A new paper titled “The Global Drivers of Wildfire” has been published in Frontiers in Environmental Science, providing insights into the environmental factors that influence wildfires worldwide. This research is the result of a collaboration between LEMONTREE and the Leverhulme Centre for Wildfires, Environment and Society, with the work led by their innovative Giraffe Team.
You might be wondering—what exactly is a Giraffe Team? Much like a ‘tiger team,’ it’s a small, cross-functional group of experts brought together to solve a specific problem or issue, but Sandy and Colin prefer giraffes to tigers!
The research team was led by Olivia Haas, with key contributions from Theo Keeping and Sandy Harrison, Colin Prentice and Jose Gomez-Dans, bringing together a diverse set of expertise to tackle the complex issue of wildfires within the Leverhulme centre.
Key Drivers of Wildfires
The paper explores the environmental controls driving wildfire activity on a global scale. Through analysis of 10 empirical global studies, the team identified several key factors that significantly influence wildfire extent and severity. The research uncovers consistent patterns that have emerged across different ecosystems, leading to an improved understanding of wildfire regimes and their environmental controls. This understanding is critical for refining the models used to predict and manage wildfire risks.
Gross Primary Productivity (GPP) as a Crucial Control
One of the most consistent findings is that variables related to vegetation amount and productivity- essentially, the rate at which plants convert sunlight into energy- plays a pivotal role in wildfires. High levels of GPP before the fire season led to increased fuel loads, creating conditions ripe for larger and more intense fires. Conversely, high GPP during the fire season suggests moist conditions that could inhibit fire spread, as the fuel is likely too wet to burn. This dual role of GPP highlights the complexity of wildfire dynamics and underscores the importance of accurately simulating GPP in wildfire models.
The study suggests that GPP, or its counterpart, Net Primary Production (NPP), could serve as effective surrogates for fuel loads in models, given the limited global field information on actual fuel loads. Accurate simulation of these variables is essential for enhancing model predictions of wildfire regimes. For instance, ensuring that the vegetation components of process-based wildfire models accurately simulate GPP could lead to more reliable forecasts of fire occurrence and behaviour under different environmental conditions.
Vegetation, Climate, and Fire Spread
Role of Vegetation Cover
The research also highlights the significant influence of vegetation cover, particularly the balance between trees and herbaceous plants, on wildfire dynamics. The type and continuity of vegetation are crucial in determining how a fire spreads. Grass-dominated landscapes, for example, tend to result in more extensive burnt areas but lower fire intensity. This is because grasslands provide continuous, low-intensity fuel that can sustain widespread fires but are less likely to produce the intense heat associated with forest fires.
Interestingly, while many fire models currently include plant functional types (PFTs) as predictors, the empirical analyses in this study did not find PFTs to have a direct influence on wildfire properties. This raises questions about the utility of PFT-specific parameterizations, which often multiply the number of parameters needed in models, many of which are poorly specified. The study suggests that instead of relying on PFTs, fire models should focus on specific plant traits that promote fire spread, such as the presence of volatile oils or ladder-fuel structures. Explicitly representing these traits in fire-enabled vegetation models could improve the accuracy of fire spread predictions.
Climatic Influences on Fire Activity
The study also emphasizes the importance of climatic factors, particularly those influencing atmospheric humidity and fuel drying, in driving wildfire activity. Variables such as vapor pressure deficit (VPD) and diurnal temperature range consistently emerge as critical controls on wildfires. These factors are directly related to the drying of fuels, making them more susceptible to ignition and sustained burning.
In contrast, while precipitation has often been used as a proxy for these factors, the study suggests that direct measures of atmospheric dryness are more reliable predictors of wildfire activity. Precipitation’s influence on wildfires is primarily through its effect on fuel wetness or dryness; however, it is the atmospheric conditions that ultimately determine the flammability of the landscape. The research proposes that in the absence of explicit simulation of the fuel bed and fuel moisture, models could replace less effective variables like soil moisture with parameterizations based on climatic factors such as VPD, which are more closely aligned with the process of fuel drying.
