Get data on forest fires, compiled annually for the National Forestry Database The National Forestry Database includes national forest data and forest management statistics to seve as a credible, accurate and reliable source of information on forest management and its impact on the forest resource. Forest fire data is grouped into eight categories, which are further broken down by geographic location. These include: * number of fires by cause class and response category * area burned by cause class and response category * number of fires by month and response category * area burned by month and response category * number of fires by fire size class and response category * area burned by fire size class and response category * area burned by productivity class, stocking class, maturity class and response category * other fire statistics, such as property losses

As of November 2024, Turkey reported nearly 120,000 hectares lost to forest fires that year, more than four times the figure recorded one year earlier. During the period in consideration, Turkey saw the largest wildfire-affected area in 2021, with fires having burnt more than 200,000 hectares across the country.

In 2023, a total of 1,059 forest fires were recorded in Germany. This was an decrease compared to 2,397 forest fires in 2022. Figures fluctuated during the specified time period, though they peaked significantly in certain years, namely 1992 and 2003.

The FinalFirePerimeter polygon layer represents final mapped wildland fire perimeters. This feature class is a subset of the FirePerimeters feature class. Incidents of 10 acres or greater in size are expected. Incidents smaller than 10 acres in size may also be included. Data are maintained at the Forest/District level, or their equivalent, to track the area affected by wildland fire. Records in FirePerimeter include perimeters for wildland fires that have corresponding records in FIRESTAT, which is the authoritative data source for all wildland fire reports. FIRESTAT, the Fire Statistics System computer application, required by the USFS for all wildland fire occurrences on National Forest System Lands or National Forest-protected lands, is used to enter and maintain information from the Individual Fire Report (FS-5100-29).National USFS fire occurrence final fire perimeters where wildland fires have historically occurred on National Forest System Lands and/or where protection is the responsibility of the US Forest Service. Knowing where wildland fire events have happened in the past is critical to land management efforts in the future.This data is utilized by fire & aviation staffs, land managers, land planners, and resource specialists on and around National Forest System Lands.*This data has been updated to match 2021 National GIS Data Dictionary Standards.Metadata and Downloads

We have curated a dataset to address the problem of forest fire detection. All images in the dataset are 3-channeled with spatial resolution of 250 × 250. We have retrieved these images by searching various search terms in multiple search engines. Afterwards, we thoroughly investigated these images to crop and remove the inappropriate components such as people, fire-extinguishing machinery etc. We ensured that each image only contain the relevant fire region. The dataset is designed for binary problem of fire or no-fire detection in the forests landscape. It is a balanced dataset consisting of 1900 images in total, where 950 images belong to each class. We have divided the dataset into 80:20 for training and testing purposes in our study.

