The Big Picture

Electric utility risk managers face a daunting challenge: accurately predicting and mitigating wildfire risk in an increasingly volatile environment.

Red Flag Warnings, issued by the National Weather Service, are often used as a critical tool in this effort. They provide a seemingly clear indication of high-risk fire weather conditions. However, the reality is far more complex.

The gap between the broad warnings and the specific needs of utilities is the core problem that must be addressed.

While Red Flag Warnings are essential for general public awareness, they fall short of providing the precise, actionable intelligence electric utilities need to protect their infrastructure and communities.

The Catch

Spread vs. Start: Why That Difference Matters

There is a critical difference between fire spread and fire starts. The primary focus of Red Flag Warnings is on the spread of existing fires. This is crucial for public safety, but it doesn’t directly translate to the risk of ignition from electric utility infrastructure or other sources.

An electric utility’s greatest concern, and ability to mitigate a fire, is often the initial spark, which can be triggered by seemingly less severe conditions than those just examined for rapid fire spread. Therefore, basing operational decisions solely on Red Flag Warnings can lead to either over or under-reaction.

Red Flag Warnings cover broad geographic areas, often spanning entire counties or regions. Electric utilities, however, need to pinpoint risks at the circuit level, so they can act to mitigate the threat.

The Red Flag Warning’s lack of granularity can lead to inefficient resource allocation and unnecessary operational disruptions, or worse yet, inaction.

Dry Lightning Adds Hidden Complexity

In addition, when electric utilities base their situational awareness on Red Flag Warnings, they can misinterpret the consequences of “dry lightning”, which are included under the Red Flag scope but represent a fundamentally different risk profile.

Dry lightning is lightning that strikes the ground absent of significant rainfall. This makes it particularly dangerous because it can easily ignite dry vegetation, leading to wildfires without the natural suppression of accompanying rain.

Why This Can Quickly Spiral

This can lead to the potential for numerous, simultaneous ignitions; a scenario that can quickly overwhelm resources regardless of wind speed or humidity.

This requires a completely different operational response than warnings driven by wind and ignoring the unique challenges of dry lightning can leave electric utilities vulnerable to widespread, uncontrollable fires.

Relying solely on Red Flag Warnings can lead to:

• Overreaction: Implementing costly and disruptive measures, like widespread PSPS, when the actual risk to infrastructure is localized or less severe and a surgical approach would have the same efficacy against the fire.
• Underreaction too: Failing to take necessary precautions when localized ignition risks are high, even if the overall Red Flag Warning doesn’t seem dire.
• Inefficient Resource: Allocation: Deploying resources across a broad area when the true risk is concentrated in specific locations ignitions.
• Liability Exposure: Making operational decisions based on incomplete data, potentially leading to preventable ignitions and subsequent legal repercussions.

How Utilities Can Respond—and Improve

Red Flag Warnings are a valuable piece of the puzzle, but not the whole picture. To change, electric utility risk managers can:

• Understand the Nuances: Recognize that not all Red Flag Warnings are created equal. Train teams to ask deeper questions about what’s happening in their environment, and how they should react to different conditions. Dry lightning warnings, for example, require a different response than warnings based on wind and low humidity.
• Operationalize Granular Data: Don’t rely solely on the broad geographic scope of Red Flag Warnings. Supplement them with more precise data that pinpoints specific areas of risk within their service territory, and operationalizes the response to the granular data in resource deployment, asset hardening, vegetation management and PSPS plans.
• Integrate with Broader Information and Risk Assessment: Use Red Flag Warnings as one input among many in a comprehensive wildfire risk assessment. Consider factors like fuel conditions, topography, and proximity to utility infrastructure.

Red Flag Warnings are a valuable starting point, but they represent only a fraction of the information electric utility risk managers need.

The core problem is the need for precise, localized, and ignition-focused risk assessments.

By recognizing the limitations of broad public warnings and actively seeking more granular data, utilities can move beyond reactive responses and develop proactive strategies that safeguard their infrastructure, communities, and financial stability.

Wildfires are a major threat to communities and their impacts can be devastating. That’s why it’s so important to have accurate information about the time and likelihood of a wildfire reaching a particular area. This is where wildfire decision trigger modeling comes into play. It helps fire agencies assess fire exposure and create evacuation plans with enough warning time.

