This dataset supports the State of Wildfires 2023-24 report under review at Earth System Science Data Discussions (Jones et al., under review, https://doi.org/10.5194/essd-2024-218). The dataset provides annual data and final-year anomalies in burned area (BA), fire carbon (C) emissions, and fire properties (e.g. distributional statistics for fire count, size, rate of growth). Annual data relate to the global fire season defined as March-February (e.g., March 2023-February 2024), aligning with an annuall lull in the global fire calendar (see Jones et al., 2024). The complete methodology is described by Jones et al. (2024).

Citation

Work utilising our regional summaries should cite both Jones et al. (2024, under review, ESSD) AND the primary reference for the variable(s) of interest as follows:

• Giglio et al. (2018) for MODIS MCD64A1 BA.
• van der Werf et al. (2017) for GFED4.1s fire C emissions.
• Kaiser er al. (2012) for GFAS fire C emissions.
• van der Werf et al. (2017) AND Kaiser er al. (2012) for the average of GFED4.1s and GFAS fire C emissions.
• Andela et al. (2019) for the Global Fire Atlas.

Input Data

Burned Area (BA)

• BA data from NASA’s MODIS BA product (MCD64A1) are extended from Giglio et al. (2018) and are available at Giglio et al. (2021, https://lpdaac.usgs.gov/products/mcd64a1v061/).
  • Period: 2001-February 2024
  • Resolution: 500m

Fire Carbon (C) Emissions

• GFED4.1s fire C emissions data are extended from van der Werf and are available at https://globalfiredata.org/.
  • Period: 2003-February 2024
  • Resolution: 0.25 degree, daily

• GFAS fire C emissions data are extended from Kaiser et al. (2012) and are available at https://confluence.ecmwf.int/display/CKB/CAMS+global+biomass+burning+emissions+based+on+fire+radiative+power+%28GFAS%29%3A+data+documentation.
  • Period: 2003-February 2024
  • Resolution: 0.1 degree, daily

Global Fire Atlas (Individual Fire Atlas and Properties)

• Global Fire Atlas are extended from Andela et al. (2019) and are available at Andela and Jones (2024, https://doi.org/10.5281/zenodo.11400062, last access: 31 May 2024).
  
  • Period: 2002-February 2024
  • Driven by 500m MODIS BA data (collection 6.1)

Regional Analysis

We performed "cookie-cutting" (spatial and temporal masking) of the above input data sets to features in each of the following regional layers (e.g. per country in the "Countries" layer).

The statistics derived from cookie-cutting are listed below. Full details in Jones et al. (2024).

Regional Statistics and Anomalies

• Burned Area (BA)
  • Calculated regional totals for each fire season.
  • Relative and standardized anomalies from historical data (since 2001).
  • Ranking amongst all recorded fire seasons.
  • Onset, peak, and cessation based on monthly deviations from climatological means.

• Carbon Emissions
  • Calculated regional totals for each fire season.
  • Relative and standardized anomalies from historical data (since 2003).
  • Ranking amongst all recorded fire seasons.
  • Onset, peak, and cessation based on monthly deviations from climatological means.
  • Statistics available for GFAS, GFED, and their mean.

• Individual Fire Properties
  • Based on ignition point vectors from the Global Fire Atlas.
  • Calculated regional count.
  • Calculated regional maxima and 95th percentiles for each fire season.
  • Relative and standardized anomalies from historical data (since 2002).
  • Ranked anomalies among all recorded fire seasons.