The forest fire map shows forest fires that occurred mainly in the territory of southern Quebec, i.e. the area located south of the territorial limit of attributable forests. This map data makes it possible to improve knowledge about fire regimes and to meet the specific needs of special management plans following forest fires. They can also be used to meet a variety of study and research needs, such as analyzing the impact of climate change, modeling post-fire regeneration, and studying ecosystem dynamics. This information is obtained from and produced from a variety of sources, including satellite images, aerial photographs, field or aerial surveys, fire scar dating, and archival documents. This data contains four types of mapping as well as fire regime mapping: • Detailed fire mapping, from 1976 to the present. This mapping includes burn types, total burn and partial burn, when information is available. In addition, for fires that have been characterized, information on the classes of burning patterns is added. The minimum mapping area can be up to 0.1 ha, depending on the source products used. This map is partially available for areas located in the north of southern Quebec. • Mapping the simplified contours of fires, from 1972 to today. This map shows the external contours of fires (without fragmentation), in order to represent them globally in a product that is easily usable and can be integrated into current information systems, GPS or others. Resulting from the fusion of detailed fire mapping, this product was designed to meet various customer needs. This map is partially available for the sectors located in the north of southern Quebec. • The mapping of the origin of fires having been listed by the protection organizations (e.g.: SOPFEU) for the period from 1972 to today. This mapping includes the date, the source of ignition (human or lightning) and the protection zone. It is available for all of Quebec. • The mapping of ancient fires concerns fires that occurred between the very end of the 19th century and 1975. This mapping comes from the information present on the forest maps of the first and second inventories, as well as from the information contained on the ecoforest maps of the third and fourth inventories. The dating of these fires is done using various methods, including the analysis of study trees bearing fire scars and the consultation of archival documents. This data is available for the following regions: Saguenay-Lac-Saint-Jean (02), Bas-Saint-Laurent (02), Bas-Saint-Laurent (01), Gaspésie-Îles-de-la-Madeleine (11), Abitibi-Témiscamingue (08), Mauricie-Centre-du-Québec (04-17), and Lanaudière-du-Québec (04-17), and Lanaudière-Laurentides (14-15). • Mapping fire regimes in southern Quebec. This map shows 13 zones with distinct fire regimes. These areas were delineated based on available information on the areas burned during the period 1890-2020 and other potentially decisive environmental variables, such as physiography, the abundance of different tree species known to be dependent on fire as well as the location of natural and anthropogenic ignitions. Fire regime mapping covers all forest areas under management as well as a more northern portion that is not managed. The detailed methodology is presented in Forest Research Paper no. 189 “Zoning fire regimes in southern Quebec” (coming soon). This zoning may be useful to ensure better consideration of the risk of fire in a forest management context. It can also serve as a territorial basis for projecting future fire activity taking into account various factors, such as climate change, fire suppression as well as changes in the types of fuels and their distribution on the territory.**This third party metadata element was translated using an automated translation tool (Amazon Translate).**

This is a regression task, where the aim is to predict the burned area of forest fires, in the northeast region of Portugal, by using meteorological and other data.

Data Set Information:

In [Cortez and Morais, 2007], the output 'area' was first transformed with a ln(x+1) function. Then, several Data Mining methods were applied. After fitting the models, the outputs were post-processed with the inverse of the ln(x+1) transform. Four different input setups were used. The experiments were conducted using a 10-fold (cross-validation) x 30 runs. Two regression metrics were measured: MAD and RMSE. A Gaussian support vector machine (SVM) fed with only 4 direct weather conditions (temp, RH, wind and rain) obtained the best MAD value: 12.71 +- 0.01 (mean and confidence interval within 95% using a t-student distribution). The best RMSE was attained by the naive mean predictor. An analysis to the regression error curve (REC) shows that the SVM model predicts more examples within a lower admitted error. In effect, the SVM model predicts better small fires, which are the majority.

Attribute Information:

For more information, read [Cortez and Morais, 2007].

1. X - x-axis spatial coordinate within the Montesinho park map: 1 to 9
2. Y - y-axis spatial coordinate within the Montesinho park map: 2 to 9
3. month - month of the year: 'jan' to 'dec'
4. day - day of the week: 'mon' to 'sun'
5. FFMC - FFMC index from the FWI system: 18.7 to 96.20
6. DMC - DMC index from the FWI system: 1.1 to 291.3
7. DC - DC index from the FWI system: 7.9 to 860.6
8. ISI - ISI index from the FWI system: 0.0 to 56.10
9. temp - temperature in Celsius degrees: 2.2 to 33.30
10. RH - relative humidity in %: 15.0 to 100
11. wind - wind speed in km/h: 0.40 to 9.40
12. rain - outside rain in mm/m2 : 0.0 to 6.4
13. area - the burned area of the forest (in ha): 0.00 to 1090.84 (this output variable is very skewed towards 0.0, thus it may make sense to model with the logarithm transform).

To use this dataset:

import tensorflow_datasets as tfds

ds = tfds.load('forest_fires', split='train')
for ex in ds.take(4):
 print(ex)

See the guide for more informations on tensorflow_datasets.

