Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2025 Sep 12;2(10):2071–2080. doi: 10.1021/acsestair.5c00035

Effects of Fire Plume Height on the Geophysical Estimation of Surface Fine Particulate Matter from Satellite Aerosol Optical Depth during North American Wildfires

Inderjeet Singh †,*, Randall V Martin †, Aaron van Donkelaar †, Chi Li †, Yuanjian Zhang †, Haihui Zhu †, Dandan Zhang †, Alexei Lyapustin ‡
PMCID: PMC12520165  PMID: 41098629

Abstract

Wildfires can inject smoke at high altitudes into the atmosphere. The resulting free tropospheric aerosols may affect inference of ground-level fine particulate matter (PM2.5) from satellite aerosol optical depth (AOD), yet the effects of accounting for plume height in this inference are poorly understood. Here, we include in the GEOS-Chem chemical transport model a fire plume height parametrization (GFAS, Global Fire Assimilation System) to examine its effect on PM2.5 inferred from satellite AOD during wildfires over the United States and Canada. Comparison with six years satellite observations of plume height reveals a low bias of a factor 1.7 in the GFAS plume height over evergreen needleleaf forests. We scale the GFAS plume height over evergreen needleleaf forests in GEOS-Chem to better represent the satellite observations, focusing on 2018 and 2020 when large wildfires yield prominent signals. Replacing the default ground-level wildfire emissions in GEOS-Chem with the scaled GFAS vertically distributed emissions reduces the bias between measured PM2.5 and PM2.5 inferred from satellite AOD, and significantly improves the consistency of simulated AOD with sun photometer measurements. Overall, this study signifies the importance of vertically distributing wildfire emissions for the inference of PM2.5 from satellite AOD.

Keywords: wildfire plume height, chemical transport modeling, PM2.5 , AOD, air quality, GEOS-Chem

graphic file with name ea5c00035_0010.jpg

graphic file with name ea5c00035_0008.jpg

Introduction

Exposure to fine particulate matter (PM2.5) is a leading environmental risk factor for respiratory and cardiovascular morbidity and mortality across the United States and Canada. − Gaps in ground-based monitoring motivate alternative measures of PM2.5 from satellite remote sensing of aerosol optical depth (AOD). However, the relationship of satellite AOD with surface PM2.5 concentration is sensitive to the vertical profile of atmospheric aerosols. , Accurate representation of the vertical distribution of aerosols is especially important during wildfires when smoke can reach high altitudes. − Inclusion of the wildfire plume height into chemical transport models used to infer PM2.5 from satellite AOD could improve the resulting satellite-derived PM2.5 data.

Wildfire smoke plume height (H) determines whether emitted smoke is a localized phenomenon limited to the atmospheric boundary layer or is injected into the free troposphere where it can travel long distances. This information can either be derived from satellite observations or from physical models solving a set of mathematical equations (e.g., 1-D Navier–Stokes). Satellite retrievals of plume height have revealed that wildfires can inject substantial quantities of smoke into the free troposphere. ,, Physical models have revealed that elevated plumes greatly affect the long-distance transport of wildfire smoke. , These physical models have included plume dynamics models such as a plume rise model (PRM) that offers information on plume vertical structure and a semiempirical model such as the integrated monitoring and modeling system for wildland fires (IS4FIRES) that provide plume height estimates. The additional information offered by PRM is attractive, but care is needed to evaluate its accuracy versus independent observations.

A recent source of satellite plume height with daily observations includes the Earth Polychromatic Imaging Camera (EPIC) onboard the Deep Space Climate Observatory (DSCOVR) spacecraft at the Sun-Earth L1 Langrange point. An attribute of this vantage point is the ability to continuously observe the sunlit earth, which is advantageous over other space-borne plume height sources with lower sampling frequency such as the Multiangle Imaging Spectroradiometer (MISR), the Multi-Angle Implementation of Atmospheric Correction (MAIAC), the Tropospheric Monitoring Instrument (TROPOMI), and the Cloud-Aerosol LiDAR with Orthogonal Projection (CALIOP). This high measurement frequency from EPIC is also enabling other air quality applications. −

Many studies involving effects of wildfires on surface PM2.5 do not consider smoke plume height; − however, a growing body of literature is drawing attention to this issue. Feng et al. used IS4FIRES plume top height to quantify the fraction of the plume in the planetary boundary layer, applied the resulting fraction to the Stochastic Time-Inverted Lagrangian Transport (STILT) model to assess the impact of Australian wildfires on surface concentration of PM2.5, and found that the plume height information greatly affects the daily PM2.5 estimates, especially during intense wildfire events. Cheeseman et al. examined the relationship of plume height-to-boundary layer height ratio versus PM2.5-to-AOD ratio and concluded that fire plumes leaving the boundary layer have little impact on surface PM2.5 concentration. In this study, we investigate the impacts of implementing smoke plume height in the GEOS-Chem chemical transport model on the inference of surface PM2.5 inferred from satellite AOD.

Data

Wildfires Characterization

Our study focuses on the US and Canada where dense ground-based measurements for AOD and PM2.5 facilitate evaluation during extreme fire periods when signals are pronounced. The NOAA’s National Center for Environmental Information (NCEI) (https://www.ncei.noaa.gov/access/monitoring/monthly-report/fire/202013, https://www.ncei.noaa.gov/access/monitoring/monthly-report/fire/201813, last accessed: July 29, 2024) annual report indicates that the years 2020 and 2018 recorded the most and second most land area burned per fire over the period 2000–2022. We focus on the summer and fall seasons, when wildfires were most extreme for 2018 and 2020. During these periods using the data from the National Interagency Fire Center (NIFC) and Canadian National Fire Database (CNFDB), we further classify as extreme fire event months if land area burned exceeds the 20 years average. This criterion yields July to November in 2018, and August to December in 2020 as extreme fire event months (Tables S1 and S2).

Satellite AOD and Plume Height

We use daily AOD data obtained from the Multi-Angle Implementation of Atmospheric Correction (MAIAC) atmospheric product MCD19A2. The MAIAC algorithm uses time series analysis and image processing to derive the surface bidirectional reflectance function at fine resolution and subsequentially calculates AOD at 1 km spatial resolution from the MODIS satellite instruments. The MODIS instruments onboard NASA’s Terra and Aqua satellites provide global coverage of AOD on a near-daily basis.