ABSTRACT: This data set provides monthly burned area, and monthly, and annual fire emissions data from July 1996 to February 2012. Emissions data are available for carbon (C), dry matter (DM), carbon dioxide (CO2), carbon monoxide (CO), methane (CH4), hydrogen (H2), nitrous oxide (N2O), nitrogen oxides (NOx), non-methane hydrocarbons (NMHC), organic carbon (OC), black carbon (BC), particulate matter 2.5 micron (PM2p5), total particulate matter (TPM), and sulfur dioxide (SO2). The C4 fraction of carbon emissions is also provided. The annual C emissions estimates were derived by combining burned area data with a biogeochemical model, CASA-Global Fire Emissions Database (CASA-GFED), that estimates fuel loads and combustion completeness for each monthly time step. The fuel loads were based on satellite derived information on vegetation characteristics and productivity to estimate carbon input and carbon outputs through heterotrophic respiration, herbivory, and fires. Note that while most emissions estimates included data for 32 variables (trace gases, aerosols, and carbon), not all data are available for all years, and not all variables (emission species) are included in each data product. Additional information may be obtained from the Global Fire Data website: http://www.globalfiredata.org/index.html. Data products include: - 0.5 degree x 0.5 degree gridded monthly burned area data (ha) for 1996 to 2012 provided as text files and as GeoTIFF files for 1996 to 2012. - 3-Hourly emssions (fraction) for 2003 to 2010 in NetCDF (.nc) format. - Daily emssions (fraction) for 2003 to 2010, in NetCDF (.nc) format. - Monthly emissions for 32 variables from 1997 to 2011, in text and GeoTIFF format. - Monthly emissions for 31 variables from specific sources (grassland and savanna, woodland, deforestation & degradation, forest, agricultural waste burning, and peat fires), both as absolute and relative emissions. The time period is for 2007 to 2011, and the files are provided in text and GeoTIFF format. - Global emission totals of C and other species from all sources, and from each individual source (forest fires, peat fires, agricultural waste burning, etc). - Annual emissions of carbon and other trace gases for all countries, for the period 1997 to 2010, provided as text files. These files are for indicative use only; they are not suitable for official reporting due to large uncertainties and potential for missing key regional aspects in the global approach used. - Ancillary data for monthly biosphere fluxes. The CASA-GFED biosphere flux sources include Net Primary Production (NPP), Heterotrophic respiration (Rh), and fires (biomass burning). These files are for the time period 1997 to 2009 and are provided as text files and in GeoTIFF format. There are 12 compressed (*.zip) files with this data set. The data are in text, NetCDF (.nc), and GeoTIFF (.tiff) formats as described above.

Satellites provide direct observations of fire activities (e.g., burned area, fire radiative power) (Andela et al., 2017; Giglio et al., 2006; Luo et al., 2024), but their temporal coverages are limited because most satellite data were available only after 1980s (Chuvieco et al., 2019). Fire modules in dynamic global vegetation models (DGVMs) are able to simulate long-term burned area and the interactions with vegetation dynamics based on climate conditions and soil properties (Sitch et al., 2015; Sitch et al., 2024), but the spatial resolution at the global scale is usually coarse due to the coarse resolution of the input meteorological forcing data, and most models failed in capturing the global trends of burned area (Andela et al., 2017; Hantson et al., 2020). The processes included and the parameterizations of fire processes are widely different across fire models, resulting in a large range of simulated burned area at both the regional and global scales (Hantson et al., 2020). Considering the limitations of satellite observations and fire models, a spatiotemporally consistent burned area dataset over the 20th century trained from present-day observations, is essential for fire modelling and can serve as publicly available benchmark for fire ecology and carbon cycle studies.

