Intelevac Wildfire® is our advanced solution for predicting city-wide evacuation times during wildfires, providing essential information that can save lives.

This tool uses high-performance computing and an open-source traffic simulation tailored for scalability. It updates evacuation routes in real time by combining road network data with information about road closures due to wildfires. The system also adjusts to changes in road capacity and runs multiple simulations to determine how quickly people can evacuate and identify where vehicles might get stranded.

To ensure the highest accuracy, our simulations use cutting-edge computer vision and machine learning to maintain precise road network data, which we gather from both street-level images and high-resolution satellite imagery.

Additionally, we use detailed population data from the U.S. Census Bureau to accurately predict how many people will need to evacuate in different areas, including projections for future population growth. This comprehensive approach makes our simulations both robust and reliable, helping to enhance wildfire emergency planning and response.

Methodology

Microscopic traffic simulations of wildfire evacuation scenarios

An open-source microscopic traffic simulator, Simulation of Urban Mobility (SUMO), is used with scalability in mind for wildfire evacuation simulations. It incrementally combines road network data and road closures caused by wildfires to update the routes that vehicles will take, based on information about their origin and destination locations. This approach also accounts for stochastically changing roadway capacities and specific roadway sections. For each wildfire scenario, this simulation step will be repeated multiple times to obtain a probabilistic distribution of egress times and the number of isolated vehicles.

Seamless up-to-date precise road network data

We employ detailed road network datasets in our microscopic traffic simulator. To construct seamless, up-to-date, and precise road network data for accurate evacuation simulations, we utilize computer vision and machine learning methods with street-level images and high-resolution remotely sensed data.

People

Spatially explicit and detailed population information is critical for accurate wildfire evacuation simulations. We use census data from the U.S. Census Bureau to estimate refined population density and evacuation demands during wildfire events. Our approach employs not only current population data but also projections of future populations.

Wildfire environments

Spatially explicit and detailed population information is critical for accurate wildfire evacuation simulations. We use census data from the U.S. Census Bureau to estimate refined population density and evacuation demands during wildfire events. Our approach employs not only current population data but also projections of future populations.

Sample Report

Below is an overview of our Intelevac Wildfire® sample report, demonstrating the comprehensive insights we provide for effective wildfire emergency planning and response:

Paradise, California

Evacuation Risk Assessment

City Overview

Population and Housing

Demographics

According to the 2022 Census data, the population of Paradise is approximately 6,666. From 2013 to 2022, the population decreased by 19,553. This represents a rate of population change of -74.6% over the period from 2012 to 2022, with a 13.76% decrease observed between 2021 and 2022 alone.

Figure 1. Population changes (from 2013 to 2022) (Paradise, CA)

 

The 2022 Census data for Paradise, California, illustrates a predominantly White population, constituting approximately 81.9% of the town’s total population of 6,666. Other racial groups include Black or African American residents at roughly 0.68%, Asian residents at about 0.41%, and individuals of two or more races making up around 4.05%. This demographic profile highlights a significant majority of White residents, indicating less racial diversity compared to the broader diversity observed in the Chico, CA Metro Area and California.

Figure 2. Population by race and ethnicity (Paradise, CA)

 

Economics

The median household income in Paradise stands at $54,842, representing roughly 80 percent of the median income in the Chico, CA Metro Area ($66,085), and approximately three-fifths of California’s median household income ($91,905).

Figure 3. Median household income (Paradise, CA)

 

Housing

According to the 2022 Census data, the total number of housing units is 3,336, with a decreasing trend observed. The median value of owner-occupied housing units is $344,100, which is approximately 90 percent of the median value in the Chico, CA Metro Area ($371,600) and about half of the median value in California ($659,300).

Figure 4. Value of owner-occupied housing units

City Overview

Roadways and Street

In Paradise, California, cul-de-sacs are commonly found within residential areas and are characteristic of the town’s layout, which include a mix of housing developments nestled among the region’s natural landscapes of forests and hills. The town also features a number of privately-maintained roads, particularly in planned unit developments, which can present challenges for access due to their often narrow widths and limited availability of on-street parking.

