An official website of the United States government
Research has shown that climate change creates warmer temperatures and drier conditions, leading to longer wildfire seasons and increased wildfire risks in the United States. These factors have, in turn, led to increases in the frequency, extent, and severity of wildfires in recent years. Given the danger posed by wildland fires to people, property, wildlife, and the environment, there is an urgent need to provide tools for effective wildfire management. Early detection of wildfires is essential to minimizing potentially catastrophic destruction. To that end, in this paper, we present our work on integrating multiple data sources into SmokeyNet, a deep learning model using spatiotemporal information to detect smoke from wildland fires. We present Multimodal SmokeyNet and SmokeyNet Ensemble for multimodal wildland fire smoke detection using satellite-based fire detections, weather sensor measurements, and optical camera images. An analysis is provided to compare these multimodal approaches to the baseline SmokeyNet in terms of accuracy metrics, as well as time-to-detect, which is important for the early detection of wildfires. Our results show that incorporating weather data in SmokeyNet improves performance numerically in terms of both F1 and time-to-detect over the baseline with a single data source. With a time-to-detect of only a few minutes, SmokeyNet can be used for automated early notification of wildfires, providing a useful tool in the fight against destructive wildfires.
more »
« less
This content will become publicly available on May 15, 2026
How a research community constructs and uses naturalness: A case study of the 2023 Lookout Fire and the H.J. Andrews Experimental Forest in Oregon, USA
Wildfire severity is increasing in the western United States. Simultaneously, many recognize that fire is a natural process and advocate for learning to live with fire. Indeed, the naturalness of fire can be an important reason provided to increase the amount of fire on a landscape. However, “naturalness” can be interpreted in incommensurate ways, such as the historic range of variability of a system or the absence of human influence. What makes wildfires feel natural or unnatural to the people who experience them, and how naturalness affects reactions to wildfires is underexplored. Using social representations theory, we examine the 2023 Lookout Fire at the H. J. Andrews Experimental Forest (HJA). We use semi-structured interviews (n = 40) to explore how the research community associated with the HJA mentally constructs and uses naturalness to emotionally process and make meaning from the wildfire. We find even in a community with advanced training in ecology, respondents use a variety of metrics to determine naturalness, including ignition source, fire behavior, and pre-fire landscape characteristics and fire history. Respondents consider a variety of factors, and there was not consensus on whether the Lookout Fire was a “natural” fire. In general, respondents who described the fire as more natural were able to come to a state of acceptance and excitement for future research opportunities sooner than respondents who described the fire as largely unnatural. This has important implications for wildfire risk communication for scientists and practitioners who want to restore fire as a natural process. While fires perceived (or framed) as natural may be more readily accepted, fires perceived as unnatural may take longer to process. Fires perceived as human-caused and especially as climate-exacerbated may be the most difficult for people to process after and during the fire, and may have the most resistance for being managed for purposes other than full suppression.
more »
« less
- Award ID(s):
- 2025755
- PAR ID:
- 10644578
- Publisher / Repository:
- Elsevier
- Date Published:
- Journal Name:
- Journal of environmental management
- Volume:
- 383
- ISSN:
- 0301-4797
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Abstract BackgroundPrescribed fires play a critical role in reducing the intensity and severity of future wildfires by systematically and widely consuming accumulated vegetation fuel. While the current probability of prescribed fire escape in the United States stands very low, their consequential impact, particularly the large wildfires they cause, raises substantial concerns. The most direct way of understanding this trade-off between wildfire risk reduction and prescribed fire escapes is to explore patterns in the historical prescribed fire records. This study investigates the spatiotemporal patterns of escaped prescribed fires in California from 1991 to 2020, offering insights for resource managers in developing effective forest management and fuel treatment strategies. ResultsThe results reveal that the months close to the beginning and end of the wildfire season, namely May, June, September, and November, have the highest frequency of escaped fires. Under similar environmental conditions, areas with more records of prescribed fire implementation tend to experience fewer escapes. The findings revealed the vegetation types most susceptible to escaped prescribed fires. Areas with tree cover ranging from 20 to 60% exhibited the highest incidence of escapes compared to shrubs and grasslands. Among all the environmental conditions analyzed, wind speed stands out as the predominant factor that affects the risk of prescribed fire escaping. ConclusionsThese findings mark an initial step in identifying high-risk areas and periods for prescribed fire escapes. Understanding these patterns and the challenges of quantifying escape rates can inform more effective landscape management practices.more » « less