We use plume height information from the EPIC instrument (https://portal.nccs.nasa.gov/datashare/maiac/DataRelease/EPIC/, last accessed: June 2024) onboard Deep Space Climate Observatory (DSCOVR). EPIC captures images of the sunlit disk of Earth every 60–100 min from the L-1 Lagrange point between the Earth and Sun at a resolution of 2048 by 2048 pixels, with a pixel resolution of about 8 km at the center of the disk. The EPIC instrument detects the Earth reflected solar radiance at 10 different wavelengths spanning from the ultraviolet to the near-infrared part of the electromagnetic spectrum. Among these wavelength channels, two wavelengths 764 nm (O2 A absorption band) and 688 nm (O2 B absorption band) are useful in sensing the aerosol and cloud heights. The EPIC retrieval algorithm utilizes these O2 A and B absorption band and their respective continuum band to derive the aerosol optical centroid height (AOCH). , Appendix A contains more details about the AOCH. The EPIC AOCH has been validated against CALIOP observed extinction weighted aerosol height and found to have a high level of accuracy with a correlation coefficient of 0.885 and RMSE of 0.91 km during the 2020 California wildfires. We regrid hourly EPIC plume height data into a rectilinear grid of 10 km × 10 km resolution and take the daily mean for comparison with daily GFAS data. We use EPIC observations for the period 2016–2022, except for the year 2019 when the spacecraft was placed in extended safe hold due to degradation of gyroscope.

GFAS Emission Inventory

We represent wildfire plume heights with GFAS, an operational biomass burning inventory with parametrized plume heights based on numerical one-dimensional plume rise models (PRMs) available at 10 km grid resolution and daily temporal resolution. The inputs to the 1-D plume rise model (Navier–Stokes equations) include convective heat flux, active fire area, and atmospheric profiles such as temperature and wind profiles. The plume rise model defines plume height as the height where most of the smoke detrainment happens after attainment of hydrostatic equilibrium. The plume height in the PRM model is estimated for fires with an area exceeding 1 ha and a MODIS subpixel fire effective temperature exceeding 600 K. We leverage the PRM-derived plume height information, which includes the mean altitude of maximum injection (MAMI), along with plume bottom information to construct a Gaussian profile (where μ = MAMI, 3σ = (MAMI-plume bottom height)). We use this Gaussian profile for a vertical distribution of mass flux for every species in the GFAS emission inventory. The detailed procedure is in Appendix B. We also enhance the particulate matter emission flux in the GFAS emission inventory by a factor of 3.4 as recommended by prior evaluations , that compare the modeled aerosol loading against the satellite observations. Neglect of this scale factor would substantially degrade simulation of AOD as discussed in the Results and Discussion section.

GEOS-Chem Simulation

We simulate the AOD to PM2.5 relationship using the chemical transport model GEOS-Chem in its High-Performance configuration (GCHP) version 13.4.0 with improvement to performance and usability. GCHP is driven by MERRA-2, a reanalysis meteorological field from the Goddard Earth Observing System from the NASA Global Modeling and Assimilation Office. The simulation is conducted at cubed sphere C180 resolution using a stretched grid simulation based on a C90 grid with a stretch factor of 2 (focused over North America, target latitude: 50° and target longitude: 265°) corresponding roughly to 50 km horizontal resolution. The spin-up time is one month.

The GEOS-Chem oxidant-aerosol simulation includes the sulfate-nitrate-ammonium system, , primary , and secondary carbonaceous aerosols, sea salt, and natural and anthropogenic dust. Aerosol thermodynamic partitioning between sulfate-nitrate-ammonium is computed with the ISORROPIA II thermodynamic module. Relative humidity-dependent aerosol size distributions and optical properties follow Latimer and Martin. Anthropogenic emissions are from the community Emissions Data System (CEDS). While the default fire emission inventory in GEOS-Chem utilizes the Global Fire Emission Data set version 4 (GFEDv4), which lacks emission height information, we address this limitation by replacing the GFED emission inventory with the Global Fire Assimilation System (GFAS) from the Copernicus Atmosphere Monitoring Service (CAMS) into our simulations.

AERONET Aerosol Optical Depth

We evaluate our plume height simulations with the Aerosol Robotic Network (AERONET), a worldwide sun photometer network that provides long-term measurement of AOD. We use the Version 3 Level 2 database, which includes an improved cloud screening algorithm. We sample AERONET AOD at satellite overpass time and interpolate to 550 nm wavelength, based on the local Ångström exponent at 440 and 670 nm.

Ground-Based Measurement of PM2.5

We evaluate our geophysical PM2.5 estimates with ground-based measurements of PM2.5 collected from the United States Environmental Protection Agency’s (EPA) Air Quality System (AQS). AQS includes federal, state, and tribal monitors across United States where PM2.5 is measured using Federal Reference Method and non-Federal Reference Methods (e.g., the IMPROVE network). PM2.5 data for Canada are sourced from the National Air Pollution Surveillance Program (NAPS). The PM2.5 data used here are 24 h averages and rastered on a regular grid of 1 km spatial resolution.

Methodology

Data Sampling

We employ the following procedure to infer PM2.5 from satellite AOD:

  • (1)
    Generation of geophysical data: We relate satellite AOD to surface PM2.5 using a geophysical approach following van Donkelaar et al.
    PM2.5,geophysical=(PM2.5AOD)GEOS‐Chem×AODsat 1
    Here, GEOS-Chem AOD is sampled at satellite overpass time and PM2.5 represents 24 h concentrations for the corresponding day.
  • (2)

    Coincident Sampling: As geophysical PM2.5 relies upon satellite AOD which has gaps during cloudy conditions, we sample the ground-based PM2.5 measurements for a given day only if AOD data exist for that day.

  • (3)

    Seasonal means are computed for both daily sampled geophysical PM2.5 and measured PM2.5. Observation sites with fewer than 10 days of coincident data are excluded when constructing seasonal averages.