Table 1. Road and Street Descirptive Statistics

City Overview

Fire Environment

Weather

Paradise, California, situated in the Sierra Nevada foothills, is influenced by a Mediterranean climate. This climate regime brings hot, dry summers and cool, wet winters to the area. During the summer months, temperatures can rise significantly, often exacerbating the dry conditions. The lack of significant rainfall from late spring through early fall adds to the arid environment, creating a landscape that is prone to drying out. In contrast, winters in Paradise are cooler and marked by increased precipitation, which helps to replenish the region’s vegetation. However, the wet season is relatively short compared to the long, dry periods that dominate the year, contributing to a cycle of growth and drying that fuels the fire season. The seasonal weather patterns, characterized by a stark contrast between wet and dry periods, play a critical role in shaping the natural landscape and influencing the fire risks in the region.

Figure 5. Monthly average precipitation and monthly average temperatures (1981-2010)

 

Topography

Paradise, California, is characterized by its distinctive topography, located in the Sierra Nevada foothills. This region features a varied landscape of rolling hills, steep canyons, and elevated ridges, providing a unique and complex terrain. The town itself is situated at varying elevations, with some areas nestled in more sheltered valleys and others perched on elevated plateaus, offering expansive views of the surrounding natural beauty. 

Figure 6. Overview of topography (Paradise, CA)

 

This variation in elevation and terrain also contributes to diverse microclimates within a relatively small geographic area, affecting local vegetation and wildlife habitats. Such topographical features have implications for land use, development, and especially wildfire behavior, as fires can move rapidly uphill and are influenced by the complex interplay of wind, slope, and fuel availability. The natural beauty of Paradise’s rolling foothills and wooded landscapes is thus intertwined with the challenges of managing natural resources and mitigating wildfire risks in a region with such varied terrain.

Fire History

Paradise, California’s fire history is marked by numerous wildfires, with the most devastating being the Camp Fire in November 2018, California’s deadliest and most destructive wildfire to date. The Camp Fire resulted in 85 deaths and the destruction of about 19,000 structures, almost completely destroying Paradise. Earlier significant fires, such as the Humboldt Fire and the Butte Fire in 2008, had already underscored the town’s vulnerability to wildfire, driven by its forested surroundings, dry conditions, and strong winds. This history highlights the critical need for fire preparedness, vegetation management, and resilient building practices in wildfire-prone areas.

Figure 7. Fire history in California (1970 – 2021)

 

Land cover / Fuel vegetation

The land cover in Paradise, California, contributing to its wildfire vulnerability, predominantly consists of dense forests and chaparral, interspersed with residential areas. The region’s forests are mainly composed of pine and oak trees, which, along with the shrubby chaparral, provide substantial fuel for wildfires. This natural vegetation, combined with the area’s Mediterranean climate—characterized by hot, dry summers and cool, wet winters—creates conditions conducive to the ignition and rapid spread of wildfires.

Figure 8. The percent tree canopy cover of Paradise, CA (2021)

 

The presence of residential communities within these fire-prone landscapes further complicates fire management efforts and increases the risk of property damage and loss during wildfire events. The land cover in Paradise, thus, plays a significant role in the area’s ongoing challenge with wildfire risk and management.

Wildfire Hazard Potential (WHP)

The Wildfire Hazard Potential (WHP) map, a raster geospatial product developed by the USDA Forest Service’s Fire Modeling Institute, stands as a vital instrument for evaluating wildfire hazards and guiding the prioritization of fuel management efforts over vast landscapes. Paradise, California, has a WHP value of 3.4, which exceeds the average WHP for the state of California.

Figure 9. Average wildfire hazard potential score (Paradise, CA)

Evacuation Planning and Preparation

Methods

This methodology leverages geospatial data and simulation tools to assess how urban network characteristics influence evacuation effectiveness and overall vulnerability to wildfires, providing valuable insights for urban planning and disaster management.

