the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
COMBUST: Gridded combustible mass estimates of the built environment in the conterminous United States (1975–2020)
Abstract. The increasing occurrence of natural hazards such as wildfires and drought, along with urban expansion and land consumption, causes increasing levels of fire risk to populations and human settlements. Moreover, increasing geopolitical instability in many regions of the world requires evaluation of scenarios related to potential hazards caused by military operations. Quantitative knowledge of burnable fuels and their spatiotemporal distribution across landscapes is crucial for risk and potential damage assessments. While there is a good understanding of the distributions of biomass fuels based on remote sensing observations, the combustible mass of the built environment has rarely been quantified in a spatially explicit and detailed manner. Therefore, we developed fine-grained estimates of urban fuels for the conterminous United States, estimating the combustible mass of building materials, building contents, and personal vehicles at 250 m spatial resolution. The resulting dataset is called COMBUST (COmbustible Mass of the Built environment in the conterminous United STates) and is currently the most comprehensive dataset of combustible mass and materials for the U.S., covering over 110 million structures. COMBUST includes different backcasting scenarios from 1975 to 2020 and is based on the integration of a variety of geospatial data sources such as Earth-observation derived data, real estate data, statistical estimates and volunteered geographic information. COMBUST is accompanied by COMBUST Plus, a set of consistently enumerated gridded datasets facilitating combustion exposure modelling of buildings and population. These datasets constitute a rich resource for ecological and social science applications, as well as for disaster risk management, planning and decision making. COMBUST is available at https://doi.org/10.5281/zenodo.15611963.
- Preprint
(20894 KB) - Metadata XML
- BibTeX
- EndNote
Status: open (until 24 Nov 2026)
- RC1: 'Comment on essd-2026-401', Anonymous Referee #1, 25 Aug 2026 reply
-
RC2: 'Comment on essd-2026-401', Anonymous Referee #2, 14 Nov 2026
reply
This manuscript introduces a 250 m gridded dataset of combustible mass in the built environment across the COUNS, i.e., COMBUST. It estimates combustible and non-combustible building materials, building contents, personal vehicles, fuel stored at gas stations and refineries, together with several supporting exposure variables. Overall, the content is interesting and the dataset is worthy of being further promoted. However, several issues regarding the models, validation, and presentation still need to be addressed. The specific comments are as follows:
1. The abstract contains descriptions related to geopolitical instability and military operation. I think this description doesn't fit in here at all. First, the continental United States did not experience any such impact during the period from 1975 to 2020. Second, there is no analysis on this aspect in the manuscript. I suggest removing these irrelevant descriptions.
2. The manuscript always links the COMBUST data directly to fire risk. However, combustible mass mainly represents potential fuel or exposure. Actual risk also depends on ignition probability, moisture, material accessibility, combustion efficiency, weather, and vulnerability. Please tone down the risk-related wording and make it clear that COMBUST is an input to risk assessment, rather than a risk product by itself.
3. Please check whether the information in Table 2 is correct. The overall mass balance is also unclear. For example, the total combustible mass values in Table 2 seems to not include all the components listed in the table and the values in mean and high scenarios are also exist issues. I suggest the authors make this calculation again and revise the corresponding description in Section 4.2.
4. Equation 1 includes gas stations and refineries but excludes personal vehicles, whereas some dataset descriptions include vehicles in the total. I’m curious whether personal vehicles fall under the category of building stock or built environment total? What are the differences between them?
5. The backcasting method is likely affected by strong bias. This is because buildings demolished before the contemporary baseline cannot be recovered, while newer replacement buildings may effectively be projected backward. It is a major limitation and deserves much more discussion or analysis.
6. The three backcasting models are described mostly in words. Please provide explicit equations showing how each historical layer is derived from the baseline.
7. Indoor area, building density, and building volume are three different structural assumptions, not repeated samples from a probability distribution. Therefore, the IQR between models should not be directly interpreted as the range of uncertainty.
8. The current validation statistics do not directly validate the historical combustible mass estimates. It is important to conduct independent validation using historical housing censuses, assessor records, or floor area statistics, even for a small number of cities or specific time periods.
9. The nominal 2020 baseline combines building data from around 2018, ZTRAX data from 2020, refinery data published in 2021, and gas-station locations from September 2024. Using 2024 facilities in a 2020 layer could introduce stations that did not yet exist. Please clarify the temporal reference of each component and ensure whether these data can represent the 2020 baseline.
10. Please add a validation in a few regions with different urban forms using assessor data, local building inventories, vehicle registrations, or facility capacities, and exhibit spatial bias rather than only national errors. The value of national errors holds no significance for users at the regional scale.
Citation: https://doi.org/10.5194/essd-2026-401-RC2
Data sets
COMBUST: Gridded combustible mass estimates of the built environment in the conterminous United States (1975-2020) J. H. Uhl et al. https://doi.org/10.5281/zenodo.15611963
Model code and software
COMBUST: Gridded combustible mass estimates of the built environment in the conterminous United States 1975-2020 - supplementary code J. H. Uhl and M. C. Cook https://github.com/johannesuhl/COMBUST
Viewed
| HTML | XML | Total | BibTeX | EndNote | |
|---|---|---|---|---|---|
| 284 | 138 | 120 | 542 | 134 | 123 |
- HTML: 284
- PDF: 138
- XML: 120
- Total: 542
- BibTeX: 134
- EndNote: 123
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1
See attached review.