Figure 1. Summary of variables selected and considered important for predictions of burnt area in global empirical analyses. The individual bars show the number of studies which included the variable as a predictor (blue), the number of times the variable was selected as the most important driver of burnt area (red) and the number of times the variable was in the top three predictors (pink).
Rethinking Human Impacts on Wildfires
One of the more surprising conclusions from the study is the limited role that ignitions—whether natural, like lightning, or human-caused—play in controlling wildfire occurrence on a global scale. While ignitions are necessary to start a fire, the availability and dryness of fuel are far more critical in determining whether a fire will spread. This challenges the focus of many current models that emphasize ignition sources, such as population density, without fully accounting for the complexities of human impact on fire regimes.
It is common in fire research to use population density as a measure for both anthropogenic ignitions and fire suppression. However, the study argues that this approach conflates two distinct processes and oversimplifies the diverse cultural practices surrounding fire use. For instance, many fires are deliberately set for agricultural purposes, such as preparing fields or removing waste. Lumping these agricultural fires together with wildfires can lead to inaccurate predictions and misunderstandings of fire dynamics. The study advocates for separating the treatment of wildfires and agricultural fires in models, given their different controls and implications for fire management.
The Importance of Landscape Fragmentation
Fragmentation of the landscape, often due to human activity, is another critical factor in wildfire dynamics. The study highlights how fragmentation can limit fire spread, with factors like crop cover and population density indirectly measuring this effect. Fragmented landscapes, characterized by breaks in vegetation continuity, often act as barriers to fire spread, thereby reducing the overall area burned.
However, the impact of fragmentation varies between ecosystems. Some ecosystems, particularly those adapted to frequent fires, may respond differently to fragmentation than those that are not. For example, in fire-adapted ecosystems, fragmentation might reduce fire intensity and frequency, potentially disrupting natural fire cycles and leading to long-term ecological changes. Understanding these differences is crucial for accurate modelling and effective fire management strategies. The study suggests that a more nuanced approach to measuring fragmentation is needed if it is to be implemented effectively in wildfire models.
Figure 2: Empirical analysis of the relative importance (as measured by t-values) of individual predictors for (A) burnt area, (B) fire size and (C) fire intensity. Green represents variables related to vegetation properties, red represents variables related to climate, and purple represents natural and anthropogenic variables related to landscape fragmentation. Redrawn from Haas et al., (2022).
Lessons for Fire Science and Modelling
The findings from this research offer several key lessons for fire science and modelling, with implications for both current practices and future developments:
- Aligning Variables with Fire Processes:
There is a clear need to align the variables used in models more closely with the processes that generate fuel loads and influence fuel drying. This means focusing on factors like GPP and atmospheric humidity, which directly impact fire behaviour, rather than relying on less effective proxies like soil moisture. - Systematic Evaluation of Model Variables:
The study calls for a more systematic approach to evaluating which variables are included in wildfire models. It is crucial to ensure that these variables are not redundant or simultaneously acting on multiple aspects of the fire regime. For instance, separating the simulation of fire intensity from fire size and burnt area in models could lead to more accurate predictions. - Improving Process-Based Models:
While existing empirical models can be useful for projecting the response of wildfires to future climate change, the study emphasizes the need for improved process-based models. These models will be essential for examining the feedbacks between fire and the climate system, ultimately helping us better understand and manage the growing threat of wildfires in a changing world. - Rethinking Human Impacts:
The research challenges current assumptions about the role of human activities in wildfire occurrence. By separating the treatment of wildfires and agricultural fires and reconsidering the role of ignitions, models can better reflect the complex realities of fire regimes. This shift in focus will be crucial for developing more effective fire management strategies in the face of climate change.
Conclusion
This study provides essential insights into the global drivers of wildfires, highlighting the need for more accurate and nuanced models to predict and mitigate the growing threat of wildfires in a changing climate. By aligning wildfire models more closely with the processes that govern fire behaviour and reconsidering the role of human activities, predictive models can be improved.
You can read the full paper here:
Haas, O, Keeping, T., Gomez-Dans, J., Prentice, I.C. & Harrison, S.P. 2024 The global drivers of wildfire. Frontiers in Environmental Science. 12, doi.org/10.3389/fenvs.2024.1438262