The UK, wildfires and the technology designed to reduce risk

    

The entire UK’s land area is a little under 25 million hectares. In 2023, globally, 384 million hectares of land fell victim to fires and the devastation caused by them is worsening.

In the last two years alone, we have witnessed devastating fires in countries like Greece, Portugal, and Canada with holiday makers being evacuated and businesses and homes consumed in minutes.

Wildfires are already beginning to appear in the UK as a result of climate change. Our wetter winters are fuelling growth that provides the ideal conditions for summer fires. And it’s not just forests that are at risk. Rare moorland habitats are very susceptible to wildfires too, which is why the National Trust has been trialing early detection methods on Marsden Moor in West Yorkshire.

Impacts of wildfires

Wildfires don’t just threaten human life and livelihoods, they destroy whole communities, consuming homes, hospitals, schools and businesses – basically, anything in their path. They devastate wildlife habitats killing the plants, animals and insects that live there, leading to a drop in available biodiversity.

The fires also pose a life-threatening danger to the firefighters deployed to tackle the blazes. Attempting to control and extinguish wildfires costs billions. To give you an idea of the financial costs, a study by University College London found that California’s 2018 wildfires alone cost the US a whopping $148.5 billion. Capital losses and health costs within the state amounted to an additional $59.9 billion.

Many wildfires start on the wildland urban interface (WUI), in other words they start in places close to human infrastructure and people. A good UK example of this is the 2022 wildfires in Dagenham and Wennington in East London. Fortunately, no lives were lost, but many homes, shops, vehicles and acres of farmland were destroyed.

Detection

It’s clear that tackling a fire once it has taken hold is far harder than catching it while it is still small. Wildfires can spread at a speed of up to 14.27 miles per hour, so it doesn’t take long for a tiny fire to become an uncontrollable monster. This is why early detection is so important.

The traditional forms of detection have been towers, cameras and satellites. Towers are tall structures, usually in a forest, that have views over the canopy of trees. A human, or a camera, in the tower looks for plumes of smoke rising above the trees. Once spotted, fire fighters can be called. However, by the time a fire is large enough for the smoke to have risen above the tree canopy it is already well established and becoming difficult to tackle. Plus, it can be tricky to pinpoint the exact location of the fire and communicate that to the teams on the ground who may have many thousands of hectares to cover before they reach the source. All of this can delay the process of starting to put out the fire, which in turn allows the fire to spread further.

Satellites can also be used to detect wildfires, but they too come with some drawbacks. The fire has to be fairly large before a satellite will ‘see’ it, and unless the satellite is geosynchronous it may not come into range until the fire is already out of control. However, satellites are a very useful tool for detecting the spread of fires and helping fire fighters understand where the fire is going and how fast it is moving – all very useful intelligence when tackling a fire.

Thanks to advances in technology, other options are becoming possible. Sensors have been developed that can be placed in the forest (or other habitats) and can ‘smell’ tiny quantities of smoke, allowing even a small smouldering pile of leaves to be detected well before it turns into a giant fire. The sensors send an alert and can give an exact location helping fire fighters reach the spot as fast as possible. The sensors can even tell the difference between the fumes of a truck driving by and the smoke from a forest fire.

The Internet of Trees

Great as these advances in technology are, they are only practically useful if the data that the sensors and cameras etc. are picking up can be relayed to a human who can then take the necessary action (e.g. call out a fire crew).

Usually, data is sent via the Internet using mobile networks, but there is often scant mobile coverage in large forests or on remote hills covered with rare heathland. Yet, this is exactly where we need it if the early detection methods are to work efficiently. Silvanet is a mobile network for the trees: the Internet of Trees. With solar-powered Gateways placed at regular intervals the sensors can speak to each other and send their data out of the forest to the people who need to receive it and take action. This doesn’t mean you can use your mobile deep within the forest, but it does mean the forest can ‘talk’ to the outside world and let it know if a fire is starting.

Prediction

Advances in technology are already making both detection and mitigation easier, faster and more accurate.  But imagine if you could predict where a fire will start. This is where Artificial Intelligence (AI) can help.

Although predicting the exact start location of a wildfire is challenging, calculating the risk can provide reasonable prediction accuracy.

Currently, fire risk is determined through a combination of weather information from satellites and, where available, enhanced with data from local weather stations. Fire risk is then calculated to around 1 km2. This is good, but not perfect.

More advanced calculations are based on VPD (vapour pressure deficit) which is the difference between the amount of moisture that’s in the air and the amount of moisture that air could hold at saturation. From a wildfire perspective, consistently elevated VPD (in other words drier air, holding less moisture than saturation) means that ecosystems can more easily ignite and spread fire, leading to the larger, higher-severity wildfires.

Calculating fire risk levels by considering various sources of information (satellite, weather stations and potentially local sensors) and then mapping the risk is a complex and tedious task. But it can be automated and enhanced in accuracy and resolution with the help of AI. Adding more fine-grained information, such as soil and air moisture levels measured by sensors embedded in the forest, would help to take into account the microclimate of the forest. In the future, we may even be able to measure the fuel moisture (grass, leaves and needles), rather than just the soil moisture. This would make risk calculation, and therefore fire prediction, even more accurate.

Firefighting technology

Once a fire is detected it needs to be extinguished. This usually means firefighters and trucks on the ground, and perhaps helicopters or planes dropping huge quantities of water on the fire from above.

If a fire can be detected sooner, then the number of people and the amount of equipment needed can be reduced considerably. Something the size of a bonfire is easily contained by a single team and one truck, but once you have acres of trees aflame, it can require hundreds of people and dozens of fire trucks, plus aerial support.

This can be compounded by accessibility. Even when a fire is spotted in the very early stages, it can take time for the fire crews to reach it. It could be in a remote location, or in a landscape that is difficult to traverse – perhaps vehicles can’t reach the spot. There are many reasons why a fire, even one we know about, can still get ‘out of hand’.

Drones could provide the solution. Imagine if the sensor could detect and pinpoint the fire, and then send an autonomous drone to the very spot it is needed to extinguish the fire. Fires could be caught quickly, safely and (relatively) cheaply.  Dryad has recently received €3.8million EU grant funding to develop drones with this capability and hopes to have the first working units available in the next 2-3 years. This could mark a genuine game changer in the fight against the destruction of wildfires.

The future

Currently, governments around the world are focusing (and spending millions) on increasing the number of firefighters, trucks and planes available to tackle the fires. In the coming years, I expect to see a gradual movement away from this type of spending and instead we’ll see investment in detection and mitigation technology.

The use of improved prediction modelling enhanced by a variety of environmental sensors including those for early detection, combined with automated firefighting methods, like autonomous drones, could see us begin to win the war against the destruction caused by severe wildfires. 

Photo: Caleb Cook

---

About the author

Carsten Brinkschulte is CEO and co-founder of Dryad Networks. Dryad provides ultra-early detection of wildfires as well as health and growth-monitoring of forests using solar-powered gas sensors in a large-scale IoT sensor network. Dryad aims to reduce unwanted wildfires, which cause up to 20% of global CO2 emissions and have a devastating impact on biodiversity. By 2030, Dryad aims to prevent 3.9m hectares of forest from burning, preventing 1.7bn tonnes of CO2 emissions.

Website:  https://www.dryad.net/