This study produced a global monthly 0.5°×0.5° burned area fraction (BAF) dataset from1901 to 2020 using machine learning models based on climate conditions, vegetation states, population density and land use data. We first divided the globe into 14 regions following the Global Fire Emission Dataset (GFED regions) (Giglio et al., 2006; van Der Werf et al., 2017) and conducted all steps described below in each GFED region individually. To better capture extreme fires, we first developed a classification model to distinguish grid cells with extreme and regular fires, using the 90th percentile of all burned area fractions within a region as the threshold to define extreme fires. We then trained separate regression models for grid cells categorized as having extreme or regular fires. The models were trained against the satellite-based burned area product (FireCCI51) during 2003-2020 excluding cropland fires, and then used to reconstruct the burned area from 1901 to 2020. In addition to validation against satellite observations that were not used for model training, we also compared our burned area predictions with charcoal records and other independent global and regional burned area datasets. In addition to the historical reconstructed burned area dataset based on the FireCCI51 mentioned above, we also produced two additional products of historical burned area with the same spatiotemporal resolution as the FireCCI51-based burned area reconstruction: 1) the GFED5-based data version, which is based on machine-learning models trained by the burned area from GFED5 which has much more fires than GFED4 (Chen et al., 2023) instead of FireCCI51, and 2) the FireCCI51-based data with burned area further calibrated using the relationship between statistic-based burned area (Mouillot and Field, 2005) and GDP (Bolt and Van Zanden, 2024) at the regional scale before 2000 (named as FireCCI51-GDP version).

This dataset can be used to benchmark historical simulations from fire modules in DGVMs, re-calculate historical fire emissions and estimate legacy effects of vegetation recovery after fires on terrestrial carbon sink. Though the temporal coverage of our product is long enough to support studies related to fire disturbance, carbon dynamics and climate change, more reliable explanatory data for model training and burned area data for validation would help further improve the accuracy of the reconstructed burned area product.

Description of variable names in netcdf files:
'lat' -- latitude of the center of grid cells
'lon' -- longitude of the center of grid cells
'ba' -- burned area fraction in 0.5°×0.5°grid cells
'ba_type' -- 0=no fire; 1=regular burned area; 2=extreme burned area. Note that regular & extreme BA is defined in each GFED region individually. Extreme BA refers to that burned area within grid cells surpassing the 90th percentile of all grid cells with burned area in each GFED region.

This dataset is currently for manuscript submission to the scientifc journal

Contains two separate datasets:

1. Global GFED-based monthly burned area (in ha) time series (1996-2016) at 1 km (downscaled using cubic-splines from 25 km);
2. Global burned area long term (2000-2012) P90 (quantile probability = 0.9) based on the ESA CCI burned area accumulated weekly product;

Original GFED monthly data is provided as HDF4 files (ftp.fuoco.geog.umd.edu/data/GFED/GFED4). Dataset is described in detail in Giglio et al. (2013). Processing steps are available here. Antarctica is not included.

To access and visualize global datasets use: https://openlandmap.org or watch this video.

If you discover a bug, artifact or inconsistency in the maps, or if you have a question please use some of the following channels:

• Technical issues and questions about the code: https://gitlab.com/openlandmap/global-layers/issues

All files provided as Cloud-Optimized GeoTIFFs / internally compressed using "COMPRESS=DEFLATE" creation option in GDAL. File naming convention:

• nhz = theme: natural hazards,
• monthly.burned.ha = variable: estimated monthly burned area in ha,
• gfed = data source GFED data,
• m = mean value,
• 1km = spatial resolution / block support: 1 km,
• s0..0cm = vertical reference: land surface,
• 2000.02 = time reference aggregated: month Feb of year 2000,
• v4 = version number: GFEDv4,