Coincident sampling is similarly used to evaluate satellite AOD versus AERONET AOD, and to evaluate GFAS plume height versus satellite plume height. We sample simulated data (AOD and PM2.5) colocated with observations on a daily basis for the evaluation.

Results and Discussion

We first compare the GFAS plume heights against the EPIC plume heights for 2016–2022 (except the missing year of 2019). Figure shows the spatial distribution of GFAS plume heights and cosampled EPIC plume heights over North America along with the ratio between the two data sets. General overall consistency is found between the two data sets, with both exhibiting greater plume heights in the North and West than in Southeast. However, the GFAS plume height is significantly lower than the EPIC plume height over the regions dominated by evergreen needleleaf forest as classified by the MODIS MCD12C1 land cover product. The bias may reflect an underestimate in the MODIS fire radiative power (FRP) data for large fires which are used as the input for plume rise model, due to the absorption of longwave infrared radiation (4 μm used in the retrieval of FRP) by the entrained gases of the thick plumes of large fires of evergreen needleleaf forest , due to high sap content and dense habitat. We find that the average GFAS plume height is 1.9 km, and average EPIC plume height is 3.2 km, an underestimation by a scale factor of 1.7 across all years. This bias persists during both the extreme years of 2018 and 2020 (Figures S1 and S2) and during nonextreme fire years of 2016, 2017, 2021, and 2022 (Figures S3 and S4). In contrast, we find minimal bias over other land types (Figure S5). We explore the effect of doubling the GFAS plume height over the needleleaf region.

1.

1

Open in a new tab

Top: Spatial distribution of EPIC plume height (left) and coincident GFAS plume height (right) for the years 2016–2018 and 2020–2022 (EPIC data were not available in 2019). Bottom: Difference between GFAS and EPIC plume height (left) and spatial distribution of the evergreen needleleaf forest from the MODIS MCD12C1 product.

Figure shows a histogram of the GFAS plume height versus the EPIC plume height for 2016–2022 (excluding 2019) over North America. Both mean and mode bias decrease due to scaled GFAS plume height over the evergreen needleleaf forest. The mean bias reduces from 0.56 to −0.14 km, whereas the mode (maximum frequency) decreases from 0.8 to 0.05 km. Measurements of the extinction vertical profile from the CALIOP satellite measurement offer additional evidence of the need to increase the aerosol loading in the middle troposphere (Figure S6).

2.

2

Open in a new tab

Histogram of EPIC (H E) minus GFAS (H G) plume height difference sampled over North America for the years 2016–2018 and 2020–2022. H peak represents the height difference at the maximum frequency. Normalized frequency defines the fraction of points in each bin. The sum of normalized frequency over the horizontal axis is equal to unity.

We set up three simulations for 2018 and 2020 based on this information. The first simulation (GFED) uses the default emission inventory in GEOS-Chem which has no plume height information. The second simulation is performed using the GFAS emission inventory (GFAS-1X) with plume height information. The third simulation doubles the GFAS plume height (GFAS-2X) over the needleleaf forest region, while retaining the default plume height elsewhere.

Figure shows the AERONET and simulated AOD averaged over months with extreme fire events for the contiguous United States and Canada for the years 2018 (top row) and 2020 (bottom row). Both the measurements and simulations exhibit greater AOD over and downwind of fires in the Pacific Northwest. The simulation using GFED exhibits distinct and concentrated hotspots relative to the simulations with GFAS, highlighting the impact of releasing fire emissions directly in the surface layer. Incorporation of plume height broadens the fire hotspots due to increased smoke transport due to increased windspeed and lifetime aloft.

3.

3

Open in a new tab

Maps showing the calculated AOD from three different simulations for the years 2018 (top row) and 2020 (bottom row). Filled circles indicate in situ AOD from AERONET. The gray area in the maps represents the water bodies. Both simulated AOD and observed AOD from AERONET are averaged over the months with significant fire events (July to November inclusive for 2018, and August to December inclusive for 2020).

Scatter plots in Figure provide a more quantitative assessment. The simulation GFAS-2X best represents the AERONET measurements in 2018 (r 2 = 0.46 and slope = 0.83), and in 2020 (r 2 = 0.88, slope = 0.79), which is substantial improvement over the default simulation for both 2018 (r 2 = 0.35, slope = 0.41) and 2020 (r 2 = 0.32, slope = 0.20). The simulation with scaled plume height (GFAS-2X) outperforms GFAS-1X indicating that implementing a scaled plume height for the needleleaf forest region in the GEOS-Chem better represents the AOD measurements. Comparison of simulated AOD and satellite AOD from MAIAC (Figure S7) for three different simulations also supports the use of plume height information from GFAS albeit with less clarity about GFAS-1X versus GFAS-2X (Figure S8). We conduct a sensitivity simulation that neglects the scale factor of 3.4 recommended by prior evaluations , and find that doing so would substantially degrade the AOD simulation (Figures S9 and S10). We retained the scale factor for all subsequent analyses.

4.

4

Open in a new tab

Scatter plots depicting the comparison between calculated AOD from three different simulations and observed AOD from AERONET for the years 2018 (top row) and 2020 (bottom row). The solid line denotes y = x, and the dashed line represents the regression line.

Figure shows the PM2.5 measured and inferred from satellite AOD and the simulated AOD-PM2.5 relationship. The default approach of using GFED emissions at the surface results in an overestimation of satellite-derived PM2.5 concentrations in source regions and downwind with slopes of 2.60° in 2018 and 2.70° in 2020. The high bias in satellite-derived PM2.5 in the source regions reflects the high bias in simulated PM2.5 in the source regions (Figures S11 and S12) such that the simulated PM2.5-to-AOD ratio is biased to be high. The high bias of the satellite-derived PM2.5 downwind reflects insufficient smoke transport downwind such that the AOD is biased low and the simulated PM2.5-to-AOD ratio is biased high. The simulation using plume height addresses both issues, thus reducing the bias in satellite-derived PM2.5 both over the source region and downwind.

5.

5

Open in a new tab

PM2.5 concentrations inferred from satellite AOD from three different GEOS-Chem simulations for 2018 (top row) and 2020 (bottom row). The gray area in the maps represents the water bodies. The geophysical PM2.5 is averaged over the months with significant fire events (from July to November inclusive for 2018, and August to December inclusive for 2020). Figure S16 provides the corresponding measurement sites.