Forest fire incidents are becoming increasingly common around the world, posing a threat to the environment, economy, and social life. These wildfires are further expected to rise in their frequency and intensity, considering the global climate change and human activities. A variety of attributes must be studied in order to analyse relationships between the probable causes of fire and the characteristics of wildfire incidents, and inform decision-making. Such attributes are available or easily collectable in various regions around the world, but they are not readily available in the South American Amazon. The Amazon rainforest covers such a large area that acquiring a useful dataset necessitates extensive effort and computer intensive pre-processing. The associated study to this dataset investigates potential data sources for the Amazon, establishes a methodological baseline, and prepares a dataset of covariates thought to be contributing to the wildfire ignition process. The dataset is intended to be used for forest fire studies, specifically spatio-temporal and statistical analysis of wildfires. The study provides three sets of (i) raw data (acquired data with a global extent), (ii) pre-processed data (source data transformed to the same projection system and same file format), and (iii) working data (cropped to Amazon region extent with spatial resolution of 500 meters and monthly temporal resolution, to enable the scientific community to work with various possibilities of forest-fire analysis, and to further encourage research in study areas in the other parts of the world.

First, we would like to thank the wildland fire advisory group. Their wisdom and guidance helped us build the dataset as it currently exists. Currently, there are multiple, freely available fire datasets that identify wildfire and prescribed fire burned areas across the United States. However, these datasets are all limited in some way. Their time periods could cover only a couple of decades or they may have stopped collecting data many years ago. Their spatial footprints may be limited to a specific geographic area or agency. Their attribute data may be limited to nothing more than a polygon and a year. None of the existing datasets provides a comprehensive picture of fires that have burned throughout the last few centuries. Our dataset uses these existing layers and utilizes a series of both manual processes and ArcGIS Python (arcpy) scripts to merge these existing datasets into a single dataset that encompasses the known wildfires and prescribed fires within the United States a ...

The "Forest Fire Dataset" is a comprehensive and meticulously curated resource, specifically designed to support the development of algorithms for forest fire detection and object detection tasks. The dataset consists of 2,974 images dedicated to fire classification, which are divided into two primary categories: the first category includes images documenting active forest fires, while the second category contains images of intact, fire-free forest environments. This clear distinction within the dataset is crucial for training models to accurately differentiate between fire-affected and unaffected areas in forested regions. In addition to the fire classification data, the dataset includes 1,690 images dedicated to object detection, enhancing its applicability in machine learning and computer vision research.

The dataset is carefully structured with a thoughtful distribution across training, validation, and test sets, with proportions of 80%, 15%, and 5%, respectively, to ensure that models trained on this data can generalize effectively to new, unseen data. The data were collected from various online sources and underwent rigorous manual filtering to maintain high data integrity. Additionally, a portion of the dataset was generated through controlled simulations of forest fires, conducted after obtaining the necessary approvals from relevant authorities. This simulated portion adds diversity and reliability to the dataset, providing a more comprehensive training ground for algorithms.

By integrating both real-world and simulated scenarios, the "Forest Fire Dataset" offers a robust foundation for developing advanced fire detection systems, significantly contributing to forest conservation and disaster management efforts.

For scientific research and advanced applications in the fields of forest fire detection and computer vision, the "Forest Fire Dataset" is a valuable tool. Researchers and practitioners are encouraged to refer to the published article that details the development of the system based on this dataset. To cite the article related to this dataset, the following citation can be used:

I. Shamta and B. E. Demir, “Development of a deep learning-based surveillance system for forest fire detection and monitoring using UAV,” ِ Artica: PLoS One, vol. 19, no. 3, p. e0299058, 2024.

Link to Article: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0299058 ORCID: https://orcid.org/my-orcid?orcid=0009-0003-1280-679X Google Academik: https://scholar.google.com/citations?user=xP6CvtQAAAAJ&hl=tr

INVESTIGATIVE WILDFIRE DATA FOR TURKEY/ NASA

BE CAREFUL OF THE FILE NAMES.

IT CONTAINS THE DATA NEEDED TO RESEARCH LATEST FOREST FIRES IN TURKEY.

PAY ATTENTION TO THE DATE INTERVALS. THESE ARE 7-11 DAILY DATA OF LAST TIMES.

• fire _ nrt _ M _ C61 _ 212465 _ all _ countries.csv

This file is important for all countries becuase it contains fire data of last 11 days for all around the world

Content

Data on recent forest fires in Turkey, published with permission from NASA Portal. The data was created based on the hotspots and obtained from the satellite.

