WildfireSight FAQ

WildfireSight is a technology-based company focused on all three distinct phases of wildfire: before, during, and after wildfire. Its solutions are designed to help organizations forecast risk, improve situational awareness, and support smarter decisions across the wildfire cycle.

Z1 is WildfireSight's AI-powered wildfire forecasting model. It provides daily wildfire risk predictions up to 6 months ahead for locations across the regions it covers. Z1 is designed to help users identify where wildfire risk is building before ignition, so they can act earlier and more effectively.

WildfireSight delivers the results of Z1 through an interactive WebGIS App. The platform allows clients to view wildfire risk forecasts spatially, monitor changing conditions, and use mapped intelligence to support planning, preparedness, operations, and risk management.

It is both. WildfireSight provides a technology platform through its WebGIS App, supported by advanced forecasting models and specialized wildfire risk intelligence services for client needs.

Z1 is designed to generate daily wildfire risk predictions up to 6 months ahead. This gives clients a planning horizon far beyond traditional short-term wildfire danger tools and helps them prepare for risk that may build well before a fire starts.

The Fire Weather Index, FWI, is Canada's official wildfire danger system. It has been built by decades of scientific work and trusted since the 1970s. However, it has an important limitation: it is fundamentally focused on short-term fire danger, typically about one day ahead.

WildfireSight's Z1 is different. It is designed to forecast where wildfire risk is building before fire starts, with a much longer forecasting horizon and a more targeted decision-making value for clients.

WildfireSight is not trying to replace the importance of FWI, but it addresses a major gap that FWI does not solve well: longer-range, location-specific wildfire risk forecasting.

Its advantages include:

• forecasting wildfire risk much farther ahead
• helping users see where risk is building before ignition
• providing more targeted insight for planning and prioritization
• reducing unnecessary alarm by focusing attention more precisely

A simple way to describe this is: less alarm, more precision.

Because longer-range forecasting is more difficult, not easier. In fact, that normally favors short-range systems like FWI. This is why WildfireSight's performance is especially meaningful.

During the 2025 fire season in Alberta, WildfireSight compared Z1's 14-day forecast results against the government's 1-day FWI forecast. Even at that longer forecasting horizon, Z1 still outperformed FWI. That matters because it shows stronger forecasting value under a more difficult prediction window.

In the 2026 fire season, Z1's forecasting capability extends even further, providing daily wildfire risk predictions up to 6 months in advance. In other words, the 14-day comparison was the specific benchmark used in the 2025 Alberta performance comparison, while the broader Z1 platform is designed to support much longer forecasting horizons.

Z1's accuracy has been validated against 55 years of wildfire data published by the Government of Canada. Based on this validation, Z1 demonstrated approximately 60% to 80% accuracy depending on fire cause.

This range reflects the reality that different wildfire causes, such as human-caused versus lightning-caused fires, behave differently and present different forecasting challenges.

During the 2025 fire season in Alberta, WildfireSight's Z1 forecasted 74.5% of all occurred wildfires, while the government's FWI forecasted about 65%.

At the same time, Z1 flagged 20% less land as high danger.

That means WildfireSight did not just forecast more fires successfully, it did so while flagging less land as high danger. In practical terms, that points to better precision, less over-warning, and more efficient attention to where risk is truly building.

It means Z1 was more selective and more precise in identifying where serious wildfire risk was developing. Instead of broadly marking larger areas as high danger, it focused risk more efficiently.

For organizations making operational or strategic decisions, this matters because excessive warning across too much land can create noise, dilute attention, and reduce practical usefulness. WildfireSight aims to provide clearer signal with less unnecessary alarm.

No. WildfireSight as a company works across before, during, and after wildfire. However, the Z1 model is primarily focused on forecasting wildfire risk before ignition, which is where early planning and prevention value can be strongest.

WildfireSight is currently available in Alberta. Expansion is underway, and the platform will soon be available in British Columbia and Saskatchewan.

WildfireSight provides services for a wide range of sectors, including:

• oil and gas
• forestry and timber supply companies
• mining
• cities and municipalities
• Indigenous and heritage nations
• insurance companies
• utilities and infrastructure operators
• land and resource management organizations
• emergency and risk management teams

WildfireSight helps organizations:

• identify wildfire risk earlier
• improve seasonal and operational planning
• prioritize assets, people, and field activities
• strengthen preparedness and mitigation decisions
• reduce uncertainty through more focused risk intelligence
• support leadership, operations, and risk communication with mapped forecasting outputs

No. WildfireSight is designed to complement public systems and conventional wildfire danger tools. It adds value by offering longer-range forecasting and more targeted wildfire risk intelligence that supports planning and decision-making beyond standard short-term danger maps.

To inquire about licensing, access, pricing, or partnership opportunities, please contact contact@wildfiresight.com.

Yes. If you would like to explore how WildfireSight may fit your organization's operations, planning, or risk management needs, please contact contact@wildfiresight.com.