IF THE area that gets affected by wildfires each year were considered a country, it would be the second largest emitter of carbon dioxide (CO₂), surpassed only by China. In 2023, wildfires globally released 7,330 million tonnes of CO₂, according to the EU’s Copernicus Atmosphere Monitoring Service (CAMS). This is significantly higher than the 6,000 million tonnes of greenhouse gases (CO2 and other gases) emitted by the US in 2022 (see ‘Big emitters’,).

This year, as many as 11 countries have reported wildfires as of May 15. In Canada, home to 9 per cent of the world’s forests, significant fires were reported as early as May 9, leading to evacuation orders in several towns in British Columbia and Alberta provinces. By mid-May, an estimated 55 million tonnes of CO₂ (MtCO₂) had already been released into the atmosphere from these fires, according to CAMS. The May 2024 North American Seasonal Fire Assessment and Outlook, issued by government agencies in the US, Canada and Mexico, highlights that the current warm, dry, and windy conditions could trigger additional wildfires. This raises concerns that the 2024 Canadian wildfires could follow the path of the record-breaking 2023 event, which destroyed 18.4 million hectares—an area bigger than Greece—and released 1,760 MtCO₂.

The actual wildfire emissions could be even higher because calculating them is challenging and “probably underestimated,” says Cynthia Whaley, a researcher at Environment and Climate Change Canada, a government department that coordinates the country’s environmental poli-cies.

Calculating wildfire emissions is complex because it is influenced by variables like temperature, wind, humidity and drought, which vary greatly and unpredictably. In the past four decades, global warming has further created conditions ripe for wildfires, leading to more frequent and intense events. World Weather Attribution, an international team of climate scientists, states that the 2023 wildfires in Quebec, Canada, were at least twice as likely and 20-50 per cent more intense due to climate change.

Wildfires are also spreading to new areas. A 2021 study published in the Proceedings of the National Academy of Sciences (PNAS) found that fires in the western US have been spreading to higher elevations, where blazes have been rare historically, because of warmer and drier conditions. The other concern is that the rising frequency and intensity of wildfires do not give forests enough time to regrow to absorb the emissions. Traditionally, 80 per cent of the carbon released during a wildfire is reabsorbed by new vegetation, while the remaining 20 per cent contributes to atmospheric CO₂, according to a 2022 paper published in Science Advances. But climate change creates a feedback cycle: Hot temperatures lead to extreme weather and droughts, which worsen wildfires and inject more CO₂ into the atmosphere. This, in turn, traps more heat, perpetuating the cycle (see ‘Caught in a loop’,).

ASSESSMENT DIFFICULTIES

“The margin of error for wildfire-related CO₂ emissions can be as high as 20 per cent,” Guido van der Werf, a researcher at Wageningen University in the Netherlands, tells Down To Earth (DTE). In contrast, the uncertainty for CO₂ emissions from fossil fuel combustion is around 6 per cent, according to a 2021 paper published in Scientific Data.

According to the guidelines for national greenhouse gas inventories by the UN Intergovernmental Panel on Climate Change (IPCC), scientists need to know four things to compute wildfire emissions. First is the extent of the burned area. Second is the biomass density or vegetation in the burned area. Third is the emission factor, which is the mass of pollu-tant produced per unit dry mass of fuel (vegetation) burned. The last is the combustion factor of the fires, which tells scientists how much coal charcoal is left after wood is burned. “Once the values are available, it is only simple multiplication,” says van der Werf. India and many other countries use the IPCC guidelines to ascertain wildfire emissions.

The Global Fire Assimilation System (GFAS) used by EU’s CAMS employs a slightly different approach. It is modelled around the fire radiative power, which represents radiant heat released from detected fires in megawatts from satellite data. “It is a measure of the brightness temperature anomalies in different parts of the infrared spectrum. A high anomaly indicates a heat source,” Mark Parrington, senior scientist at CAMS, tells DTE. Next, they convert the observations to ascertain the vegetation consumed by the fire and then apply emission factors to estimate the emissions. “We use this measurement as it is available in near real-time,” he adds.

No matter which method is used, large uncertainties remain in measuring emissions from wildfires. One key source of uncertainty is fire detection. The accuracy of satellite data can be affected by weather conditions, such as cloud cover. The most widely used tool to measure fires is the Moderate Resolution Imaging Spectroradiometer (MODIS), an instrument on two low-Earth NASA satellites Terra and Aqua, which scan the Earth’s surface at least once every day or two. “But we do not know about a lot of fires that occur outside of those measurement times,” says Parrington. The two satellites, for instance, cross tropical India at about 10 am and about 1 pm every day. The other option is to use geosta-tionary satellites, which are positioned in a circular orbit at the Earth’s equator. They record the same image at brief intervals, proving to be par-ticularly useful for observations of weather conditions. But the downside is that the image resolution suffers because they are stationed at a great distance of about 36,000 km from the Earth’s equator.