Figure shows scatter plots of satellite-derived PM2.5 versus ground-based measurements. The simulation GFAS-2X (Figure ) achieves the best geophysical PM2.5 performance for 2018 (slope = 1.39, r 2 = 0.62) and 2020 (slope = 0.98, r 2 = 0.67). Improvements reflect both decreased simulated PM2.5 values (Figures S11 and S12) and increased AOD downwind (Figure ). Prior work has similarly found that injecting smoke aloft can decrease PM2.5 in source regions, , increase AOD downwind, and decrease the AOD-to-PM2.5 ratio. ,

6.

6

Open in a new tab

Scatter plots depicting the comparison between geophysical PM2.5 from three different GEOS-Chem simulations and PM2.5 from ground-based measurements for the years 2018 (top row) and 2020 (bottom row). The solid line denotes y = x, while the dashed line represents the regression line.

Figure shows the simulated PM2.5-to-AOD ratio for the three simulations. The simulation using GFED without plume height information is highly heterogeneous with pronounced PM2.5AOD over wildfires. Adding (GFAS-1X) and increasing (GFAS-2X) plume height yield a more homogeneous horizontal distribution of the PM2.5-to-AOD ratio. This homogeneity implies that the resultant satellite-derived PM2.5 horizontal distribution of fire emissions will be less sensitive to the precise horizontal distribution of the fire emissions.

7.

7

Open in a new tab

PM2.5-to-AOD ratio from three different GEOS-Chem simulations for 2018 (top row) and 2020 (bottom row). The gray area in the maps represents the water bodies. The PM2.5-to-AOD ratio is averaged over the months with significant fire events (from July to November inclusive for 2018, and August to December inclusive for 2020).

We further isolate a wildfire event in western North America (latitude: 34°N-49°N; longitude: 126°W–115°W) that experienced large fire activity in 2020. For this case study, we analyze a 7-day period (September 11–17) to assess the effect of plume height on simulated surface concentrations. Figure S13 presents 7-day average maps of geophysical PM2.5 for 2020 (top row) and corresponding scatter plots comparing simulated and observed values (bottom row). Consistent with our earlier findings, the use of plume height in simulations significantly reduces biases in surface concentrations near the source regions.

Since Cheeseman et al. found that fire plumes leaving the boundary layer have little impact on surface PM2.5 concentrations, we next test the hypothesis that removing AOD pixels where satellite injection height exceeds boundary layer height would improve geophysical estimate of surface PM2.5. Figures S14 and S15 show the effect of the removal of the daily AOD pixels, where satellite injection height is greater than boundary layer height. Comparison with Figure indicates some improvement in the slope versus PM2.5 measurements, but the r 2 value also decreases implying a reduction in skill and the sampling of daily PM2.5 during such scenarios. A larger improvement was achieved by refining the AOD-PM2.5 relationship through a better representation of smoke plume height.

Outlook

In summary, incorporating smoke plume height into the chemical transport model GEOS-Chem enables a better representation of the surface PM2.5 concentration inferred from satellite AOD. This conclusion was reached by examining the impact of smoke plume height on the geophysical PM2.5 concentration inferred from satellite AOD, simulated AOD, and simulated PM2.5 during extreme fire events of 2018 and 2020 over North America. Evaluation of the modeled plume height from GFAS against satellite retrieval of plume height (EPIC) indicates that GFAS plume height is underestimated by a factor of 1.7 over the evergreen needleleaf forest during all six years examined (2016–2018, 2020–2022), and by a factor of 1.8 during 2018 and 2020. We find that EPIC satellite observations of plume height over North America are better represented by doubled rather than default GFAS plume height over evergreen needleleaf forest. This scaling may reflect the high fuel load of this biome that lofts the fire plumes to higher altitudes than the model predicts. Integrating the resulting scaled height into GEOS-Chem improves the consistency of both simulated AOD and satellite-derived PM2.5 with ground-based measurements, over that of a simulation using the default surface emissions from either GFED or GFAS. This improvement reflects the reduced surface PM2.5 concentration in source regions and enhanced AOD downwind, both reduce the overall PM2.5-to-AOD ratio.

Overall, our findings suggest that information about plume height is valuable for geophysical inference of PM2.5 from satellite AOD. Doubling the GFAS plume height over needleleaf forest regions best represents the collective observations of the EPIC plume height, AERONET AOD, and PM2.5 monitors over North America. Although plume height information from 1-D PRM is subject to uncertainties in FRP, observational constraints to the modeled plume height (GFAS) using satellite data improve accuracy for air quality applications. Further development of a more robust simulation of smoke plume height with more constraints from satellite retrievals has the potential to achieve significant advances in modeling AOD and PM2.5 affected by wildfires. Such advances will be further complemented by additional developments to the many processes affecting the AOD to PM2.5 relationship. ,,

Appendix A

Mean Altitude of Maximum Injection (MAMI)

The GFAS emission inventory has three plume height parameters: plume bottom height, plume top height, and the mean altitude of maximum injection (MAMI) above the ground. The GFAS used a 1D plume rise model to calculate these parameters. The governing equations include a 1D Navier–Stokes equation for momentum and mass conservation and the first law of thermodynamics. Rémy et al. present a schematic description of the plume rise model. The fire radiative power (FRP), fire area, temperature, and wind profiles are used as an input to calculate the detrainment profile, which is used to define three plume height parameters. The plume top height and plume bottom height represent the top and bottom heights of the detrainment profile, respectively. The MAMI is defined as the average of plume heights where detrainment exceeds half of the maximum value. This essentially describes the MAMI as the height at which bulk of plume mass accumulates, which leads to mass shedding to the atmosphere (detrainment). This shedding is minimal at the bottom due to strong buoyancy driving high mass flux in the vertical direction.

Aerosol Optical Centroid Height (AOCH)

In aerosol layer height retrieval algorithms, an aerosol extinction profile is assumed to have a quasi-Gaussian distribution with the following form:

τ(z)=Wexp(−σ(z−AOCH))[1+exp(−σ(z−AOCH))]2 2

where τ is the aerosol extinction coefficient at height z, W is a normalization constant derived from AOD and corresponds to the columnar loading, and σ represents the half-width. At z = AOCH, the aerosol extinction is largest and equals W. This indicates that AOCH is the height where aerosol extinction is maximum which corresponds to the maximum aerosol loading at that height given that all emitted aerosols are uniformly distributed within fire plume boundary.

