PyroStack: A Multi-Band Spatio-Temporal Sub-Daily Dataset for Wildfires in the United States
Authors: Arya Kondur, Giosue Migliorini, Cameron Schmitt, Francesco Immorlano, Tairan Wang, Rebecca C. Scholten, Efi Foufoula-Georgiou, Gary Johnson, +9 more
Organizations: Department of Computer Science, University of California, Irvine, CA 92697, USA · Department of Statistics, University of California, Irvine, CA 92697, USA · Department of Civil and Environmental Engineering, University of California, Irvine, CA 92697, USA · Learning the Earth with Artificial Intelligence and Physics (LEAP) Center, Columbia University, New York, NY 10027, USA · Department of Earth System Science, University of California, Irvine, CA 92697, USA · Spatial Informatics Group, Pleasanton, CA 94566, USA · CloudFire Inc., Auburn, CA 95603, USA · Institute for Biocomputation and Physics of Complex Systems (BIFI), University of Zaragoza, Spain · Department of Theoretical Physics, University of Zaragoza, Spain
Wildfires are an increasing hazard to ecosystems, air quality, and human systems, creating a growing need for datasets that support systematic development and evaluation of models for predicting fire spread across diverse landscapes. Effective prediction requires integrating meteorological conditions, fuels, vegetation, and topography at spatial and temporal resolutions suitable for both physical simulation and data-driven approaches. However, existing datasets often lack the resolution and coverage needed to capture these interacting controls. The PyroStack dataset addresses this gap by providing a harmonized, event-based collection of wildfire and environmental data across the contiguous United States and Alaska. It integrates satellite-derived fire observations with atmospheric reanalysis, vegetation, fuel characteristics, and topographic information into a unified framework spanning 6994 wildfires that occurred between 2012 and 2024 across a wide range of ecosystems and climate conditions. PyroStack offers spatial resolutions ranging from 30 m to 9 km and hourly temporal resolution, along with fire progression data at 12-hour intervals to support model initialization and evaluation. By combining broad spatial coverage with fine spatial and temporal detail, the dataset enables systematic analysis of wildfire dynamics and supports both physics-based and machine learning approaches, providing a foundation for benchmarking and improving fire spread models, with future extensions aimed at incorporating additional regions and fire suppression data streams to further advance wildfire prediction.
Figures & tables
Name
Reference
Region
Temporal res.
Spatial res.
Num. of fire trajectories 1
Num. of vars
Variable types 2
Fire Data Source 3
Years
CalWildFire
De Rango et al. (2026)
Calabria (Italy)
Daily
100 m
8,609
21
M, T, FFA, FM, FS, A
F
2008–2018
Canadian Fire Spread Dataset
Barber et al. (2024)
Canada
Daily
180 m
3,296
50
M, V, T, PFA, FM, FS, A
M, V, L, S
2002–2021
Wildfire- SpreadTS+
Lahrichi et al. (2026)
W-Central US
Daily
375 m– 27 km
1,005
40
M, V, T, FPX, FM
V, M, L
2016–2023
TS-SatFire
Zhao et al. (2025)
CONUS
1 h–1 yr
375 m
179
27
M, V, T, FPX
V, M
2017–2021
FireSpread_MedEU 4
Müller et al. (2026)
Mediterranean & EU
Daily
3 m
103
21
V, PFA, A
P
2017–2023
IberFire 5
Erzibengoa et al. (2025)
Spain
Daily
1–9 km
–
120
M, V, T, FPX, FS, FM, A
S
2008–2024
Table 1: Comparison of existing fire tracking datasets with the PyroStack dataset introduced in this study.
Layer Category
Data Source
Spatial Resolution
Temporal Resolution
Directory Name
Fire Characteristics
FEDS-MTBS ( Chen et al., 2026 )
300 m
12-hr
fire_spread
Low Resolution Climate
ERA5-Land ( Muñoz-Sabater et al., 2021 )
9000 m
1-hr
low_res_climate
High Resolution Climate
RTMA via CloudFire Inc. Worldgen Server
600 m
1-hr
high_res_climate
Fuel Structure
LANDFIRE via CloudFire Inc. Worldgen Server
30 m
Static
fuel_structure
Vegetation, Fuel Model, and Topography
LANDFIRE ( Ryan and Opperman, 2013 )
30 m
Static
veg_fm_topo
Table 2: Summary of dataset layer categories and data sources. Spatial and temporal resolution correspond to processed data in PyroStack, rather than the original data source.
Figure 1: Burn map for progression of the Caldor fire. The extent of the plot indicates the bounding box as determined by the final fire area ( farea ). Fire progression (with perimeters from FEDS-MTBS) fire areas are shown for (a) three individual timesteps corresponding to days 2, 10, and 40, with active firelines ( fline ) shown in red at each timestep, and (b) 12-hour intervals through the entire fire duration.
