WildFiresUA is a system that uses (AI) to detect forest fires, analyze their development, and model the emissions emitted

Prediction of fire evolution and dispersion of hazardous by-products such as PM2.5 and PM10 particulate matter, as well as gases such as O3 (ozone), NO2 (nitrogen dioxide), SO2 (sulfur dioxide), and CO (carbon monoxide).

After the onset of the full-scale war in Ukraine on February 24, 2022, our team of scientists, shaken by the shock, immediately began seeking ways in which we could provide assistance and be of value. By March, the issue of radiation safety had become a pressing concern, as the adversary engaged in alarming provocations, occupying the Chernobyl Exclusion Zone and shelling the territories around the Chornobyl and Zaporizhzhia Nuclear Power Plants.

We swiftly pivoted our focus to developing both a module for visualizing the direction of airflows and a simulation of potential emissions in case of damage to one of the reactors at Chornobyl and Zaporizhzhia. Additionally, we worked on creating a service for tracking radiation levels in Ukraine and Europe. For a more detailed account of this, please refer to the Case Study - Modeling of NPP Accident Damage.

Another critical task was to assist government agencies in monitoring and environmental departments in analyzing and forecasting forest fires. This encompassed forecasting fire spread and the dispersion of hazardous by-products, including solid particles like PM2.5 and PM10, as well as gases such as O3 (ozone), NO2 (nitrogen dioxide), SO₂ (sulfur dioxide), and CO (carbon monoxide).

decor

The problems

The lack of accurate and on-time data on fires makes it challenging to address this problem affectively

decor

Prognostic analysis of fires

decor

Air impact assessment

pic map

Our mission was clear: to develop a comprehensive system capable of monitoring, analyzing, and predicting regional air pollution resulting from landscape fires in Ukraine.

This marked the inception of our journey towards creating a monitoring and analysis system for forest fires in Ukraine.

WildFiresUA is our solution—a system designed for the analysis and prediction of regional air pollution caused by gases and aerosols from landscape fires in Ukraine, commencing from February 24, 2022.

WildFiresUA is designed to allow the State Emergency Service of Ukraine (SESU), environmental agencies, and the public to visualize the analysis of air pollution following large-scale fires, mainly forest and field fires. We are also working on a real-time analysis service. It will help understand the dispersion of polluted air and its potential consequences for people.

It's crucial to comprehend that large-scale fires pose a significant threat to human life and health, even in regions distant from the fire zone, as smoke can spread over tens of kilometers and beyond.

The system uses AI to detect fires using cameras that can be installed on tall telecommunications towers and other structures, and then supplements it with satellite data to pinpoint the fire and analyze the area of the fire.

Then we process all the data, including weather conditions and wind direction, to accurately model the spread of emissions.

Satellite based detection
decor

Wildfires

decor

Video camera detection

decor

Satellite based detection

decor

Measurement of pollution levels by ground stations

decor

Meteorological forecasting and modeling dispersionatmospheric pollution

decor

Visualization of the forecast pollution on the map

decor

Users notification

decor

Response system

decor

The system displays fires detected using satellites with varying levels of detail. On the interactive map, we provide information about specific fires, including temperature, emissions quantities (PM2.5, PM10, CO, CO2, NO2, SO2), coordinates, and area in square kilometers. We also classify emissions by types and show their concentration over specific time periods.

Scientifically oriented startup YourAirTest, with the support of partners, contributed to the development of a fuel map for the entire Ukraine using satellite imagery. This map, created using high-resolution optical and infrared data from multiple satellite platforms, offers detailed information about the types and quantities of fuel available in different regions of the country based on the Fuel Characteristic Classification System (FCCS) standards.

By utilizing the fuel map, we were able to assess emissions of various air pollutants resulting from fires in Ukraine due to the war.

An important step was the addition of AI to our system. It uses AI to detect fires using cameras that can be mounted on tall telecommunications towers and other structures, and then supplements this data with satellite imagery to more accurately locate the fire and analyze the area of the fire. All data, including weather conditions and wind direction, are then processed to accurately model the spread of emissions. We were able to start training our AI YAT-Meteo (weather forecasting model), which was able to create meteorological data for the CALPUFF model (a multilayer scattering model with an integrated Lagrangian modeling system) to simulate the spread of emissions from fires and Aermod.

The development of WildFiresUA faced significant challenges, especially in accurately detecting fires and estimating their size. We experimented with different methods, such as thermal anomalies and satellite images, but satisfactory results were hard to come by during pollution verification. Ultimately, we overcame these difficulties by using camera data and supplementing it with Sentinel and MODIS satellite data.

The Copernicus Sentinel mission, consisting of two 180°-phase satellites, played a key role. It allowed us to track the state of the Earth's surface over a wide swath (290 km) with long revisit periods, which facilitated continuous monitoring of the affected areas. Thus, our system reflects the state of these regions before, during and after forest fires.

info

WildFiresUA is a valuable solution that provides critical analysis and forecasting of regional air pollution caused by gases and aerosols from landscape fires in Ukraine, starting from February 24, 2022.

For our users, WildfiresUA offers a user-friendly interface.

To achieve this goal, we have developed seven core directions for our development

YourAirTest — air quality monitoring system

WildFiresUA — evolution analysis and emissions modeling from forest fires in Ukraine and Europe

WRF-YAT Meteo — meteorological research and weather forecasting

Predictive Modeling for NPP Accident Scenarios

MonAir OUT&IN — outdoor and indoor air quality monitoring station

Automatic Gaussian Plume Model (AGPM)

Mobile YourAirTest App

WildFiresUA offers a user-friendly interface for our users and provides accurate and timely information that helps in effective response to fires and their consequences. We are proud to support you in ensuring clean air and protecting the environment.