GFAS4HTAP vegetation fire emissions 2003-2023

Overview This dataset contains emission flux from wildfires for various species and combustion rate. The data based on daily dry matter burnt estimates (DM) from CAMS GFASv1.2, downloaded at https://ads.atmosphere.copernicus.eu/datasets/cams-global-fire-emissions-gfas. Subsequently, an updated spurius signal mask is applied and emissions are calculated with a new land cover map derived from ESA CCI and PEATMAP for 2018, and emission factors from NEIVAv1.1 (https://gmd.copernicus.org/articles/17/7679/2024/) and further literature. Each archive contains a folder with the daily emissions for one species in the *_daily.nc file. Other NetCDF files with approximative fields at monthly, annual and 21-year resolution and plots have been added for illustration. The archives of the injection height parameters MAMI and APT contain the Mean Altitude of Maximal Injection and Altitude of Plume Top, respectively. They have been downloade from CAMS GFAS and converted to the standard date format of this reository. A detailed description is available in Remy et al. (2017) at https://acp.copernicus.org/articles/17/2921/2017/. The data is available in netCDF4 format, where the data for each species is contained in individual files.The dataset contains daily data from 01/01/2003 to 31/12/2023 on a regular lat-lon grid with 0.1deg resolution.The date of the time coordinate identifies the validity period. For example for daily data, "2003-01-01 00:00:00" denotes emissions during 00:00:00-23:59:59 UTC of the first of January 2003.All emission data is in [kg m**-2 s**-1]. The CO2 data is the instantaneous emission of CO2 from wildfires. On a longer timescale CO2 will increase due to oxidation of, primarily, CO and CH4. C, PM2.5 and TPC, and only these, constitute a double-counting with other included species. Due to the data volume restriction of Zenodo, "toxic" emissions are provided in this sister repository: 10.5281/zenodo.15721938 A paper explaining the methods and data used in the creation of the dataset is being worked on. Until it has been published, please cite https://gmd.copernicus.org/articles/18/3265/2025/ when using GFAS4HTAP. Calculating emissions locally The G4H archive contains software and static data (emission factor table and land cover mask), with which users can calculate emission consistently with GFAS4HTAP from any dry matter burnt field: Install and activate the conda environment env_g4h.yml, adapt the configuration section in the main() routine of the emissions.py file and run the script. Q&As Q1: Are emissions in beta and v2 for their common periods/species the same? No, all emissions have changed: All are shifted by one day (fixing a "feature" of the CAMS ADS netCDF conversion) and the emmission factor for CO in savannah has been updated. Use of the beta version is discouraged. If bandwidth is an issue consider calculating the emissions locally. Additionally, the metadata in the NetCDF files has been completed. Q2: Should VOCs not explicitly treated in the used model chemistry be ignored or lumped with other species to preserve the total mass? Is NMOC_g the total VOC mass? The NEVIA database includes measurements of a lot of gaseous emissions of larger organic molecules, which have not been represented in emissions estimates or chemical mechanisms in the past. These are reported in the database as NMOC_g. Thus, NMOC_g is the mass of gaseous non methane organic carbon that is NOT included in the mass of other individual or lumped species. It is what is left over in the unspeciated bin after individual species have been accounted for. It can be very large, around half of the organic mass. In other words, total gaseous non-methane organic carbon = sum of all individual VOC species provided + sum of lumped VOC species provided + NMOC_g So, what do you do with NMOC_g or other explicit species that are not in your mechanism when preparing emissions inputs? … It depends on the mechanism that you are putting it into. A reasonable default approach may be to