evaluation efforts across the community. Exhaustive efforts have been made towards standardising almost every facet of provided information from the major public reporting networks, saved in 21 data variables, and 163 metadata variables. Extensive effort in particular is put towards the standardisation of measurement process information, and station classifications. Extra complementary information is also associated with measurements, such as metadata from various popular gridded datasets (e.g. land use), and temporal classifications per measurement (e.g. day / night). A range of standardised network quality assurance flags are associated with each individual measurement. GHOST own quality assurance is also performed and associated with measurements. Measurements prefiltered by some default GHOST quality assurance are also provided.  

The dataset is 1.39 TB in total size (121 GB compressed). The data is separated out per network, per temporal resolution, per component, and is saved as netCDF4 files, per year and month. There is additionally one synthetic network entitled "GHOST-PUBLIC", which aggregates data across all networks. The dataset is compressed as .zip files per network. Beneath each network, collections of files per temporal resolution, per component, are compressed as tar.xz files.

 

Each network .zip file can be decompressed via the following syntax:
unzip [network].zip

 

Component tar.xz files can be decompressed via the following syntax:
tar -xf [component].tar.xz

If you have any other doubts of queries regarding the dataset, please email: dene.bowdalo@bsc.es

We gratefully acknowledge all data providers for the substantial work done in establishing and maintaining the measuring stations that provide the data contained in this dataset. We would also like to warmly thank all data providers who met with GHOST authors through this work, and for all support given, from helping resolve data rights issues, to giving suggestions for improvements.

We acknowledge the computing resources of MareNostrum, and the technical support provided by the Barcelona Supercomputing Center (AECT-2020-1-0007, AECT-2021-1-0027, AECT-2022-1-0008, and AECT-2022-3-0013). We also acknowledge the Red Temática ACTRIS España (CGL2017-90884-REDT), and the H2020 project ACTRIS IMP (\#871115).

The research leading to the creation of this dataset has received funding from the grant RTI2018-099894-BI00 funded by MCIN/AEI/ 10.13039/501100011033 (BROWNING), the EU H2020 Framework Programme under grant agreement No. GA 821205 (FORCES), the European Research Council under the Horizon 2020 research and innovation programme through the ERC Consolidator Grant grant agreement No. 773051 (FRAGMENT), the AXA Research Fund (AXA Chair on Sand and Dust Storms at the Barcelona Supercomputing Center), and the Department of Research and Universities of the Government of Catalonia through the Atmospheric Composition Research Group (code 2021 SGR 01550).