Wildfires are increasingly destructive globally, causing significant societal and ecological damage. While climate change, fuel accumulation, and ignition patterns are recognized drivers of increased fire activity, human intervention through fire suppression plays a crucial, yet often underestimated role. Fire suppression acts as a filter, allowing only certain types of fires (typically more intense ones) to burn while extinguishing others. This creates a "suppression bias," where the remaining fires are disproportionately severe. This differs from the "fire suppression paradox," where suppressing fires leads to increased fuel loads and consequently more intense future fires. This research focuses on the direct, immediate impact of the suppression bias, independent of the long-term effects of the paradox, using a simulation model to quantify its effects relative to climate change and fuel accumulation.

Existing literature acknowledges the fire suppression paradox – the unintended consequence of increased fuel loads from fire suppression, leading to more intense future fires. However, the "suppression bias" – the disproportionate removal of low-intensity fires leading to a skewed representation of fire severity – has received less attention. Previous studies have tangentially discussed this bias, but its emergent impacts haven't been comprehensively assessed due to limitations in isolating suppression's effect using empirical data. The ubiquity of suppression makes it difficult to find control landscapes without any suppression; even remote areas experience some fire suppression. Furthermore, the measurement of suppression effort is challenging, and data often only exist for larger fires, obscuring the many smaller fires suppressed during initial attack. This study addresses these limitations using a simulation approach.

The researchers developed a simulation model to assess the magnitude of the suppression bias on wildfire behavior and ecological impacts. The model incorporates fundamental components of fire behavior, including fuel moisture, topography, weather, ignitions, and fuel loading. It simulates fire spread using Huygens' principle and the Rothermel fire spread model, accounting for initial attack and subsequent suppression efforts. Different suppression scenarios were modeled: three regressive scenarios (Moderate, High, Maximum suppression, where lower-intensity fires are suppressed more heavily) and one progressive scenario (Progressive suppression, where higher-intensity fires are suppressed more heavily), in addition to a control with no suppression. The model was run across a range of fuel aridity (vapor pressure deficit, VPD) and fuel loading conditions representative of North American forest ecosystems, encompassing a 240-year time period of modeled increases in VPD and fuel loading, reflecting climate change and fuel accumulation. For each fire simulation, metrics such as proportion burned at high severity, average fire severity, daily and total fire size, and fire severity diversity were calculated. The influence of suppression was compared to the effects of climate change and fuel accumulation to isolate the impact of the suppression bias.

The simulation results demonstrate that regressive suppression significantly increases the ecological impacts of wildfires, leading to higher fire severity. Under maximum suppression, the proportion of area burned at high severity was more than double compared to unsuppressed fires. The mean fire severity increased by 0.21 CBI (Composite Burn Index) units, equivalent to the effect of 102 years of additional fuel accumulation. Conversely, progressive suppression reduced high-severity burning. The study found that regressive suppression accentuates trends of increasing severity due to climate change and fuel accumulation. The area burned increased at a much faster rate under maximum suppression compared to no suppression or progressive suppression scenarios. For instance, across the gradient of increasing fuel aridity, the yearly burned area doubled nearly three times faster under maximum suppression compared to no suppression. Regressive suppression also decreased the diversity of fire effects, resulting in less variability in burn severity. The study also reveals that while regressive suppression keeps fires at smaller sizes, leading to a lower absolute burned area compared to no suppression, it counterintuitively leads to a faster increase in burned area over time in the face of climate change and fuel accumulation. Progressive suppression, in contrast, leads to less extreme fire behavior and effects, effectively reversing the impact of climate change or fuel accumulation.

The findings highlight the significant and previously underappreciated influence of the suppression bias on wildfire behavior and ecological consequences. Regressive suppression, while seemingly reducing immediate impacts by keeping fires small, actually exacerbates long-term problems by selecting for and magnifying the most extreme fire events. This bias also has profound social consequences, shaping public perceptions and potentially hindering support for active fire management strategies. The results highlight the importance of shifting from a regressive to a progressive suppression approach, where higher-intensity fires are prioritized for suppression. While challenges exist in implementing progressive suppression, including social and cultural barriers and practical limitations, the study demonstrates that even less aggressive regressive suppression can significantly reduce the bias. The study suggests incorporating prescribed burning and adaptive management frameworks to better address the challenges and limitations of progressive suppression.

This study demonstrates that the suppression bias is a major driver of wildfire severity and ecological impacts. It emphasizes the critical need to shift towards progressive suppression strategies, allowing for more low- and moderate-intensity fires while effectively managing high-intensity events. This approach is crucial for mitigating the escalating wildfire crisis and ensuring ecosystem resilience in the face of climate change and fuel accumulation. Future research should focus on testing these findings empirically across various ecological and cultural settings, refining simulation models to incorporate greater complexity, and developing effective strategies to overcome the social and logistical barriers to implementing progressive suppression.

The simulation model, while incorporating fundamental aspects of fire behavior, does not capture all real-world complexities. For instance, it does not account for spatial variability in topography, wind direction, or fuel loading within individual fires. It also does not incorporate the feedback loop of the fire suppression paradox (fuel accumulation influencing future fires). Therefore, the results might be a conservative estimate of the overall impact of regressive suppression. The model's generalizability across different ecological contexts also depends on the accuracy of the parameterization for different fuel models and ignition patterns.