3 SEPARATE SATELLITE DATA:

• MODIS C6.1
• SUOMI VIIRS C2
• J1 VIIRS C1

GENERAL ATTRIBUTES

• Latitude Center of nominal 375 m fire pixel

• Longitude Center of nominal 375 m fire pixel

• Bright_ti4 (Brightness temperature I-4) VIIRS I-4: channel brightness temperature of the fire pixel measured in Kelvin.

• Scan (Along Scan pixel size) The algorithm produces approximately 375 m pixels at nadir. Scan and track reflect actual pixel size.

• Track (Along Track pixel size) The algorithm produces approximately 375 m pixels at nadir. Scan and track reflect actual pixel size.

• Acq_Date (Acquisition Date) Date of VIIRS acquisition.

• Acq_Time (Acquisition Time) Time of acquisition/overpass of the satellite (in UTC).

• Satellite N Suomi National Polar-orbiting Partnership (Suomi NPP)

• Confidence This value is based on a collection of intermediate algorithm quantities used in the detection process. It is intended to help users gauge the quality of individual hotspot/fire pixels. Confidence values are set to low, nominal and high. Low confidence daytime fire pixels are typically associated with areas of sun glint and lower relative temperature anomaly (15K) temperature anomaly in either day or nighttime data. High confidence fire pixels are associated with day or nighttime saturated pixels.

Please note:

• Low confidence nighttime pixels occur only over the geographic area extending from 11° E to 110° W and 7° N to 55° S. This area describes the region of influence of the South Atlantic Magnetic Anomaly which can cause spurious brightness temperatures in the mid-infrared channel I4 leading to potential false positive alarms. These have been removed from the NRT data distributed by FIRMS.

• Version Version identifies the collection (e.g. VIIRS Collection 1) and source of data processing: Near Real-Time (NRT suffix added to collection) or Standard Processing (collection only).

"1.0NRT" - Collection 1 NRT processing.

"1.0" - Collection 1 Standard processing.

• Bright_ti5 (Brightness temperature I-5) I-5 Channel brightness temperature of the fire pixel measured in Kelvin.

• FRP (Fire Radiative Power) FRP depicts the pixel-integrated fire radiative power in MW (megawatts). Given the unique spatial and spectral resolution of the data, the VIIRS 375 m fire detection algorithm was customized and tuned in order to optimize its response over small fires while balancing the occurrence of false alarms. Frequent saturation of the mid-infrared I4 channel (3.55-3.93 µm) driving the detection of active fires requires additional tests and procedures to avoid pixel classification errors. As a result, sub-pixel fire characterization (e.g., fire radiative power [FRP] retrieval) is only viable across small and/or low-intensity fires. Systematic FRP retrievals are based on a hybrid approach combining 375 and 750 m data. In fact, starting in 2015 the algorithm incorporated additional VIIRS channel M13 (3.973-4.128 µm) 750 m data in both aggregated and unaggregated format.

Satellite measurements of fire radiative power (FRP) are increasingly used to estimate the contribution of biomass burning to local and global carbon budgets. Without an associated uncertainty, however, FRP-based biomass burning estimates cannot be confidently compared across space and time, or against estimates derived from alternative methodologies. Differences in the per-pixel FRP measured near-simultaneously in consecutive MODIS scans are approximately normally distributed with a standard deviation (ση) of 26.6%. Simulations demonstrate that this uncertainty decreases to less than ~5% (at ±1 ση) for aggregations larger than ~50 MODIS active fire pixels. Although FRP uncertainties limit the confidence in flux estimates on a per-pixel basis, the sensitivity of biomass burning estimates to FRP uncertainties can be mitigated by conducting inventories at coarser spatiotemporal resolutions.

http://cedadocs.ceda.ac.uk/770/1/SEVIRI_FRP_documentdesc.pdf

• Type (Inferred hot spot type) 0 = presumed vegetation fire

1 = active volcano

2 = other static land source

3 = offshore detection (includes all detections over water)

• DayNight (Day or Night)