The AOCH and MAMI both represent the height with maximum aerosol loading or mass and, hence, can be compared with each other.

Appendix B

Offline Implementation of GFAS Plume Height

We use a Gaussian profile to distribute the GFAS emission fluxes vertically following studies , in which the plume height retrieval algorithms assume a quasi-Gaussian distribution function with height with maximum aerosol extinction represents the mean and an assumed half-width of 1 km represents the extent of the plume. In this study, we use MAMI information to represent the mean and plume bottom information to calculate the extent of the plume. Following is the step-by-step procedure for the implementation of GFAS plume height in GEOS-Chem.

  • (1)

    We first create daily virtual atmosphere with 72 hybrid sigma levels using surface pressure from Merra-2 meteorology.

  • (2)

    Then, we create daily Gaussian vertical profiles using mean μ = mean altitude of maximum injection (MAMI) and standard deviation (σ) to define plume top and bottom at μ ± 3σ.

  • (3)

    We bin these Gaussian vertical profiles onto the Merra-2 grid to calculate the fractional contribution of each bin to the total column.

  • (4)

    Then we implement these 3-D emission field into GEOS-Chem through The Harmonised Emission Component (HEMCO).

Supplementary Material

ea5c00035_si_001.pdf (1.3MB, pdf)

Acknowledgments

This work was supported by NASA Grant 80NSSC21K0508 and NSF Grant 2244984.

The python codes used for data analysis and implementation of plume height into GEOS-Chem are available at https://github.com/inder-math/Project2/. The GFAS and MERRA-2 data sets used in this study can be downloaded from http://geoschemdata.wustl.edu/ExtData/.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsestair.5c00035.

  • Summary of the monthly burned area per fire in the United States and Canada, distinguishing extreme and nonextreme fire event months (Tables S1 and S2); spatial distributions and histograms comparing GFAS and EPIC plume heights across years and land types (Figures S1–S5); vertical extinction profiles, satellite AOD maps, and sensitivity tests of GEOS-Chem AOD against AERONET observations (Figures S6–S10); simulated PM2.5 maps and their comparison with ground-based measurements during peak fire periods (Figures S11–S13); satellite-derived PM2.5 obtained by excluding AOD pixels where EPIC plume height is above boundary layer height against observations (Figures S14 and S15); and ground-based PM2.5 measurement sites for 2018 and 2020 (Figure S16). (PDF)

The authors declare no competing financial interest.

