A Commercial Approach to Wildfire Management?

(first published in the UK RSPSoc "Sensed" journal, July 2023) https://www.rspsoc.org.uk/images/SENSED/SENSED_202307_July2023_WEB.pdf )

Wildfires are definitely more in the news here in Canada than they are in the UK – but still felt like somewhat of a distant problem when living in a city in the south of the country. That changed this summer. One day last week (June 25th to be exact), Montreal had the worst air quality of anywhere in the world. For the last few weeks poor air quality brought about by wildfires in Quebec has made the news globally, having impacted millions in Canada and the U.S. This is already recorded as the worst fire season on record – and we are only just into July.

Wildfires in Canada are hardly uncommon – there is a significant fire season every year, generally starting in May. The problem is the frequency and severity of the fires, fire season duration, and how this is changing. Climate change means that winters are becoming less severe – the Arctic is warming at a much faster rate than the global average. This means greater mid-season and/or early snow melt, which can lead to less saturation and dryer conditions. Add to this warmer, dryer and earlier summers and the number of fires in Canada is only expected to rise. This will lead to more cases in which fire management resources become overwhelmed. The poor air quality we have had over the last few weeks is more likely to become a new normal, rather than be a one-off.

Given the significant land mass of Canada, satellites are the most utilized tools for monitoring fire conditions – yes, there are in-situ watch towers, crowd-sourced data, and even passing commercial aviation, but satellites are able to cover the whole land mass regularly. (As a side note – covid really paid havoc with wildfire detection from commercial aircraft…)

So, which satellite sensors? VIIRS and MODIS are the fire monitoring work horses. Both are well established with government departments building current models using these datasets. Data from these sensors provides information on fire extent through heat maps. Each sensor has its advantages: MODIS images more often (up to 4 times a day across Canada), but at coarser resolution; VIIRS has a high ground resolution at 300m, which provides more accurate data on fire extent, but revisit is <3 times daily. Latency is also important – data needs to be acquired and in the hands of decision makers as soon as possible. MODIS (on NASA’s Aqua and Terra satellites) was also not designed as an operational mission; it is R&D, with no direct replacement when its lifespan comes to an end.

These sensors also image earlier in the day. Decision making as to how to fight fires the next day (or triaging) is made with the most up-to-date information. Why is this important? In a country like Canada (and indeed elsewhere in the world) not all fires are extinguished – they are let to burn naturally. There are neither the resources nor necessity to put these fires out. Rather, fires are managed based on the danger they pose to the population, infrastructure and industry. A second factor is the “peak burn” period. This is the time where the fire has the greatest radiative power – or is burning most intensely. A fire may be small, but if it is burning intensely, and if (for example) winds pick up, it could go on a significant run, quickly. This peak burn period tends to be later in the afternoon – after the last acquisitions of the day from MODIS or VIIRS.

Here is where the Canadian Space Agency’s WildFireSat mission will aim to step in. It will collect daily imagery at the time of the peak burn period, and collect data on fire radiative power (FRP) specifically. The satellite is in-development with a launch anticipated for 2029. As it will be the “last image of the day”, data collected should be able to greatly improve triaging, i.e., knowing which fires to tackle, and in what order, the following day. Data from the satellite will be made freely available – as is VIIRS and MODIS – and expected to be utilized with these other datasets.

Data and services to support wildfire management are considered as for the public good – similar to weather services – therefore, can there be a commercial business around this topic? There is also another aspect to consider; whilst public fire management services are not perfect – they work well. A commercial company is unlikely to “replace” known public services – the information collected is too critical to public safety (amongst other factors). Indeed, the same argument could be made for commercial weather services.

However, this is not stopping several companies emerging with a commercial play on how to support fire management. What these companies are doing is looking at other parameters to measure fires and/or to add a robustness to fire data collection. OroraTech in Germany is one such company, with a constellation of thermal imaging satellites they aim to provide a 24/7 fire detection service. It will also use in-orbit processing to reduce latency. The company already has two satellites and has raised €15 million to support development.

The Canadian government through the Canadian Space Agency (CSA) is exploring what companies such as OroraTech can achieve. The CSA recently awarded a contract to Spire Global’s subsidiary ExactEarth to “deliver preparatory work for implementation phases of a wildfire monitoring satellite” as an initial step towards the CSA’s WildFireSat mission. Spire is partnering with OroraTech, for the contract to utilize its “wildfire intelligence service.”

Wildfires have become a hot-topic… so to speak. Other companies, whilst not fully-focused on wildfire applications, are further looking at how they can support the application: for instance, Hydrosat intends to build a thermal and multispectral imaging constellation for a variety of applications (mostly agriculture focused), but does have interest in how its data can be used for assessing wildfire risk. Hyperspectral operators are also looking at how to use the short-wave in heat detection and/or to again assess risk.

What is likely is that any commercial outcome would be with the government as a customer, at least initially. The government procures a service, and in turn provides free solutions for the public good. There are industries which operate in high fire risk environments (obviously the forestry industry, as well as the oil and gas sector) which may have use for tailored solutions, but have become used to public services.

It is also likely that the commercial solution would be used in combination with known, existing fire models which leverage government satellites and ground-based networks. It would be a question of improving existing fire detection, monitoring and management rather than offering something completely different which is yet to be fully proven. Nonetheless, given that wildfire prevalence is only expected to increase, it would seem only wise for governments to look at what the private sector can achieve to improve or augment wildfire management services.