D= Daytime fire

N= Nighttime fire

The FireOccurrence point layer represents ignition points, or points of origin, from which individual USFS wildland fires started. Data are maintained at the Forest/District level, or their equivalent, to track the occurrence and the origin of individual USFS wildland fires. Forests are working to include historical data, which may be incomplete.National USFS fire occurrence locations where wildland fires have historically occurred on National Forest System Lands and/or where protection is the responsibility of the US Forest Service. Knowing where wildland fire events have happened in the past is critical to land management efforts in the future.This data is utilized by fire & aviation staffs, land managers, land planners, and resource specialists on and around National Forest System Lands. The attributes included within the FireOccurrence point layer are needed to meet the needs of the US Forest Service, for data exchange between interagency data systems, to relate to the FirePerimeter polygon data layer and various fire data systems, and to track the locations of wildland fires.*This data has been updated to match 2021 National GIS Data Dictionary Standards.Metadata and Downloads

Oregon saw the largest area burned by wildfires across the United States in 2024. That year, about 2,232 individual wildfires burned in the northwestern state, ravishing almost 1.89 million acres. Texas followed second, with roughly 1.3 million acres burned due to wildfires that year. Fire season 2021 and California’s wildfire suppression costs As one of the most wildfire-prone states in the country, California spends a significant amount of money on their suppression. Estimates suggest wildfire suppression expenditure in California climbed to 1.2 billion U.S. dollars in the fiscal year ending June 2022. The fiscal year, which includes the summer and fall months of 2021, was among the most devastating fire seasons on record, with that year’s Dixie fire becoming the second-largest California wildfire by acres burned. The Dixie fire was responsible for over 963,000 acres burned across the state that year. Wildfire causes Wildfires are uncontrolled fires burning across any type of combustible vegetation such as grass- and brushland, forests, and agricultural fields. They are also referred to as wildland fires, forest fires, or bushfires, with the latter term particularly common in Australia. Wildfires regularly occur on all continents of the world, except for Antarctica, but are particularly common in dry regions with dense vegetation. As the rise in average global temperatures is changing weather patterns and resulting in more and more countries being affected by dry, hot weather conditions, the severity and rapid spread of wildfires have increased in recent years. The most common causes of wildfires are natural phenomena such as lightning strikes as well as human activity. The area burned due to human-caused wildfires in the U.S. surpassed 1.5 million acres in 2023.

The FirePerimeter polygon layer represents daily and final mapped wildland fire perimeters. Incidents of 10 acres or greater in size are expected. Incidents smaller than 10 acres in size may also be included. Data are maintained at the Forest/District level, or their equivalent, to track the area affected by wildland fire. Records in FirePerimeter include perimeters for wildland fires that have corresponding records in FIRESTAT, which is the authoritative data source for all wildland fire reports. FIRESTAT, the Fire Statistics System computer application, required by the USFS for all wildland fire occurrences on National Forest System Lands or National Forest-protected lands, is used to enter and maintain information from the Individual Fire Report (FS-5100-29).National USFS fire occurrence final fire perimeters where wildland fires have historically occurred on National Forest System Lands and/or where protection is the responsibility of the US Forest Service. Knowing where wildland fire events have happened in the past is critical to land management efforts in the future.This data is utilized by fire & aviation staffs, land managers, land planners, and resource specialists on and around National Forest System Lands.*This data has been updated to match 2021 National GIS Data Dictionary Standards.Metadata and DownloadsThis record was taken from the USDA Enterprise Data Inventory that feeds into the https://data.gov catalog. Data for this record includes the following resources: ISO-19139 metadata ArcGIS Hub Dataset ArcGIS GeoService CSV Shapefile GeoJSON KML For complete information, please visit https://data.gov.

The Final Risk Charter represents the possibility of a fire with serious consequences for the socio-economic and environmental reality of a given area. It is derived from the sum of the maps of “Probability”, “Intensity” and “Vulnerability”. Each area has been classified from 1 low risk to 4 high risk.

The first Algerian forest fires dataset consist of data on forest fires occurrence in Algeria related to meteorological observations and the fire weather indices. Our dataset includes mainly the daily meteorological observations and the Fire Weather Index (FWI) system components. Given the lack of publically available datasets on data on forest fires occurrence in Algeria we have created this one to study the feasibility of the appliance of machine learning algorithms as models for forest fires prediction in the context of Algeria.