Figure 10. Overview of  Intelevec: Wildfire

 

Network Connectivity Measure Calculation

Four key measures of network connectivity for selected cities: diameter (d), alpha (a), beta (b), and gamma (g). These measures are determined based on the cities’ Open Street Map (OSM) network data and provide insights into the network’s size, complexity, redundancy, and connectivity. The calculations aim to assess the cities’ network vulnerability to wildfires

 

Table 2. Network connectivity measurement metrics

 

Table 3. Network connectivity calculation

 

 

Microscopic Traffic Simulations of Evacuation Scenarios

The framework evaluates network vulnerability through agent-based evacuation simulations under various scenarios, including no wildfire and several wildfire scenarios with varying degrees of roadway closures (1%, 3%, 5%, 7%, and 10%). These scenarios reflect potential disruptions caused by wildfires, such as ember fires. The simulations, conducted using the SUMO traffic simulation tool, calculate the number of isolated vehicles (IV) and the evacuation time estimate (ETE) for each scenario, with each simulation run 50 times to obtain median and distribution values.

Figure 11. Flow of agent-based evacuation simulation



Evacuation Planning and Preparation

Results

Connectivity Metrics

Paradise is the least connected in that it has the lowest values of a, b, and g.

Paradise’s network size is larger and the city has a higher d than the other cities. It is not possible to compare the connectivity of Paradise and the other cities based on only their ds because the high d indicates a large network or poor network connectivity. However, it can be checked that Paradise is the least connected in that it has the lowest values of a, b, and g

 

Table 4. Network Sizes and Network Connectivity Measures of Diameter (d), Alpha (a), Beta (b), and Gamma (g)  [THESE ARE GENERATED NUMBERS]

 

 

Evacuation Measure

  • Isolated Vehicles (IV)

It is evident that increased likelihoods of roadway closures due to wildfires lead to a higher number of isolated vehicles (IV). However, the extent of this increase varies across different cities. Paradise, identified as the city with the lower network connectivity, exhibits the most significant impact from wildfire-induced roadway closures. Specifically, when 10 % of Paradise’s original network is rendered inaccessible by wildfires, approximately 45 % of evacuees find themselves unable to leave the area. In contrast, Atascadero has a much smaller portion of vehicles (about 19%) are isolated. 

Figure 12. Comparison of isolated vehicles (IVs) – 10% of roadway closure [THESE ARE GENERATED NUMBERS], Might be better to compare to County and State

 

Evacuation Measure

  • Evacuation Time Estimates (ETE)

The analysis of Evacuation Time Estimates (ETE) with provided route information, as depicted in Figure 12, reveals several key insights. Firstly, an increase in the probability of roadway closures directly correlates with longer ETEs. Secondly, the degree to which ETEs lengthen varies by city. For example, Paradise and Atascadero initially exhibit comparable ETEs under no-wildfire conditions. However, Paradise’s ETE significantly escalates with increased roadway closures. This suggests a higher sensitivity in Paradise’s road network to the random closure of roadways, as evidenced by the broad interquartile range in the data. 

Figure 13. Comparison of evacuation time estimates (ETEs) – 10% of roadway closure [THESE ARE GENERATED NUMBERS], Might be better to compare to County and State

Evacuation Measure

This variance indicates the presence of critical roadway links, the closure of which can lead to notable congestion. Figure 12 illustrates this with several outliers, indicating significant variations in ETE for both Paradise and Atascadero, highlighting the presence of crucial roadways that substantially influence ETE. Also, Figure 12 explores ETEs without route information, revealing an opposite trend: ETE decreases as more roadways are closed. This counterintuitive finding suggests that increased involuntary vehicle (IV) presence actually mitigates traffic congestion by reducing the number of non-IVs on the network, thereby facilitating more efficient evacuation for those remaining. However, this does not suggest that a higher number of IVs is beneficial for evacuation scenarios overall. The reduction of vehicles primarily aids in congestion management rather than evacuation efficiency, as evacuees without vehicles may face longer evacuation times on foot, along with direct exposure to wildfire hazards. An extreme scenario where all vehicles are immobilized highlights the dire consequences of such situations, underscoring the critical need for effective evacuation planning and the importance of managing IVs during such emergencies.

 

Click the button below to access the link for downloading a Sample Report provided by Corners Global.

FREE DOWNLOAD

Send download link to:

How to Request a Report

To request a customized Intelevac Wildfire® report for your region, please contact us with your specific needs and geographic focus. Our team will provide a detailed, actionable report to enhance your wildfire emergency planning and response capabilities.