Wildfires in Africa, 2026: a hundred and thirty-four times Europe

While Europe closes out a summer that has already passed the halfway point of the record-breaking 2025 season, on the other side of the Mediterranean an entire continent is burning on a scale that makes those figures look almost negligible. Data from GWIS (the Global Wildfire Information System), the Copernicus Emergency Management Service system that has monitored fires globally since the early 2000s by combining MODIS and VIIRS satellite detections, show that between January 1 and August 31, 2026, more than 85 million hectares burned across 52 African countries and territories: over a hundred and thirty times the area burned in the same period across the European Union. And yet, as the comparison with the 2012-2025 average for the same period shows, this African season is not an anomaly. It is substantially below the norm.

Figure 1. Share of national territory burned in each African country: the average for the same period (January 1 to August 31) over 2012-2025, compared with the 2026 season updated to August 31. Both maps use the same color scale, so the colors are directly comparable. Source: GWIS, Global Wildfire Information System (JRC).

A scale with no equivalent in Europe

The first striking fact is purely one of magnitude. Across the 52 African countries and territories tracked by GWIS, the area burned from January 1 to August 31, 2026 exceeds 85 million hectares. Over the same period, the 23 EU countries for which GWIS records meaningful figures totaled roughly 635,000 hectares: the Democratic Republic of Congo alone, with over 20 million hectares burned, accounts for more than thirty-one times the entire European Union. Angola, with nearly 14.7 million hectares, is twenty-three times the EU total. Even a single mid-sized country such as Zambia or South Sudan burns, on its own, more hectares than the whole of the European continent tracked by EFFIS.

This is nothing new: the scientific literature attributes to Africa somewhere between 60% and 70% of all burned area worldwide each year, most of it in tropical savannas and grasslands, where fire has long been part of traditional land management, used for grazing and to renew agricultural land.

This is also why some countries reach such extreme shares, over 30% of the entire national territory burned in a single year for South Sudan, Angola and Zambia. Across the vast grass plains and seasonal wetlands that cover much of these countries, such as the Sudd region in South Sudan, pastoral communities systematically burn dry vegetation every year, at the end of the dry season, to encourage fresh grass regrowth for livestock, to help with hunting, and to prepare land for the following agricultural season. It is a cycle that repeats on the same land year after year: this is why the figure does not describe a sequence of disasters, but a largely intentional, recurring fire regime, very different in nature from the wildfires that strike the Mediterranean parts of Europe. Independent confirmation comes from the scientific literature: a study of a savanna reserve in West Africa measured that 71.9% of its area burns on average every year (Zoffoun et al., 2023), an order of magnitude consistent with what is observed in South Sudan.

Figure 2. Hectares burned in 2026 (data as of August 31) in the ten African countries with the largest burned area, compared with the total for the EU countries tracked by GWIS over the same period. Source: GWIS.

A season below average, not a disaster

The more interesting finding, however, comes from the historical comparison. The 2012-2025 average for the same period of the year, across the same 52 countries, is around 231 million hectares: the 2026 season, at 85 million hectares, stands at roughly 37% of that average. In other words, the continent that on its own burns more than a hundred times as much as Europe is, this year, experiencing a fire season markedly below its own historical norm.

This is not a one-off feature of 2026: it confirms a long-term trend documented by several scientific studies. Global burned area has declined by roughly 24% over the past two decades, with the sharpest drop occurring precisely in African savannas, and agricultural expansion is identified as one of the main drivers of this change, together with landscape fragmentation, which makes it harder for large fires to spread (Andela et al., 2017). An analysis focused on Central Africa alone estimated a decline in burned area of 2.7 to 3.2 million hectares per year between 2003 and 2017, about 1.3% per year, driven mainly by a reduction in the number and size of large fires (Jiang et al., 2020). A separate continent-wide assessment covering 2002-2016 measured an 18.5% decline in burned area, with 80% of that reduction concentrated in the northern hemisphere (Zubkova et al., 2019).

The causes of this decline remain partly debated: studies agree that agricultural expansion and intensification play a role, but not all of them find it to be the main explanation, and the same 2002-2016 study finds that variation in rainfall and soil moisture accounts for a substantial share of the decline, in some cases more than land-use change does (Zubkova et al., 2019). What appears solid in the recent literature is that the African continent remains by far the largest single contributor to burned area worldwide, even as that area shrinks year after year (Andela et al., 2017; Jiang et al., 2020).

Figure 3. Burned area in 2026 (data as of August 31) compared with the average for the same period over 2012-2025, for the ten African countries with the largest burned area this year. In all ten cases 2026 is markedly below the average. Source: GWIS.

A note on methodology

The data in this piece come from the GWIS Statistics Portal (“Estimates” section), which estimates burned area through a spatial-temporal clustering algorithm applied to thermal anomalies detected by the MODIS and VIIRS sensors. This differs from the approach EFFIS uses for Europe, which relies on near-real-time detection from 250-meter resolution MODIS data supplemented by Sentinel-2: the two methodologies are not directly comparable hectare for hectare, even though both belong to the wider Copernicus ecosystem. It should also be noted that GWIS data can include intentional land-management fires (controlled burns for grazing or agricultural renewal), which are especially common in African savannas, and does not systematically distinguish these from unplanned wildfires, a factor worth bearing in mind when interpreting the large absolute figures cited in this article. Data were collected from the GWIS portal on August 31, 2026 and cover the period from January 1 to August 31 of each year considered.

---

Sources: GWIS, Global Wildfire Information System, Copernicus Emergency Management Service, JRC (data updated to August 31, 2026); Andela et al., “A human-driven decline in global burned area”, Science, 2017; Jiang et al., “Observed changes in fire patterns and possible drivers over Central Africa”, Environmental Research Letters, 2020; Zubkova et al., “Changes in Fire Activity in Africa from 2002 to 2016 and Their Potential Drivers”, Geophysical Research Letters, 2019; Roteta et al., “Development of a Sentinel-2 burned area algorithm: Generation of a small fire database for sub-Saharan Africa”, Remote Sensing of Environment, 2019; Zoffoun et al., “Distribution patterns of fire regime in the Pendjari Biosphere Reserve, West Africa”, Journal of Arid Land, 2023; map and statistical processing by Piero Boccardo, Politecnico di Torino.