References

  1. Brauer M., Roth G. A., Aravkin A. Y., Zheng P., Abate K. H., Abate Y. H., Abbafati C., Abbasgholizadeh R., Abbasi M. A., Abbasian M.. et al. Global Burden and Strength of Evidence for 88 Risk Factors in 204 Countries and 811 Subnational Locations, 1990–2021: A Systematic Analysis for the Global Burden of Disease Study 2021. Lancet. 2024;403(10440):2162–2203. doi: 10.1016/S0140-6736(24)00933-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Burke M., Driscoll A., Heft-Neal S., Xue J., Burney J., Wara M.. The Changing Risk and Burden of Wildfire in the United States. Proc. Natl. Acad. Sci. U.S.A. 2021;118(2):e2011048118. doi: 10.1073/pnas.2011048118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Korsiak J., Pinault L., Christidis T., Burnett R. T., Abrahamowicz M., Weichenthal S.. Long-Term Exposure to Wildfires and Cancer Incidence in Canada: A Population-Based Observational Cohort Study. Lancet Planet. Health. 2022;6(5):e400–e409. doi: 10.1016/S2542-5196(22)00067-5. [DOI] [PubMed] [Google Scholar]
  4. Zhou X., Josey K., Kamareddine L., Caine M. C., Liu T., Mickley L. J., Cooper M., Dominici F.. Excess of COVID-19 Cases and Deaths Due to Fine Particulate Matter Exposure during the 2020 Wildfires in the United States. Sci. Adv. 2021;7(33):eabi8789. doi: 10.1126/sciadv.abi8789. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Pinder R. W., Klopp J. M., Kleiman G., Hagler G. S. W., Awe Y., Terry S.. Opportunities and Challenges for Filling the Air Quality Data Gap in Low- and Middle-Income Countries. Atmos. Environ. 2019;215:116794. doi: 10.1016/j.atmosenv.2019.06.032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Hoff R. M., Christopher S. A.. Remote Sensing of Particulate Pollution from Space: Have We Reached the Promised Land? J. Air Waste Manage. Assoc. 2009;59(6):645–675. doi: 10.3155/1047-3289.59.6.645. [DOI] [PubMed] [Google Scholar]
  7. Zhao T., Mao J., Gupta P., Zhang H., Wang J.. Observational Constraints on the Aerosol Optical Depth–Surface PM2.5 Relationship during Alaskan Wildfire Seasons. ACS ES&T Air. 2024;1(9):1164–1176. doi: 10.1021/acsestair.4c00120. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Andreae M. O., Rosenfeld D., Artaxo P., Costa A. A., Frank G. P., Longo K. M., Silva-Dias M. A. F.. Smoking Rain Clouds over the Amazon. Science. 2004;303(5662):1337–1342. doi: 10.1126/science.1092779. [DOI] [PubMed] [Google Scholar]
  9. Kahn R. A., Chen Y., Nelson D. L., Leung F.-Y., Li Q., Diner D. J., Logan J. A.. Wildfire Smoke Injection Heights: Two Perspectives from Space. Geophys. Res. Lett. 2008;35(4):e1. doi: 10.1029/2007GL032165. [DOI] [Google Scholar]
  10. Zhu H., Martin R. V., van Donkelaar A., Hammer M. S., Li C., Meng J., Oxford C. R., Liu X., Li Y., Zhang D.. et al. Importance of Aerosol Composition and Aerosol Vertical Profiles in Global Spatial Variation in the Relationship between PM2.5 and Aerosol Optical Depth. Atmos. Chem. Phys. 2024;24(20):11565–11584. doi: 10.5194/acp-24-11565-2024. [DOI] [Google Scholar]
  11. Fromm M., Bevilacqua R., Servranckx R., Rosen J., Thayer J. P., Herman J., Larko D.. Pyro-Cumulonimbus Injection of Smoke to the Stratosphere: Observations and Impact of a Super Blowup in Northwestern Canada on 3–4 August 1998. J. Geophys. Res. 2005;110(D8):307. doi: 10.1029/2004JD005350. [DOI] [Google Scholar]
  12. Feng X., Mickley L. J., Bell M. L., Liu T., Fisher J. A., Val Martin M.. Improved Estimates of Smoke Exposure during Australia Fire Seasons: Importance of Quantifying Plume Injection Heights. Atmos. Chem. Phys. 2024;24(5):2985–3007. doi: 10.5194/acp-24-2985-2024. [DOI] [Google Scholar]
  13. Rémy S., Veira A., Paugam R., Sofiev M., Kaiser J. W., Marenco F., Burton S. P., Benedetti A., Engelen R. J., Ferrare R., Hair J. W.. Two Global Data Sets of Daily Fire Emission Injection Heights since 2003. Atmos. Chem. Phys. 2017;17(4):2921–2942. doi: 10.5194/acp-17-2921-2017. [DOI] [Google Scholar]
  14. Sofiev M., Ermakova T., Vankevich R.. Evaluation of the Smoke-Injection Height from Wild-Land Fires Using Remote-Sensing Data. Atmos. Chem. Phys. 2012;12(4):1995–2006. doi: 10.5194/acp-12-1995-2012. [DOI] [Google Scholar]
  15. Nelson D. L., Garay M. J., Kahn R. A., Dunst B. A.. Stereoscopic Height and Wind Retrievals for Aerosol Plumes with the MISR INteractive eXplorer (MINX) Remote Sens. 2013;5(9):4593–4628. doi: 10.3390/rs5094593. [DOI] [Google Scholar]
  16. Lyapustin A., Wang Y., Korkin S., Kahn R., Winker D.. MAIAC Thermal Technique for Smoke Injection Height From MODIS. IEEE Geosci. Remote Sens. Lett. 2020;17(5):730–734. doi: 10.1109/LGRS.2019.2936332. [DOI] [Google Scholar]
  17. Nanda S., de Graaf M., Veefkind J. P., Sneep M., ter Linden M., Sun J., Levelt P. F.. A First Comparison of TROPOMI Aerosol Layer Height (ALH) to CALIOP Data. Atmos. Meas. Tech. 2020;13(6):3043–3059. doi: 10.5194/amt-13-3043-2020. [DOI] [Google Scholar]
  18. Winker D. M., Vaughan M. A., Omar A., Hu Y., Powell K. A., Liu Z., Hunt W. H., Young S. A.. Overview of the CALIPSO Mission and CALIOP Data Processing Algorithms. J. Atmos. Ocean. Technol. 2009;26(11):2310–2323. doi: 10.1175/2009JTECHA1281.1. [DOI] [Google Scholar]
  19. Marshak A., Lyapustin A., Schuster G. L., Szabo A., Eckman R.. Editorial: DSCOVR EPIC/NISTAR: 5 Years of Observing Earth From the First Lagrangian Point. Front. Remote Sens. 2022;3:963660. doi: 10.3389/frsen.2022.963660. [DOI] [Google Scholar]
  20. Lyapustin A., Go S., Korkin S., Wang Y., Torres O., Jethva H., Marshak A.. Retrievals of Aerosol Optical Depth and Spectral Absorption From DSCOVR EPIC. Front. Remote Sens. 2021;2:645794. doi: 10.3389/frsen.2021.645794. [DOI] [Google Scholar]