But there’s one key aspect that hasn’t received much attention: input data uncertainty. To address this, Technosylva led a 2019 peer-reviewed study, published in the academic journal, Science of The Total Environment, showcasing a new stochastic fire simulation decision trigger modeling method that considers potential variations in input data to assess the probability of wildland fire impact.

Insight from the Research

Accurately predicting wildfire behavior is challenging due to uncertainties in input data and the inherent limitations of fire models. Wildfire simulations, which use various models and data to predict fire behavior, have been crucial in managing wildfires. These simulations help fire management agencies plan controlled burns, assess fire risks, and develop real-time strategies for suppression.

Several types of fire spread models have been developed, ranging from physical-based models to empirical and mathematical ones. These models consider various factors like vegetation, weather conditions, and fuel moisture to estimate critical fire behavior parameters such as the rate of spread and flame intensity.

Protecting vulnerable areas, such as wildland-urban interfaces (WUIs), is a primary concern for fire agencies. They need accurate predictions of when and where fires might impact these areas to make informed decisions about evacuations or shelter-in-place strategies. However, the accuracy of such predictions depends on the uncertainty associated with input data, including factors like topography, fuel type, and weather conditions.

To address these challenges, the study presents an innovative approach leveraging wildfire decision trigger modeling. This uses well-established semi-empirical fire spread models and incorporates stochastic elements to simulate decision trigger buffers around areas to be protected. These buffers represent the time it would take for a fire to reach the protected area under various input conditions, providing probabilities of impact.

Defining The Modeling in the Research

The model incorporates random elements into a simulation, which better reflects the chaotic and unpredictable nature of wildfires. By using random data inputs, you can produce a range of possible outcomes, rather than just one deterministic prediction. By simulating different scenarios, it produces a probability map of the fire’s arrival to areas that need to be protected.

To demonstrate the decision trigger modeling concept, the study used data from the Tubbs fire, one of California’s most destructive wildfires. The analysis ran fire simulations backward in time from selected ignition points to the edge of the WUI around Santa Rosa. The simulations estimated the time it would take for a fire to reach the WUI, and the results were compared with the actual fire progression to validate the accuracy of the decision trigger modeling method.

The decision trigger modeling method uses an ensemble approach, varying input conditions such as wind speed, wind direction, and fuel moisture content to account for input data uncertainties. It generates multiple fire simulations under different conditions to provide probabilistic estimates of fire arrival times.

Research Findings

The study found that the fire simulations using the decision trigger modeling method had a high degree of accuracy when compared to the actual fire spread during the Tubbs fire. The evacuation trigger buffers estimated by decision trigger modeling correctly predicted the time of arrival of the fire at the WUI edge. The ensemble method used for weather conditions showed that the predicted rate of spread (ROS) varied significantly depending on the input conditions. This variability highlighted the importance of considering weather uncertainty in fire behavior modeling.

The decision trigger modeling method offers a valuable framework for estimating the probability of wildfire impact in areas to be protected in real time, accounting for input data uncertainties. It provides decision-makers with information about potential fire spread and the associated probabilities, aiding in evacuation planning and the development of fire suppression strategies.

The study emphasized the need for considering uncertainties beyond weather conditions, including model limitations, fire-atmosphere interactions, and data quality. While the decision trigger modeling method showed promise, it is essential to combine it with expert knowledge and judgment for more accurate decision-making.

Next Steps from the Research

The wildfire decision trigger modeling approach presented in this study offers a novel way to predict wildfire impacts on vulnerable areas. By incorporating input data uncertainties and providing probabilistic estimates,decision trigger modeling enhances decision support for fire management agencies.

However, the study highlights the importance of considering various sources of uncertainty to improve the accuracy of fire behavior predictions in operational environments. Decision trigger modeling has the potential to become a valuable tool for assessing fire risk and protecting communities in the face of increasing wildfire threats.

Learn more about how this science is put into practice.