description: The Landsat Burned Area Product Validation dataset was collected to determine the accuracy of The Landsat Burned Area Essential Climate Variable (BAECV) product, developed by the U.S. Geological Survey (USGS). The BAECV maps burned areas across the conterminous United States (CONUS) for the entire Landsat archive (1984 2015). Rigorous validation of such products is critical for their proper usage and interpretation. The sampling design used to derive this validation dataset was adapted from the methods used by European Space Agency s (ESA) Climate Change Initiative (CCI) fire_cci project to generate the first statistically rigorous global reference dataset for a burned area product that meets the CEOS LPVS stage 3 validation requirements. Our validation dataset consists of 28 Landsat path/rows across the CONUS which were selected using a stratified sampling scheme across the major Olson biomes, as summarized by the fire_cci project (Olson et al., 2001; Padilla et al. 2014). Within the CONUS this included temperate forest, Mediterranean forest, temperate grassland and savannah, tropical and subtropical grasslands and savannah, and other which included desert/xeric shrub and flooded grasslands (Padilla et al. 2014). Path/rows selected within each biome were meant to represent high and low burned areas as specified by the Global Fire Emissions Database (GFED) version 3 (Giglio et al, 2009, 2010). We used systematic sampling to select 5 validation years spaced out in 5 year increments (2008, 2003, 1998, 1993 and 1988). The validation dataset was then independently generated by three different analysts. Each analyst mapped new burned areas using Landsat pre-fire and post-fire image pairs. The burned area polygons were generated using the Burned Area Mapping Software (BAMS), which is a semi-automated algorithm developed by the University of Alcala, Madrid, and implemented by the fire_cci project (Bastarrika et al., 2014; Padilla et al., 2014). The outputs were manually edited using visual interpretation. From these outputs, three renditions of the validation datasets were generated in which burned area extent ranged from liberal (or inclusive) (Level 1) to conservative (Level 3). Burned area extent was defined as (1) at least one analyst identified a given pixel as burned (Level 1), (2) at least two of the three analysts were required to agree a given pixel was burned (Level 2), (3) all three analysts were required to agree a pixel was burned (Level 3). Full details of the methods used to derive this validation dataset are provided in Vanderhoof et al. (2017).; abstract: The Landsat Burned Area Product Validation dataset was collected to determine the accuracy of The Landsat Burned Area Essential Climate Variable (BAECV) product, developed by the U.S. Geological Survey (USGS). The BAECV maps burned areas across the conterminous United States (CONUS) for the entire Landsat archive (1984 2015). Rigorous validation of such products is critical for their proper usage and interpretation. The sampling design used to derive this validation dataset was adapted from the methods used by European Space Agency s (ESA) Climate Change Initiative (CCI) fire_cci project to generate the first statistically rigorous global reference dataset for a burned area product that meets the CEOS LPVS stage 3 validation requirements. Our validation dataset consists of 28 Landsat path/rows across the CONUS which were selected using a stratified sampling scheme across the major Olson biomes, as summarized by the fire_cci project (Olson et al., 2001; Padilla et al. 2014). Within the CONUS this included temperate forest, Mediterranean forest, temperate grassland and savannah, tropical and subtropical grasslands and savannah, and other which included desert/xeric shrub and flooded grasslands (Padilla et al. 2014). Path/rows selected within each biome were meant to represent high and low burned areas as specified by the Global Fire Emissions Database (GFED) version 3 (Giglio et al, 2009, 2010). We used systematic sampling to select 5 validation years spaced out in 5 year increments (2008, 2003, 1998, 1993 and 1988). The validation dataset was then independently generated by three different analysts. Each analyst mapped new burned areas using Landsat pre-fire and post-fire image pairs. The burned area polygons were generated using the Burned Area Mapping Software (BAMS), which is a semi-automated algorithm developed by the University of Alcala, Madrid, and implemented by the fire_cci project (Bastarrika et al., 2014; Padilla et al., 2014). The outputs were manually edited using visual interpretation. From these outputs, three renditions of the validation datasets were generated in which burned area extent ranged from liberal (or inclusive) (Level 1) to conservative (Level 3). Burned area extent was defined as (1) at least one analyst identified a given pixel as burned (Level 1), (2) at least two of the three analysts were required to agree a given pixel was burned (Level 2), (3) all three analysts were required to agree a pixel was burned (Level 3). Full details of the methods used to derive this validation dataset are provided in Vanderhoof et al. (2017).

The SeasFire Cube is a scientific datacube for seasonal fire forecasting around the globe. Apart from seasonal fire forecasting, which is the aim of the SeasFire project, the datacube can be used for several other tasks. For example, it can be used to model teleconnections and memory effects in the earth system. Additionally, it can be used to model emissions from wildfires and the evolution of wildfire regimes.