  21. Xu X., Wang J., Wang Y., Zeng J., Torres O., Reid J. S., Miller S. D., Martins J. V., Remer L. A.. Detecting Layer Height of Smoke Aerosols over Vegetated Land and Water Surfaces via Oxygen Absorption Bands: Hourly Results from EPIC/DSCOVR in Deep Space. Atmos. Meas. Tech. 2019;12(6):3269–3288. doi: 10.5194/amt-12-3269-2019. [DOI] [Google Scholar]
  22. Kaulfus A. S., Nair U., Jaffe D., Christopher S. A., Goodrick S.. Biomass Burning Smoke Climatology of the United States: Implications for Particulate Matter Air Quality. Environ. Sci. Technol. 2017;51(20):11731–11741. doi: 10.1021/acs.est.7b03292. [DOI] [PubMed] [Google Scholar]
  23. Kollanus V., Prank M., Gens A., Soares J., Vira J., Kukkonen J., Sofiev M., Salonen R. O., Lanki T.. Mortality Due to Vegetation Fire-Originated PM2.5 Exposure in Europe-Assessment for the Years 2005 and 2008. Environ. Health Perspect. 2017;125(1):30–37. doi: 10.1289/EHP194. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Liu J. C., Mickley L. J., Sulprizio M. P., Dominici F., Yue X., Ebisu K., Anderson G. B., Khan R. F. A., Bravo M. A., Bell M. L.. Particulate Air Pollution from Wildfires in the Western US under Climate Change. Clim Change. 2016;138(3):655–666. doi: 10.1007/s10584-016-1762-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Xu R., Ye T., Yue X., Yang Z., Yu W., Zhang Y., Bell M. L., Morawska L., Yu P., Zhang Y.. et al. Global Population Exposure to Landscape Fire Air Pollution from 2000 to 2019. Nature. 2023;621(7979):521–529. doi: 10.1038/s41586-023-06398-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Cheeseman M., Ford B., Volckens J., Lyapustin A., Pierce J. R.. The Relationship Between MAIAC Smoke Plume Heights and Surface PM. Geophys. Res. Lett. 2020;47(17):e2020GL088949. doi: 10.1029/2020GL088949. [DOI] [Google Scholar]
  27. Lyapustin A., Wang Y., Korkin S., Huang D.. MODIS Collection 6 MAIAC Algorithm. Atmos. Meas. Tech. 2018;11(10):5741–5765. doi: 10.5194/amt-11-5741-2018. [DOI] [Google Scholar]
  28. Lu Z., Wang J., Xu X., Chen X., Kondragunta S., Torres O., Wilcox E. M., Zeng J.. Hourly Mapping of the Layer Height of Thick Smoke Plumes Over the Western U.S. in 2020 Severe Fire Season. Front. Remote Sens. 2021;2:766628. doi: 10.3389/frsen.2021.766628. [DOI] [Google Scholar]
  29. Freitas S. R., Longo K. M., Chatfield R., Latham D., Silva Dias M. a. F., Andreae M. O., Prins E., Santos J. C., Gielow R., Carvalho J. A. J.. Including the Sub-Grid Scale Plume Rise of Vegetation Fires in Low Resolution Atmospheric Transport Models. Atmos. Chem. Phys. 2007;7(13):3385–3398. doi: 10.5194/acp-7-3385-2007. [DOI] [Google Scholar]
  30. Kaiser J. W., Heil A., Andreae M. O., Benedetti A., Chubarova N., Jones L., Morcrette J.-J., Razinger M., Schultz M. G., Suttie M., van der Werf G. R.. Biomass Burning Emissions Estimated with a Global Fire Assimilation System Based on Observed Fire Radiative Power. Biogeosciences. 2012;9(1):527–554. doi: 10.5194/bg-9-527-2012. [DOI] [Google Scholar]
  31. Pan X., Ichoku C., Chin M., Bian H., Darmenov A., Colarco P., Ellison L., Kucsera T., da Silva A., Wang J.. et al. Six Global Biomass Burning Emission Datasets: Intercomparison and Application in One Global Aerosol Model. Atmos. Chem. Phys. 2020;20(2):969–994. doi: 10.5194/acp-20-969-2020. [DOI] [Google Scholar]
  32. Eastham S. D., Long M. S., Keller C. A., Lundgren E., Yantosca R. M., Zhuang J., Li C., Lee C. J., Yannetti M., Auer B. M.. et al. GEOS-Chem High Performance (GCHP V11–02c): A next-Generation Implementation of the GEOS-Chem Chemical Transport Model for Massively Parallel Applications. Geosci. Model Dev. 2018;11(7):2941–2953. doi: 10.5194/gmd-11-2941-2018. [DOI] [Google Scholar]
  33. Martin R. V., Eastham S. D., Bindle L., Lundgren E. W., Clune T. L., Keller C. A., Downs W., Zhang D., Lucchesi R. A., Sulprizio M. P.. et al. Improved Advection, Resolution, Performance, and Community Access in the New Generation (Version 13) of the High-Performance GEOS-Chem Global Atmospheric Chemistry Model (GCHP) Geosci. Model Dev. 2022;15(23):8731–8748. doi: 10.5194/gmd-15-8731-2022. [DOI] [Google Scholar]
  34. Bindle L., Martin R. V., Cooper M. J., Lundgren E. W., Eastham S. D., Auer B. M., Clune T. L., Weng H., Lin J., Murray L. T.. et al. Grid-Stretching Capability for the GEOS-Chem 13.0.0 Atmospheric Chemistry Model. Geosci. Model Dev. 2021;14(10):5977–5997. doi: 10.5194/gmd-14-5977-2021. [DOI] [Google Scholar]
  35. Park R. J., Jacob D. J., Field B. D., Yantosca R. M., Chin M.. Natural and Transboundary Pollution Influences on Sulfate-Nitrate-Ammonium Aerosols in the United States: Implications for Policy. J. Geophys. Res. 2004;109(D15):e30. doi: 10.1029/2003JD004473. [DOI] [Google Scholar]
  36. Pye H. O. T., Liao H., Wu S., Mickley L. J., Jacob D. J., Henze D. K., Seinfeld J. H.. Effect of Changes in Climate and Emissions on Future Sulfate-Nitrate-Ammonium Aerosol Levels in the United States. J. Geophys. Res. 2009;114(D1):e77. doi: 10.1029/2008JD010701. [DOI] [Google Scholar]
  37. Heald C. L., Coe H., Jimenez J. L., Weber R. J., Bahreini R., Middlebrook A. M., Russell L. M., Jolleys M., Fu T.-M., Allan J. D.. et al. Exploring the Vertical Profile of Atmospheric Organic Aerosol: Comparing 17 Aircraft Field Campaigns with a Global Model. Atmos. Chem. Phys. 2011;11(24):12673–12696. doi: 10.5194/acp-11-12673-2011. [DOI] [Google Scholar]
  38. Park R. J., Jacob D. J., Chin M., Martin R. V.. Sources of Carbonaceous Aerosols over the United States and Implications for Natural Visibility. J. Geophys. Res. 2003;108(D12):e33. doi: 10.1029/2002JD003190. [DOI] [Google Scholar]
  39. Pai S. J., Heald C. L., Pierce J. R., Farina S. C., Marais E. A., Jimenez J. L., Campuzano-Jost P., Nault B. A., Middlebrook A. M., Coe H.. et al. An Evaluation of Global Organic Aerosol Schemes Using Airborne Observations. Atmos. Chem. Phys. 2020;20(5):2637–2665. doi: 10.5194/acp-20-2637-2020. [DOI] [Google Scholar]
