Satellite sensors and federal fire records tell a consistent story: fire-friendly conditions are lasting longer in many regions, and decades of U.S. ignition records show the dominant spark is often not lightning. A major peer-reviewed analysis of U.S. wildfire starts from 1992 through 2012 found that people accounted for about 84% of recorded ignitions, substantially extending the wildfire season beyond the window defined by natural lightning. Layered on top of climate-driven increases in fuel aridity documented in the research literature, that combination can narrow the margin of safety for communities near wildland areas and strain firefighting capacity during peak periods.
What is verified so far
The most detailed national accounting of U.S. wildfire starts comes from a federal database covering 1992 through 2015. That dataset, compiled from federal, state, and local reporting systems, contains approximately 1.88 million georeferenced wildfire records spanning roughly 140 million acres. Each record includes the fire’s discovery date, cause classification, and location, making it possible to compare human-caused and lightning-caused ignitions on a national scale. The database is publicly accessible through a Forest Service research archive, which documents the methods used to harmonize and quality-check the underlying reports.
A peer-reviewed analysis published in the Proceedings of the National Academy of Sciences used a subset of that national record, covering 1992 through 2012, and found that human ignitions account for approximately 84% of recorded U.S. wildfires. The authors drew on those 1.5 million-plus records to map where and when fires started, distinguishing between lightning and human causes such as debris burning, equipment use, and power line failures. Their work, available through a public repository, showed that human-caused fires substantially extend the wildfire season beyond the window defined by natural lightning, and that those fires ignite in times and places with higher fuel moisture, conditions where lightning fires would be unlikely to spread. In practical terms, people are expanding the geographic and seasonal “fire niche” of the country.
The National Interagency Fire Center reinforces these findings with its own operational reporting. Using standardized National Wildfire Coordinating Group cause categories, NIFC compiles national 10-year averages showing that human activity remains the leading source of wildfire ignitions. Its guidance on fire investigation practices describes how causes are assigned and underscores that debris burning, equipment, and other human actions consistently dominate ignition statistics. The consistency between the PNAS analysis and NIFC’s operational summaries supports treating the 84% estimate as a useful baseline rather than an outlier, while recognizing that the underlying georeferenced national occurrence dataset described above ends in 2015.
On the climate side, a separate body of evidence tracks fire weather conditions rather than ignition counts. A study in Nature Communications, covering 1979 through 2013, documented a measurable increase in global mean fire weather season length across a substantial fraction of vegetated land. Using indices that combine temperature, humidity, wind, and precipitation, the authors found that more landscapes now experience an extended window of high fire danger. Their results, detailed in the published analysis, indicate that climate trends are creating more days each year when fires can start and spread rapidly.
In the western United States specifically, a separate attribution study in PNAS concluded that human-caused climate change has materially increased forest fire area by drying fuels and increasing the number of days with extreme fire weather. By comparing modeled worlds with and without anthropogenic greenhouse gas forcing, the authors linked a significant share of the observed growth in burned area to warming-driven changes in vapor pressure deficit and snowpack. The study, accessible through a medical and scientific index, supports the conclusion that climate change is not just coincident with larger fires but is a causal factor.
NASA’s observational infrastructure adds a near-real-time dimension to these longer-term records. The agency’s Fire Information for Resource Management System distributes active fire detections from MODIS sensors aboard Terra and Aqua satellites and VIIRS instruments on S-NPP, NOAA-20, and NOAA-21 platforms, with data typically available within 24 to 48 hours. Through the online FIRMS portal, users can visualize hotspots, download fire detection shapefiles, and monitor daily changes in fire activity. A companion product, known as MCD64A1, maps burn scars at 500-meter resolution, capturing both the extent and timing of burned areas; NASA’s Earthdata catalog describes the burned area product and its processing chain. Together, these tools allow researchers to track not just where fires burn but when burning begins and ends each year, feeding directly into season-length analyses and supporting independent checks on ground-based reports.