Ignition Date Range
CONUS Source
AK Source
Start Date
End Date
LF Version
Code
LF
CF
LF
CF
1 Oct 2024
31 Dec 2024
LF 2024
2.5.0
API
API
API
Archive
1 Oct 2023
30 Sep 2024
LF 2023
2.4.0
API
API
API
Archive
1 Jan 2023
30 Sep 2023
LF 2022
2.3.0
API
API
API
Archive
1 Jan 2021
31 Dec 2022
LF 2020
2.2.0
Archive
API
Archive
Archive
1 Jan 2017
31 Dec 2020
LF 2016
2.0.0
API
API
API
Archive
Table 3: Summary of LANDFIRE (LF) version selection across temporal ranges with their respective sources. These versions and sources were used for the retrieval of time-varying fuel layers (e.g., evt, fbfm13, fbfm40, cc, ch, cbd, cbh) and not for time-invariant layers provided by LANDFIRE (e.g., aspect, elevation, slope). In the source columns, LF and CF refer to fuel layers retrieved through the LANDFIRE API and CloudFire Inc. Worldgen server, respectively.
Layer Name 1
Description
Category
Data Source
Units
farea
Fire area
Fire Characteristics
FEDS
Binary
fline
Active fireline
Fire Characteristics
FEDS
Binary
nfp
New fire pixels
Fire Characteristics
FEDS
Binary
frp
Fire radiative power
Fire Characteristics
FEDS
W/m 2
d2m
2-meter dewpoint temperature
Low-Res Climate
ERA5-Land
Kelvin (K)
sp
Surface pressure
Low-Res Climate
ERA5-Land
Pascal (Pa)
Table 4: Brief description of the data layers available for each fire in PyroStack.
Fire Size Category
Fire Count
Total Active Fire Obs.
Total Fire Line Length
Total Burned Area
Category
Threshold
Count
%
Length (km)
%
Area ( 103 km 2 )
%
Top 1%
>600 km2
70
4,098
7
102
21
71
27
Top 5%
>160 km2
350
13,600
24
235
48
150
58
Top 10%
>70 km2
700
22,551
41
314
64
187
72
Top 50%
>7 km2
3497
47,900
86
459
93
246
95
All fires
–
6994
55,680
100
495
100
259
100
Table 5: Cumulative contribution of the largest fires to the total number of 12-hourly active fire observations, total fireline length, and total burned area. Data are grouped by fire size categories, where each category represents the top percentile of fires sorted by burned area.
Figure 2: Map of all fires included in the PyroStack dataset. Point size scales with the final burned area of each fire, while point color denotes fire type (wildfire, prescribed, or unknown).
Figure 3: Characteristics of wildfires and prescribed fires in PyroStack. Distributions are shown by (a) total burned area, (b) burn duration, defined as the elapsed time between the first and last FEDS time step, (c) ignition month, (d) initial elevation, (e) dominant vegetation type based on the LANDFIRE existing vegetation physiognomy product, and (f) year of occurrence. In panel (f), total annual burned area is also shown for comparison.
Figure 4: Time series of selected variables from PyroStack for the Caldor Fire, which burned in 2021 in California’s Sierra Nevada. Panels show (a) cumulative burned area, (b) incremental burned area, and (c) active fireline length, together with selected hourly environmental drivers aggregated over the fire domain: (d) vapor pressure deficit, derived from the PyroStack temperature and dewpoint temperature layers; (e) wind speed; and (f) precipitation. The red-shaded intervals on 15–16 August 2021 indicate the three 12-hour periods used to initialize and evaluate the machine-learning- and physics-based fire-spread simulations described in Sections 3.3 and 3.4 (see Figure 6 ).
Figure 5: Spatial distribution of PyroStack covariate layers (i.e. non-fire-spread layers) for the Caldor Fire. Data correspond to observations at 12PM PST on August 15, 2021.
Figure 6: Comparison of Pyretechnics ensemble and our machine learning model on fire spread forecasts for the Caldor Fire. Each panel shows predicted burn probability overlaid on terrain shading, with the initial cold perimeter (dark blue), initial active fireline (cyan), and observed final FEDS-MTBS perimeter (black) shown for reference. The cold perimeter distinguishes the inactive segments of the initial FEDS-MTBS perimeter from the active fireline at time t , while the final FEDS-MTBS perimeter represents the updated observed fire extent at time t+12 . Panels (a, d) show forecasts initialized at 14:00 PDT on 15 August 2021, panels (b, e) show forecasts initialized at 02:00 PDT on 16 August 2021, and panels (c, f) show forecasts initialized at 14:00 PDT on 16 August 2021. The Pyretechnics ensemble (a, b, c) produces spatially concentrated, high-confidence predictions closely tied to the active fireline, reflecting its physics-based propagation rules. The ViT (d, e, f) produces smoother, more spatially diffuse probability fields, reflecting the model’s tendency to spread probability mass over a broader region.
Faculty of Science, University of Alberta · Interdisciplinary Lab for Mathematical Ecology & Epidemiology (ILMEE), University of Alberta · Department of Mathematical and Statistical Sciences, University of Alberta
Department of Electrical and Computer Engineering, University of Illinois Chicago, Chicago, IL, USA · Informatics Institute, Istanbul Technical University, ˙Istanbul, Türkiye · USDA Forest Service Pacific Wildland Fire Sciences Laboratory, Washington, USA +1