Despite California being a major fire hotspot in the Americas, there is no extensive scientific analysis of operational fire spread models allowing analysis of their performance and drivers leading to model inaccuracies.  Recent advances in technology have allowed monitoring the fire progression of most wildfires every 15 min in the United States through the National Fireguard Detections platform. This data, when available for use on a fire, provides unprecedented capabilities to analyze factors influencing fire behavior and compare the observed and predicted wildfire rate-of-spread (ROS) modeling in fires distributed across different and complex landscapes.

Building on other studies that analyzed these modeling techniques, Technosylva joined with CALFIRE and led a 2023 peer-reviewed study, published in the International Journal of Wildland Fire that assesses the performance of fire spread models used in California by comparing observed fire growth data with simulated data. The analysis reviewed operational settings under different environmental conditions using 1853 California wildfires from 2019 to 2021 to determine what conditions the current models may over, or underestimate ROS and subsequently, the burned area and associated fire impacts on buildings and other assets.

Insight from the Research

Wildfire spread models play a crucial role in predicting how fires propagate, but their accuracy is influenced by various factors, including fuel availability, topography, and weather. Among these models, Rothermel’s semi-empirical model has been widely used for its simplicity and computational efficiency. However, the inherent limitations and assumptions of these models, along with input data quality, can impact their reliability.

This study, conducted in California, aimed to assess the predictive accuracy of wildfire spread models under different environmental conditions. It utilized high-resolution data from the National Fireguard Detections product to compare observed and predicted Rates of Spread (ROS) for 1853 wildfires occurring from 2019 to 2021. The analysis sought to identify conditions under which the models overestimate or underestimate ROS, ultimately affecting the burned area and fire impacts on buildings and assets.

Key observations and findings from the Research

• Fire Progression Data: The study used the National Fireguard Detections product data, offering high temporal resolution to monitor fire progression every 15 minutes. A grid-growing clustering algorithm was employed to classify polygons into individual fire incidents, enabling a quantitative analysis of fire behavior.
• Fire Modeling with WFA-e: Fire simulations were conducted using WFA-e, incorporating various fire spread models, including Rothermel’s surface and crown fire spread models. Fuel type, topography, and weather data were integrated to run simulations.
• Statistical Analysis: The accuracy of the fire spread models was assessed using error metrics such as ROS residuals, mean absolute error (MAE), mean bias error (MBE), and mean absolute percentage error (MAPE).
• Environmental Factors: The study revealed that the accuracy of fire spread predictions was influenced by environmental variables such as wind speed and fuel moisture content (both live and dead). Low wind speeds and high fuel moisture levels tended to lead to underestimations of ROS, while high wind speeds resulted in overestimations.
• Fuel Types: Different fuel types played a significant role in the accuracy of predictions. Models performed relatively well for shrub, grass, and grass-shrub fuel types, while they consistently underpredicted ROS for timber fuel types.
• Overall Model Accuracy: The models had an average MAPE of 47% for automatic fire simulations, with better performance in shrub, grass, and grass-shrub fuel types. Timber fuel types exhibited the highest MAPE (approximately 67%).

The study found that the model errors and biases were reasonable for simulations performed automatically. It identified environmental variables that might bias ROS predictions, particularly in timber areas where some fuel models might underestimate ROS. Overall, the performance of fire spread models for California aligns with studies developed in other regions, and the models are deemed accurate enough to be used in real-time to assess initial attack fires.

Next Steps from the Research

The study highlighted challenges related to pyroconvection, local wind fields, and the estimation of ROS in timber areas. It recommended the development of improved fire spread models to address these challenges and enhance prediction accuracy.

The study found that while current fire spread models have limitations and biases, they are accurate enough to be used in real-time operational settings, particularly with the capability for manual adjustments and calibration. However, there is a need for ongoing improvements, especially for modeling fire spread in timber areas, predicting crown fire behavior, and considering the effects of pyroconvection. This research contributes valuable insights to wildfire prediction and management, emphasizing the importance of continuously refining and enhancing predictive models in the face of growing wildfire threats.

The research underscores the importance of wildfire simulators in supporting planning and incident analysis in real-time, despite the potential uncertainties derived from input data quality and model inaccuracies. The study additionally provides insights into the performance of fire spread models in California, offering a foundation for understanding and potentially improving upon current operational models in the future.

Learn more about how this science is put into practice.