  40. Jaeglé L., Quinn P. K., Bates T. S., Alexander B., Lin J.-T.. Global Distribution of Sea Salt Aerosols: New Constraints from in Situ and Remote Sensing Observations. Atmos. Chem. Phys. 2011;11(7):3137–3157. doi: 10.5194/acp-11-3137-2011. [DOI] [Google Scholar]
  41. Fairlie T. D., Jacob D. J., Dibb J. E., Alexander B., Avery M. A., van Donkelaar A., Zhang L.. Impact of Mineral Dust on Nitrate, Sulfate, and Ozone in Transpacific Asian Pollution Plumes. Atmos. Chem. Phys. 2010;10(8):3999–4012. doi: 10.5194/acp-10-3999-2010. [DOI] [Google Scholar]
  42. Philip S., Martin R. V., van Donkelaar A., Lo J. W.-H., Wang Y., Chen D., Zhang L., Kasibhatla P. S., Wang S., Zhang Q.. et al. Global Chemical Composition of Ambient Fine Particulate Matter for Exposure Assessment. Environ. Sci. Technol. 2014;48(22):13060–13068. doi: 10.1021/es502965b. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Fountoukis C., Nenes A.. ISORROPIA II: A Computationally Efficient Thermodynamic Equilibrium Model for K+–Ca2+–Mg2+–NH4 +–Na+–SO4 2––NO3 ––Cl––H2O Aerosols. Atmos. Chem. Phys. 2007;7(17):4639–4659. doi: 10.5194/acp-7-4639-2007. [DOI] [Google Scholar]
  44. Latimer R. N. C., Martin R. V.. Interpretation of Measured Aerosol Mass Scattering Efficiency over North America Using a Chemical Transport Model. Atmos. Chem. Phys. 2019;19(4):2635–2653. doi: 10.5194/acp-19-2635-2019. [DOI] [Google Scholar]
  45. Hoesly R. M., Smith S. J., Feng L., Klimont Z., Janssens-Maenhout G., Pitkanen T., Seibert J. J., Vu L., Andres R. J., Bolt R. M.. et al. Historical (1750–2014) Anthropogenic Emissions of Reactive Gases and Aerosols from the Community Emissions Data System (CEDS) Geosci. Model Dev. 2018;11(1):369–408. doi: 10.5194/gmd-11-369-2018. [DOI] [Google Scholar]
  46. van der Werf G. R., Randerson J. T., Giglio L., van Leeuwen T. T., Chen Y., Rogers B. M., Mu M., van Marle M. J. E., Morton D. C., Collatz G. J.. et al. Global Fire Emissions Estimates during 1997–2016. Earth Syst. Sci. Data. 2017;9(2):697–720. doi: 10.5194/essd-9-697-2017. [DOI] [Google Scholar]
  47. Giles D. M., Sinyuk A., Sorokin M. G., Schafer J. S., Smirnov A., Slutsker I., Eck T. F., Holben B. N., Lewis J. R., Campbell J. R.. et al. Advancements in the Aerosol Robotic Network (AERONET) Version 3 Database – Automated near-Real-Time Quality Control Algorithm with Improved Cloud Screening for Sun Photometer Aerosol Optical Depth (AOD) Measurements. Atmos. Meas. Tech. 2019;12(1):169–209. doi: 10.5194/amt-12-169-2019. [DOI] [Google Scholar]
  48. van Donkelaar A., Martin R. V., Park R. J.. Estimating Ground-Level PM2.5 Using Aerosol Optical Depth Determined from Satellite Remote Sensing. J. Geophys. Res. 2006;111(D21):e97. doi: 10.1029/2005JD006996. [DOI] [Google Scholar]
  49. Schroeder W., Ellicott E., Ichoku C., Ellison L., Dickinson M. B., Ottmar R. D., Clements C., Hall D., Ambrosia V., Kremens R.. Integrated Active Fire Retrievals and Biomass Burning Emissions Using Complementary Near-Coincident Ground, Airborne and Spaceborne Sensor Data. Remote Sens. Environ. 2014;140:719–730. doi: 10.1016/j.rse.2013.10.010. [DOI] [Google Scholar]
  50. Doran, J. D. ; Randall, C. K. ; Long, A. J. . Fire in the Wildland-Urban Interface: Selecting and Maintaining Firewise Plants for Landscaping University of Florida: Gainesville, FL; 2004. [Google Scholar]
  51. Li Y., Tong D., Ma S., Freitas S. R., Ahmadov R., Sofiev M., Zhang X., Kondragunta S., Kahn R., Tang Y.. et al. Impacts of Estimated Plume Rise on PM2.5 Exceedance Prediction during Extreme Wildfire Events: A Comparison of Three Schemes (Briggs, Freitas, and Sofiev) Atmos. Chem. Phys. 2023;23(5):3083–3101. doi: 10.5194/acp-23-3083-2023. [DOI] [Google Scholar]
  52. Jin X., Fiore A. M., Curci G., Lyapustin A., Civerolo K., Ku M., van Donkelaar A., Martin R. V.. Assessing Uncertainties of a Geophysical Approach to Estimate Surface Fine Particulate Matter Distributions from Satellite-Observed Aerosol Optical Depth. Atmos. Chem. Phys. 2019;19(1):295–313. doi: 10.5194/acp-19-295-2019. [DOI] [Google Scholar]
  53. Chu D. A., Ferrare R., Szykman J., Lewis J., Scarino A., Hains J., Burton S., Chen G., Tsai T., Hostetler C.. et al. Regional Characteristics of the Relationship between Columnar AOD and Surface PM2.5: Application of Lidar Aerosol Extinction Profiles over Baltimore–Washington Corridor during DISCOVER-AQ. Atmos. Environ. 2015;101:338–349. doi: 10.1016/j.atmosenv.2014.11.034. [DOI] [Google Scholar]
  54. Chen X., Wang J., Xu X., Zhou M., Zhang H., Castro Garcia L., Colarco P. R., Janz S. J., Yorks J., McGill M.. et al. First Retrieval of Absorbing Aerosol Height over Dark Target Using TROPOMI Oxygen B Band: Algorithm Development and Application for Surface Particulate Matter Estimates. Remote Sens. Environ. 2021;265:112674. doi: 10.1016/j.rse.2021.112674. [DOI] [Google Scholar]
  55. Kim H., Chen X., Wang J., Lu Z., Zhou M., Carmichael G. R., Park S. S., Kim J.. Aerosol Layer Height (ALH) Retrievals from Oxygen Absorption Bands: Intercomparison and Validation among Different Satellite Platforms, GEMS, EPIC, and TROPOMI. Atmos. Meas. Tech. 2025;18(2):327–349. doi: 10.5194/amt-18-327-2025. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ea5c00035_si_001.pdf (1.3MB, pdf)

Data Availability Statement

The python codes used for data analysis and implementation of plume height into GEOS-Chem are available at https://github.com/inder-math/Project2/. The GFAS and MERRA-2 data sets used in this study can be downloaded from http://geoschemdata.wustl.edu/ExtData/.

Articles from ACS Es&t Air are provided here courtesy of American Chemical Society

RESOURCES