represent as much of the mass of explicit species provided in GFAS4HTAP as makes sense for the proxy/lumping scheme in your mechanism. There is little understanding of how to represent NMOC_g and assigning its mass to other species in the mechanism may well create too much hydrocarbon reactivity. So, a reasonable default approach may be to ignore NMOC_g. Different modelers are going to make different choices, and it will be useful for each model to provide their emissions inputs (total VOC and if possible speciated VOC) along with the outputs for comparison. Q3: The daily file has emission rate in kg/m2/s – Is this a flat rate for the day (GMT)? Yes, this is correct. Q4: Where is the vegetation map? It is part of the package for calculating emissions locally, i.e. in the file G4H.tgz. Q5: Why did the Zenodo link change? Zenodo provides one link/DOI for all version of a repository, which ends with "1". Additionally, each version has its own link/DOI, counting up in the last digit. Q6: How do I get total particulate matter (TPM)? The GFAS4HTAP emissions are based on NEIVA EFs and GFED5 speciation. Most PM measurements are now operationally defined, e.g., based on inlet cutoffs, so TPM is rarely reported. However, if TPM or PM10 are needed, it is recommended take the provided PM2.5 emissions and multiply them by 1.2 (inflate by 20%). Q7: Which enhancement factor should be used for aerosol/PM emissions? It is recommended to tune PM/aerosol emissions to each model setup with (at least) one universal scaling/enhancement parameter. As reference, atmospheric observations targeted by the model can be used. If no such reference is available, the total atmospheric load from an aerosol observation-constrained (re)analysis, e.g. from CAMS, might be use4d as reference. The underlying reason for the need to tune is that the fast aerosol chemistry in the smoke plumes near fires are represented to different degrees in different models and model configurations. List of included species species long_name C carbon combustion (C in CO2, CO, CH4, TPC) CO2 carbon dioxide CO carbon monoxide CH4 methane NMOC_g gaseous non-methane organic compounds not included otherwise H2 hydrogen NOx nitrogen oxides(NOx as NO) N2O nitrous oxide PM2p5 PM 2.5 (particulate matter <2.5u) TPC total particulate carbon (OC+BC) OC organic carbon (carbon in organic matter) BC black carbon SO2 sulfur dioxide C2H6 ethane CH3OH methanol C2H5OH ethanol C3H8 propane C2H2 acetylene C2H4 ethylene C3H6 propylene C5H8 isoprene C10H16 terpenes C7H8 toluene C6H6 benzene C8H10 xylene Higher_Alkenes C4H8 + c5H10 + C6H12 + C8H16 (1 butene + i butene + tr-2-butene + cis-2-butene + 1 pentene + 2 pentene + hexene + octene) Higher_Alkanes C4H10 + C5H12 + C6H14 + C7H16 (n-butane + i-butane + n-pentane + i-pentane(me-butane) + n-hexane + i-hexane + Heptane) CH2O formaldehyde C2H4O acetaldehyde C3H6O acetone NH3 ammonia C2H6S dimethyl sulfide (DMS) HCN hydrogen cyanide HCOOH formic acid CH3COOH acetic acid MEK methyl Ethyl Ketone / 2-butanone CH3COCHO methylglyoxal HOCH2CHO hydroxyacetaldehyde PCDD2378 2,3,7,8-TeCDD PCDD12378 1,2,3,7,8-PeCDD PCDD123478 1,2,3,4,7,8-HxCDD PCDD123678 1,2,3,6,7,8-HxCDD PCDD123789 1,2,3,7,8,9-HxCDD PCDD1234678 1,2,3,4,6,7,8-HpCDD OCDD OctaCDD PCDF2378 2,3,7,8-TeCDF PCDF12378 1,2,3,7,8-PeCDF PCDF23478 2,3,4,7,8-PeCDF PCDF123478 1,2,3,4,7,8-HxCDF PCDF123678 1,2,3,6,7,8-HxCDF PCDF123789 1,2,3,7,8,9-HxCDF PCDF234678 2,3,4,6,7,8-HxCDF PCDF1234678 1,2,3,4,6,7,8-HpCDF PCDF1234789 1,2,3,4,7,8,9-HpCDF OCDF OctaCDF NAP Naphthalene ACY Acenaphthylene ACE Acenaphthene FLO Fluorene PHE Phenanthrene ANT Anthracene FLA Fluoranthene PYR Pyrene BaA Benz(a)anthracene CHR Chrysene BbF Benzo(b)fluoranthene BkF Benzo(k)fluoranthene BaP Benzo(a)pyrene IcdP Indeno(1,2,3-cd)pyrene DahA Dibenz(a,h)anthracene BghiP Benzo(g,h,i)perylene Hg Mercury as Hg0+HgP List of other parameters short name description MAMI Mean Altitude of Maximal Injection APT Altitude of Plume Top G4H software for calculating enissions locally, including emission factor table and land cover mask