What remains uncertain
The strongest national wildfire occurrence database ends at 2015. No publicly available georeferenced dataset of comparable scope has been released for the years since, which means claims about post-2015 trends in human ignition rates rely on secondary reporting, state-level summaries, and operational tallies rather than standardized, peer-reviewed data. That gap matters because the period from 2017 through 2023 included several of the most destructive fire seasons on record in parts of the West, and without updated occurrence data, it is difficult to confirm whether the 84% human-ignition share has held steady, risen, or shifted regionally.
Direct statements from NASA scientists characterizing recent fire season intensity in institutional releases are also limited in the available evidence. The NASA Earth Observatory has published explanatory summaries linking longer and more frequent fire seasons to climate variables such as snowmelt timing and vapor pressure deficit, and noting ignition sources as a key factor (see NASA Earth Observatory). But those summaries draw primarily on the older peer-reviewed studies described above rather than offering new agency-level assessments of the most recent seasons. Readers should treat present-tense claims about “current” fire season severity with caution unless they cite a specific, dated source or a clearly documented data product.
Investigation records for individual major wildfires, including cause determinations under NWCG standards, are not publicly aggregated in a single searchable format. Local and state agencies may release findings for high-profile incidents, but there is no unified, real-time national database that compiles final cause classifications for every large fire. News coverage often attributes specific fires to human activity based on preliminary reports, but the full evidentiary chain from investigation to final classification can take months or years to complete. That lag means some widely circulated attributions remain provisional, and retrospective corrections may not receive the same attention as initial claims.
One hypothesis that current data cannot fully test involves the interaction between shifting human mobility patterns and peak fire danger windows. Urban sprawl, recreation trends, and the growth of wildland–urban interface communities have placed more ignition sources in fire-prone areas during the driest months. At the same time, telework and changing tourism patterns may be altering when and where people spend time outdoors. Whether that synchronization between human presence and high fire danger is measurably amplifying effective fire season severity in suburban–wildland zones is a question the existing databases were not designed to answer, and no peer-reviewed study in the available evidence has quantified it. Without finer-resolution data on human activity, fuel conditions, and small fire starts that never grow large enough to be widely reported, this remains an informed but untested idea.
How to read the evidence
The evidence base for this story rests on three distinct pillars, and they do not all measure the same thing. The first pillar is ignition records: the federal wildfire occurrence database and NIFC’s operational statistics. These data say that, over multiple decades, humans have been responsible for the majority of wildfire starts in the United States and that those starts occur in seasons and locations that lightning alone would not produce. They support confident statements about the historical dominance of human ignitions and their role in lengthening the practical fire season.
The second pillar is fire weather and climate attribution research. Studies using global fire weather indices and regional climate models show that warming temperatures and drying fuels are expanding the window of dangerous fire conditions and increasing burned area in western forests. These analyses do not track individual ignitions, but they clarify the background conditions that determine how likely a given spark is to turn into a large, fast-moving fire. They justify linking broader climate trends to changes in fire behavior, while also reminding readers that climate is one factor among several, including land management and suppression histories.
The third pillar is satellite-based monitoring. Systems like FIRMS and the MODIS burned area products offer consistent, spatially explicit observations of where and when fires burn, updated on timescales of days to weeks. They are powerful tools for mapping active fires, validating model-based estimates of season length, and checking whether reported incidents align with observed burn scars. However, they do not identify causes on their own and cannot yet replace on-the-ground investigation for attribution questions.
When these pillars are combined, a coherent picture emerges: human activity has dramatically increased the number and timing of wildfire ignitions, while human-driven climate change has made the landscape more receptive to burning. What remains uncertain are the precise trends since 2015, the detailed breakdown of causes for the most recent catastrophic fires, and the ways evolving human behavior may be interacting with peak fire weather. Until a new, comprehensive ignition database is released and more recent attribution studies are published, the most defensible statements are those anchored in the established record: that people start most U.S. wildfires, that climate change is lengthening and intensifying fire-friendly conditions, and that together these forces are reshaping the nation’s fire regimes in ways communities and policymakers must confront with clear-eyed attention to both what is known and what still needs to be measured.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