It has been created in the context of the SeasFire project, which deals with "Earth System Deep Learning for Seasonal Fire Forecasting" and is funded by the European Space Agency (ESA) in the context of ESA Future EO-1 Science for Society Call.

It contains 21 years of data (2001-2021) in an 8-days time resolution and 0.25 degrees grid resolution. It has a diverse range of seasonal fire drivers. It expands from atmospheric and climatological ones to vegetation variables, socioeconomic and the target variables related to wildfires such as burned areas, fire radiative power, and wildfire-related CO2 emissions.

Datacube properties

Feature

Value

Spatial Coverage

Global

Temporal Coverage

2001 to 2021

Spatial Resolution

0.25 deg x 0.25 deg

Temporal Resolution

8 days

Number of Variables

54

Tutorial Link

https://github.com/SeasFire/seasfire-datacube

    Full name
    DataArray name
    Unit
    Contact *




    Dataset: ERA5 Meteo Reanalysis Data





    Mean sea level pressure
    mslp
    Pa
    NOA


    Total precipitation
    tp
    m
    MPI


    Relative humidity
    rel_hum
    %
    MPI


    Vapor Pressure Deficit
    vpd
    hPa
    MPI


    Sea Surface Temperature
    sst
    K
    MPI


    Skin temperature
    skt
    K
    MPI


    Wind speed at 10 meters
    ws10
    m*s-2
    MPI


    Temperature at 2 meters - Mean
    t2m_mean
    K
    MPI


    Temperature at 2 meters - Min
    t2m_min
    K
    MPI


    Temperature at 2 meters - Max
    t2m_max
    K
    MPI


    Surface net solar radiation
    ssr
    MJ m-2
    MPI


    Surface solar radiation downwards
    ssrd
    MJ m-2
    MPI


    Volumetric soil water level 1
    swvl1
    m3/m3
    MPI







          Volumetric soil water level 2




    swvl2
    m3/m3
    MPI


    Volumetric soil water level 3
    swvl3
    m3/m3
    MPI


    Volumetric soil water level 4
    swvl4
    m3/m3
    MPI


    Land-Sea mask
    lsm
    0-1
    NOA


    Dataset: Copernicus

CEMS

    Drought Code Maximum
    drought_code_max
    unitless
    NOA


    Drought Code Average
    drought_code_mean
    unitless
    NOA


    Fire Weather Index Maximum
    fwi_max
    unitless
    NOA


    Fire Weather Index Average
    fwi_mean
    unitless
    NOA


    Dataset: CAMS: Global Fire Assimilation System (GFAS)





    Carbon dioxide emissions from wildfires
    cams_co2fire
    kg/m²
    NOA


    Fire radiative power
    cams_frpfire
    W/m²
    NOA


    Dataset: FireCCI - European Space Agency’s Climate Change Initiative





    Burned Areas from Fire Climate Change Initiative (FCCI)
    fcci_ba
    ha
    NOA


    Valid mask of FCCI burned areas
    fcci_ba_valid_mask
    0-1
    NOA



    Fraction of burnable area
    fcci_fraction_of_burnable_area
    %
    NOA


    Number of patches
    fcci_number_of_patches
    N
    NOA


    Fraction of observed area
    fcci_fraction_of_observed_area
    %
    NOA


    Dataset: Nasa MODIS MOD11C1, MOD13C1, MCD15A2





    Land Surface temperature at day
    lst_day
    K
    MPI


    Leaf Area Index
    lai
    m²/m²
    MPI


    Normalized Difference Vegetation Index
    ndvi
    unitless
    MPI


    Dataset: Nasa SEDAC Gridded Population of the World (GPW), v4





    Population density
    pop_dens
    persons per square kilometers
    NOA


    Dataset: Global Fire Emissions Database (GFED)