This data set provides high-resolution surface reflectance, thermal imagery, burn severity metrics, and LiDAR-derived structural measures of forested areas in the Sierra Nevada Mountains, California, USA, collected before and after the August 2013 Rim and September 2014 King mega forest fires. Pre-fire data were paired with post-fire collections to assess pre- and post-fire landscape characteristics and fire severity. Field estimates of fire severity were collected to compare with derived remote sensing indices. Reflectance measurements for the spectroscopic AVIRIS and MASTER sensors are distributed as multi-band geotiffs for each megafire and acquisition date. Derived operational metric products for each sensor are provided in individual GeoTIFFs. GeoTIFFs produced from LiDAR point data depict first order topographic indices and summary statistics of vertical vegetation structure.

The California Department of Forestry and Fire Protection's Fire and Resource Assessment Program (FRAP) annually maintains and distributes an historical wildland fire perimeter dataset from across public and private lands in California. The GIS data is developed with the cooperation of the United States Forest Service Region 5, the Bureau of Land Management, California State Parks, National Park Service and the United States Fish and Wildlife Service and is released in the spring with added data from the previous calendar year. Although the dataset represents the most complete digital record of fire perimeters in California, it is still incomplete, and users should be cautious when drawing conclusions based on the data.

This data should be used carefully for statistical analysis and reporting due to missing perimeters (see Use Limitation in metadata). Some fires are missing because historical records were lost or damaged, were too small for the minimum cutoffs, had inadequate documentation or have not yet been incorporated into the database. Other errors with the fire perimeter database include duplicate fires and over-generalization. Additionally, over-generalization, particularly with large old fires, may show unburned "islands" within the final perimeter as burned. Users of the fire perimeter database must exercise caution in application of the data. Careful use of the fire perimeter database will prevent users from drawing inaccurate or erroneous conclusions from the data. This data is updated annually in the spring with fire perimeters from the previous fire season. This dataset may differ in California compared to that available from the National Interagency Fire Center (NIFC) due to different requirements between the two datasets. The data covers fires back to 1878. As of May 2025, it represents fire24_1.

Please help improve this dataset by filling out this survey with feedback:

Historic Fire Perimeter Dataset Feedback (arcgis.com)

Current criteria for data collection are as follows:

CAL FIRE (including contract counties) submit perimeters ≥10 acres in timber, ≥50 acres in brush, or ≥300 acres in grass, and/or ≥3 impacted residential or commercial structures, and/or caused ≥1 fatality.

All cooperating agencies submit perimeters ≥10 acres.

Version update:

Firep24_1 was released in April 2025. Five hundred forty-eight fires from the 2024 fire season were added to the database (2 from BIA, 56 from BLM, 197 from CAL FIRE, 193 from Contract Counties, 27 from LRA, 8 from NPS, 55 from USFS and 8 from USFW). Six perimeters were added from the 2025 fire season (as a special case due to an unusual January fire siege). Five duplicate fires were removed, and the 2023 Sage was replaced with a more accurate perimeter. There were 900 perimeters that received updated attribution (705 removed “FIRE” from the end of Fire Name field and 148 replaced Complex IRWIN ID with Complex local incident number for COMPLEX_ID field). The following fires were identified as meeting our collection criteria but are not included in this version and will hopefully be added in a future update: Addie (2024-CACND-002119), Alpaugh (2024-CACND-001715), South (2024-CATIA-001375). One perimeter is missing containment date that will be updated in the next release.

Cross checking CALFIRS reporting for new CAL FIRE submissions to ensure accuracy with cause class was added to the compilation process. The cause class domain description for “Powerline” was updated to “Electrical Power” to be more inclusive of cause reports.

Includes separate layers filtered by criteria as follows:

California Fire Perimeters (All): Unfiltered. The entire collection of wildfire perimeters in the database. It is scale dependent and starts displaying at the country level scale.

Recent Large Fire Perimeters (≥5000 acres): Filtered for wildfires greater or equal to 5,000 acres for the last 5 years of fires (2020-January 2025), symbolized with color by year and is scale dependent and starts displaying at the country level scale. Year-only labels for recent large fires.

California Fire Perimeters (1950+): Filtered for wildfires that started in 1950-January 2025. Symbolized by decade, and display starting at country level scale.