-
GrantWilliamson (Ed.)Increasing wildfire activities across the Great Plains has raised concerns about the effectiveness and safety of prescribed fire as a land management tool. This study analyzes wildfire records from 1992 to 2020 to assess spatiotemporal patterns in wildfire risk and evaluate the role of prescribed fires through the combined analysis of wildfire and prescribed fire data. Results show a threefold increase in both wildfire frequency and area burned, with fire size increasing from east to west and frequency rising from north to south. Wildfire seasons are gradually occurring earlier due to climate change. Negative correlation between prescribed fires in spring and wildfires in summer indicated the effectiveness of prescribed fire in mitigating wildfire risk. Drought severity accounted for 51% of the interannual variability in area burned, while grass curing accounted for 60% of monthly variability of wildfires in grasslands. The ratio of wildfire area burned to total area burned (dominated by prescribed fires) declined from over 20% in early March to below 1% by early April. The results will lay a foundation for the development of a localized fire risk assessment tool that integrates various long-term, mid-term, and short-term risk factors, and support more effective fire management in this region.more » « less
-
Abstract Prescribed fire has been increasingly promoted to reduce wildfire risk and restore fire‐adapted ecosystems. Yet, the complexities of forest ecosystem dynamics in response to disturbances, climate change, and drought stress, combined with myriad social and policy barriers, have inhibited widespread implementation. Using the forest succession model LANDIS‐II, we investigated the likely impacts of increasing prescribed fire frequency and extent on wildfire severity and forest carbon storage at local and landscape scales. Specifically, we ask how much prescribed fire is required to maintain carbon storage and reduce the severity and extent of wildfires under divergent climate change scenarios? We simulated four prescribed fire scenarios (no prescribed fire, business‐as‐usual, moderate increase, and large increase) in the Siskiyou Mountains of northwest California and southwest Oregon. At the local site scale, prescribed fires lowered the severity of projected wildfires and maintained approximately the same level of ecosystem carbon storage when reapplied at a ~15‐year return interval for 50‐year simulations. Increased frequency and extent of prescribed fire decreased the likelihood of aboveground carbon combustion during wildfire events. However, at the landscape scale, prescribed fire did not decrease the projected severity and extent of wildfire, even when large increases (up to 10× the current levels) of prescribed fire were simulated. Prescribed fire was most effective at reducing wildfire severity under a climate change scenario with increased temperature and precipitation and on sites with north‐facing aspects and slopes greater than 30°. Our findings suggest that placement matters more than frequency and extent to estimate the effects of prescribed fire, and that prescribed fire alone would not be sufficient to reduce the risk of wildfire and promote carbon sequestration at regional scales in the Siskiyou Mountains. To improve feasibility, we propose targeting areas of high concern or value to decrease the risk of high‐severity fire and contribute to meeting climate mitigation and adaptation goals. Our results support strategic and targeted landscape prioritization of fire treatments to reduce wildfire severity and increase the pace and scale of forest restoration in areas of social and ecological importance, highlighting the challenges of using prescribed fire to lower wildfire risk.more » « less
-
Each year wildfires result in billions of dollars in property damage. Being one of the major natural hazards, wildfires nowadays are also a global affair whose negative impact is particularly devastating in developing countries. As wildfires are expected to become more frequent and severe, more accurate models to predict wildfires are vital to mitigating risks and developing more informed decision-making. Artificial intelligence (AI) has a potential to enhance wildfire risk analytics on multiple fronts. For example, deep learning (DL) has been successfully used to classify active fires, burned scars, smoke plumes and to track the spread of active wildfires. Since wildfire spread tends to exhibit highly complex spatio-temporal dependencies which often cannot be accurately described with conventional Euclideanbased approaches, we postulate that the tools of topological and geometric deep learning, specifically designed for non-Euclidean objects such as manifolds and graphs, may offer a more competitive solution. We validate the proposed methodology to predict wildfire occurrences in Greece and several regions of Africa. Our results indicate that the Firecast Zigzag Convolutional Network (F-ZCN) outperforms the current baseline methods for wildfire prediction and opens a path for more accurate wildfire risk analytics, even in scenarios of limited and noisy data records.more » « less