    Burned Areas from GFED (large fires only)
    gfed_ba
    hectares (ha)
    MPI


    Valid mask of GFED burned areas
    gfed_ba_valid_mask
    0-1
    NOA


    GFED basis regions
    gfed_region
    N
    NOA


    Dataset: Global Wildfire Information System (GWIS)





    Burned Areas from GWIS
    gwis_ba
    ha
    NOA


    Valid mask of GWIS burned areas
    gwis_ba_valid_mask
    0-1
    NOA


    Dataset: NOAA Climate Indices





    Arctic Oscillation Index
    oci_ao
    unitless
    NOA


    Western Pacific Index
    oci_wp
    unitless
    NOA


    Pacific North American Index
    oci_pna
    unitless
    NOA


    North Atlantic Oscillation
    oci_nao
    unitless
    NOA


    Southern Oscillation Index
    oci_soi
    unitless
    NOA


    Global Mean Land/Ocean Temperature
    oci_gmsst
    unitless
    NOA


    Pacific Decadal Oscillation
    oci_pdo
    unitless
    NOA


    Eastern Asia/Western Russia
    oci_ea
    unitless
    NOA


    East Pacific/North Pacific Oscillation
    oci_epo
    unitless
    NOA


    Nino 3.4 Anomaly
    oci_nino_34_anom
    unitless
    NOA


    Bivariate ENSO Timeseries
    oci_censo
    unitless
    NOA


    Dataset: ESA CCI





    Land Cover Class 0 - No data
    lccs_class_0
    %
    NOA


    Land Cover Class 1 - Agriculture
    lccs_class_1
    %
    NOA


    Land Cover Class 2 - Forest
    lccs_class_2
    %
    NOA


    Land Cover Class 3 - Grassland
    lccs_class_3
    %
    NOA


    Land Cover Class 4 - Wetlands
    lccs_class_4
    %
    NOA


    Land Cover Class 5 - Settlement
    lccs_class_5
    %
    NOA


    Land Cover Class 6 - Shrubland
    lccs_class_6
    %
    NOA


    Land Cover Class 7 - Sparse vegetation, bare areas, permanent snow and ice
    lccs_class_7
    %
    NOA


    Land Cover Class 8 - Water Bodies
    lccs_class_8
    %
    NOA


    Dataset: Biomes





    Dataset: Calculated





    Grid Area in square meters
    area
    m²
    NOA

*The datacube specifications (temporal, spatial resolution, chunk size) have been set up by the Max Planck Institut (MPI) team. For the variables that the contact is MPI, Lazaro Alonso (lalonso bgc-jena.mpg.de) has led the efforts to collect and process them. For the variables that the contact is NOA, Ilektra Karasante (ile.karasante noa.gr) has led the efforts to collect and process them.

MiCASA is an extensive revision of CASA-GFED3. CASA-GFED3 derives from Potter et al. (1993), diverging in development since Randerson et al. (1996). CASA is a light use efficiency model: NPP is expressed as the product of photosynthetically active solar radiation, a light use efficiency parameter, scalars that capture temperature and moisture limitations, and fractional absorption of photosynthetically active radiation (fPAR) by the vegetation canopy derived from satellite data. Fire parameterization was incorporated into the model by van der Werf et al. (2004) leading to CASA-GFED3 after several revisions (van der Werf et al., 2006, 2010). Development of the GFED module has continued, now at GFED5 (Chen et al., 2023) with less focus on the CASA module. MiCASA diverges from GFED development at version 3, although future reconciliation is possible. Input datasets include air temperature, precipitation, incident solar radiation, a soil classification map, and several satellite derived products. These products are primarily based on Moderate Resolution Imaging Spectroradiometer (MODIS) Terra and Aqua combined datasets including land cover classification (MCD12Q1), burned area (MCD64A1), Nadir BRDF-Adjusted Reflectance (NBAR; MCD43A4), from which fPAR is derived, and tree/herbaceous/bare vegetated fractions from Terra only (MOD44B). Emissions due to fire and burning of coarse woody debris (fuel wood) are estimated separately.