Detailed metadata is included in the following documents:

Wildland Fire Perimeters (Firep24_1) Metadata

For any questions, please contact the data steward:

Kim Wallin, GIS Specialist

CAL FIRE, Fire & Resource Assessment Program (FRAP)

kimberly.wallin@fire.ca.gov

The California Department of Forestry and Fire Protection's Fire and Resource Assessment Program (FRAP) annually maintains and distributes an historical wildland fire perimeter dataset from across public and private lands in California. The GIS data is developed with the cooperation of the United States Forest Service Region 5, the Bureau of Land Management, California State Parks, National Park Service and the United States Fish and Wildlife Service and is released in the spring with added data from the previous calendar year. Although the dataset represents the most complete digital record of fire perimeters in California, it is still incomplete, and users should be cautious when drawing conclusions based on the data.

This data should be used carefully for statistical analysis and reporting due to missing perimeters (see Use Limitation in metadata). Some fires are missing because historical records were lost or damaged, were too small for the minimum cutoffs, had inadequate documentation or have not yet been incorporated into the database. Other errors with the fire perimeter database include duplicate fires and over-generalization. Additionally, over-generalization, particularly with large old fires, may show unburned "islands" within the final perimeter as burned. Users of the fire perimeter database must exercise caution in application of the data. Careful use of the fire perimeter database will prevent users from drawing inaccurate or erroneous conclusions from the data. This data is updated annually in the spring with fire perimeters from the previous fire season. This dataset may differ in California compared to that available from the National Interagency Fire Center (NIFC) due to different requirements between the two datasets. The data covers fires back to 1878. As of May 2025, it represents fire24_1.

Please help improve this dataset by filling out this survey with feedback:

Historic Fire Perimeter Dataset Feedback (arcgis.com)

Current criteria for data collection are as follows:

CAL FIRE (including contract counties) submit perimeters ≥10 acres in timber, ≥50 acres in brush, or ≥300 acres in grass, and/or ≥3 impacted residential or commercial structures, and/or caused ≥1 fatality.

All cooperating agencies submit perimeters ≥10 acres.

Version update:

Firep24_1 was released in April 2025. Five hundred forty-eight fires from the 2024 fire season were added to the database (2 from BIA, 56 from BLM, 197 from CAL FIRE, 193 from Contract Counties, 27 from LRA, 8 from NPS, 55 from USFS and 8 from USFW). Six perimeters were added from the 2025 fire season (as a special case due to an unusual January fire siege). Five duplicate fires were removed, and the 2023 Sage was replaced with a more accurate perimeter. There were 900 perimeters that received updated attribution (705 removed “FIRE” from the end of Fire Name field and 148 replaced Complex IRWIN ID with Complex local incident number for COMPLEX_ID field). The following fires were identified as meeting our collection criteria but are not included in this version and will hopefully be added in a future update: Addie (2024-CACND-002119), Alpaugh (2024-CACND-001715), South (2024-CATIA-001375). One perimeter is missing containment date that will be updated in the next release.

Cross checking CALFIRS reporting for new CAL FIRE submissions to ensure accuracy with cause class was added to the compilation process. The cause class domain description for “Powerline” was updated to “Electrical Power” to be more inclusive of cause reports.

Includes separate layers filtered by criteria as follows:

California Fire Perimeters (All): Unfiltered. The entire collection of wildfire perimeters in the database. It is scale dependent and starts displaying at the country level scale.

Recent Large Fire Perimeters (≥5000 acres): Filtered for wildfires greater or equal to 5,000 acres for the last 5 years of fires (2020-January 2025), symbolized with color by year and is scale dependent and starts displaying at the country level scale. Year-only labels for recent large fires.

California Fire Perimeters (1950+): Filtered for wildfires that started in 1950-January 2025. Symbolized by decade, and display starting at country level scale.

Detailed metadata is included in the following documents:

Wildland Fire Perimeters (Firep24_1) Metadata

For any questions, please contact the data steward:

Kim Wallin, GIS Specialist

CAL FIRE, Fire & Resource Assessment Program (FRAP)

kimberly.wallin@fire.ca.gov