“So, estimates of how much is being emitted are at a lower bound,” says Parrington. Low-Earth satellites also miss small fires, which occupy an area of less than 100 ha. A 2021 PNAS study concludes that the current estimates of global burn based on sensors such as MODIS should be seen as “very conservative”. The team analysed a new set of images obtained using the multispectral instrument sensor aboard the European Space Agency’s Sentinel-2 satellite and found 4.89 million square kilometres (km2) of land burned in Africa in 2016, about 16 per cent of the continent’s total area. Compare this to MODIS’ detection of only 2.72 million km2 burned area in the same year. About 87 per cent of the differ-ence was attributed to the inclusion of small fires. The analysis also found that wildfires in Africa emitted 5,280 Mt of CO₂. This is 31-101 per cent higher than previous estimates.

Another source of uncertainty is smouldering fires, which are the slow, low-temperature, flameless burning of vegetation. Smouldering wildfires have, till date, received little attention. But recent studies show that they are a global concern because they emit very large amounts of carbon, and are difficult to detect and suppress. Smouldering is predominantly observed in peatlands, a type of wetland built of partially decayed plant matter. Covering only 3 per cent of the world’s landmass, peatlands exist across 180 countries and store at least twice as much carbon as any other vegetation type. Fires in peatlands can reach deep layers several metres below the surface, making it difficult to spot the extent of damage.

Then, there are emission factors, which add to the uncertainties. These are derived from laboratory- or field-based ex-periments. Scientists like van der Werf conduct field surveys to capture smoke in bags, which are then transported to the laboratory to calculate the amount of CO₂ emitted from a given mass. Some researchers use aeroplanes to sample the smoke plumes. However, most estimates come from the US and Europe, with little representation from countries of the Global South. Getting emission factors from different regions is important because the values can vary with geographic zones and vegetation groups. Emission factors are not available for India. N H Ravindranath, a retired professor at the Indian Institute of Science, Bengaluru, says the country uses emission factors from tropical regions provided by IPCC.

India’s Third Biennial Update Report to the UN Framework Convention on Climate Change states that the country’s emissions from forest fires contribute a mere 1-1.5 per cent of all global emissions from wildfires, despite being home to about 2 per cent of the total global forest area. The report warns that this could change with climate change because temperature rise would dry out vegetation, making it more fire-prone, especially in Himachal Pradesh and Uttarakhand.

Another issue with the current fire emission inventories is that they do not often arrive at consistent estimates. The estimations between GFAS and Global Fire Emissions Database (GFED), the most widely used fire emissions inventory among scientists, can vary by a factor of two or three, says Whaley. A 2015 study published in Environmental Pollution compared the average annual CO₂ emissions of the most widely used datasets in 2002-11. It found that GFED version 3 reported 6,521 MtCO₂, GFED version 4s reported 7,158 MtCO₂, GFAS reported 6,698 MtCO2, and the Fire Inventory (FINN) by the US-based National Center for Atmospheric Research reported 6,919 MtCO₂.

There are also regional differences. Among the tested datasets, GFED versions 3 and 4s, and GFAS give relatively consistent emissions, but there are large differences in the Northern Hemisphere, South America, Boreal Asia, Central America and Mexico. Also, FINN’s estimate of CO₂ emissions in South America and Southeast Asia (which includes India) is higher than in the other three inventories.

Considering the challenges, the scientific community has come together to address this problem. Parrington calls the work on emissions uncertainties a new science that is rapidly grow-ing. In December 2022, International Global Atmospheric Chemistry, an international community of atmospheric scien-tists, created an initiative called Biomass Burning Uncertainty: Reactions, Emissions, and Dynamics (BBURNED). It aims to better quantify the uncertainty and variability in biomass burning emission estimation by coordinating the international scientific community to improve understanding of the current and future impacts of wildfires on public health and climate.

BBURNED organised its first workshop in November 2023, focusing on emissions inventories and their methodologies. The next workshop in September 2024 will deliberate on fire emissions and modelling to foster collaboration and data sharing. “With more certain estimates, we can improve the models and our simulations of not only what is happening right now but what could happen under different climate change scenarios,” says Whaley.