Synthesis of fuel load and fuel consumption field measurements accompanying the publication:

"Global biomass burning fuel consumption and emissions at 500-m spatial resolution based on the Global Fire Emissions Database (GFED)"

Dave van Wees1, Guido R. van der Werf1, James T. Randerson2, Brendan M. Rogers3, Yang Chen2, Sander Veraverbeke1, Louis Giglio4, and Douglas C. Morton5

1Department of Earth Sciences, Vrije Universiteit, Amsterdam, 1081 HV, The Netherlands
2Department of Earth System Science, University of California, Irvine, CA 92697, USA
3Woodwell Climate Research Center, Falmouth, MA 02540, USA
4Department of Geographical Sciences, University of Maryland, College Park, MD 20742, USA
5Biospheric Sciences Laboratory, NASA Goddard Space Flight Center, Greenbelt, MD 20771, USA

DOI: https://doi.org/10.5194/gmd-15-8411-2022

Units are g C / m2

This is a 16 member ensemble of simulations with CESM2 under the SSP2-4.5 forcing scenario from 2015 to 2100. These simulations can be compared with the CESM2 Large Ensemble and provide the opportunity to compare and contrast climate change under a lower forcing scenario. One difference from the official CMIP6 SSP2-4.5 forcing is that slightly modified biomass burning emissions are used at the beginning of the simulation. As is discussed in the CESM2 Large Ensemble reference paper (Rodgers et. al. 2021), the second 50 members of the CESM2 Large Ensemble use smoothed biomass burning emissions over the GFED era of the late 20th/early 21st centuries. While the GFED emissions were not prescribed in the SSP scenario, there is a minor effect of the smoothing into the first years of the SSP scenario and these simulations have been branched from historical simulations that used the smoothed biomass burning emissions. As such, this medium ensemble is complementary to the second 50 member of the CESM2 Large Ensemble. More information about these simulations can be found on the CESM CVCWG's CESM2 2-4.5 ensemble website.

Data generated for the paper entitled "Subtropical southern Africa fire emissions of nitrogen oxides and ammonia obtained with satellite observations and GEOS-Chem" submitted for peer-review to Royal Society's Environmental Science: Atmospheres journal.All the data are gridded to 0.25 degree latitude by 0.3125 degree longitude for the year 2019.The data include:(1) IASI NH3 monthly mean column densities for June-October from IASI instruments onboard MetOp-A and MetOp-B sensors reprocessed using GEOS-Chem a priori profiles. The IASI retrieval product used is version 4.0.0. Data are in files named with the format "iasi-metopa-metopb-gc-prior-[inventory]-nh3-cols.nc", where [inventory] is either "finnv25" for FINN version 2.5, "gfedv4s" for GFED version 4 with small fires, or "gfasv12" for GFAS version 1.2 for the three separate biomass burning inventories that were used to drive the model. (2) Top-down NOx and NH3 biomass burning emissions for 2019 obtained by applying a mass-balance approach to satellite observations and GEOS-Chem. NOx emissions are for June-October. NH3 emissions are for July-October. Only non-negative NH3 emissions are summed and use to report monthly and biomass burning season totals in the accompanying manuscript. Emissions data are in files named with the format "top-down-bb-[compound]-emis.nc", where [compound] is either NOx or NH3.(3) GEOS-Chem monthly mean tropospheric column densities of NO2 and total column densities of NH3 for June-October 2019. These are obtained by co-sampling with TROPOMI and applying the TROPOMI averaging kernel for NO2 and co-sampling with IASI for NH3. Model data are averaged over 13h00-14h00 local solar time (LST) for coincidence with IASI and 09h00-10h00 LST for coincidence with IASI. Model data are in files named with the format "geos-chem-no2-nh3-cols-[inventory].nc", where [inventory] is either "finnv25" for FINN version 2.5, "gfedv4s" for GFED version 4 with small fires, or "gfasv12" for GFAS version 1.2.