Forest and land fires in Indonesia, known in Indonesian as kebakaran hutan dan lahan (abbreviated karhutla), are recurring wildfires that burn forests, peatlands, scrubland and agricultural land across the archipelago, most severely on Sumatra and Kalimantan (Indonesian Borneo) and, increasingly, in Papua. The fires occur every dry season, but become catastrophic in years of strong drought linked to the El Niño–Southern Oscillation (ENSO) and, in some years, the positive phase of the Indian Ocean Dipole (IOD). Smoke from the fires produces a seasonal air-pollution phenomenon known as the Southeast Asian haze, which regularly crosses borders into Singapore, Malaysia, Brunei and Thailand.
Although drought creates the conditions for burning, fires in Indonesia are overwhelmingly started by people, usually to clear land for agriculture, plantations or settlement, or to claim land in disputed areas. Fires on drained peatlands are especially harmful: they can smoulder underground for weeks, are difficult to extinguish, and release large amounts of carbon and toxic smoke. The most severe episodes occurred in 1997–98, 2015, 2019 and 2023, and a further major episode began in 2026.
Karhutla is the standard official term used by the Indonesian Ministry of Forestry (previously the Ministry of Environment and Forestry, KLHK) and by the National Disaster Mitigation Agency (BNPB). It covers fires in forest areas (hutan) and on other land (lahan), including peatland, mineral soil, and concessions for oil palm and timber plantations. Official burned-area statistics are published through the SiPongi monitoring system, while hotspot counts from satellites such as MODIS and VIIRS indicate fire activity but do not measure area burned.
Indonesia's equatorial climate is humid, and fire is rare in undisturbed wet tropical forest. Severe fire seasons are triggered when drought dries vegetation and lowers the water table. Strong El Niño events delay and weaken the monsoon rains over Indonesia, and a positive Indian Ocean Dipole can add to the dryness in the second half of the year. Indonesia's meteorological agency (BMKG) has emphasised that drought alone does not start fires. An ignition source, almost always human activity, is still required.
Fire is widely regarded as the cheapest method of preparing land for cultivation. The World Bank has noted that fires are used to clear land and also to assert claims in areas where land ownership is unclear, disputes exist, or law enforcement is weak. Fire may be used by smallholders, by migrants, by local brokers who clear and sell land, and by companies or their contractors operating plantations. Allocation of responsibility is contested: companies often argue that fires started on community land or spread from outside their concessions, while communities and civil-society groups point to fires within concession boundaries. In 2026, Indonesian authorities again cited cases of land being illegally sold and then burned for oil palm cultivation.
Indonesia contains an estimated 20.6 million hectares of peatland, and, according to the government's Peatland Restoration Agency, more than 45 per cent of the world's tropical peatland area. In their natural state, peat swamps are waterlogged and resistant to fire. Drainage canals dug to prepare land for plantations and agriculture lower the water table and dry the peat, turning it into fuel. Once ignited, peat fires burn below the surface, are hard to suppress, and release far more smoke and carbon per hectare than fires on mineral soil. In 2019, about 44 per cent of the burned area was on peatland.
Repeated burning also changes ecosystems. One analysis estimated that forest converted to grassland through repeated fires accounted for about one fifth of total forest loss in Indonesia between 2001 and 2016, and degraded landscapes become more fire-prone. Greenpeace Indonesia reported that around 1.3 million hectares of the area burned in 2023 had already burned at least once between 2015 and 2022.
The 1997–98 fires occurred during a powerful El Niño and are among the most severe fire events recorded worldwide. Economist Luca Tacconi, in a study for the Center for International Forestry Research (CIFOR), revised the estimate of area affected by fire in 1997–98 from 9.7 million to 11.7 million hectares. He estimated economic costs of US$1.62–2.7 billion from forest degradation and deforestation, and US$674–799 million from smoke-haze pollution, noting that the haze figure was probably an underestimate. The episode prompted regional haze diplomacy that led to the ASEAN Agreement on Transboundary Haze Pollution in 2002.
The 2015 fires took place during a strong El Niño. Official estimates by the Ministry of Environment and Forestry put the burned area at 2.6 million hectares, although independent satellite-based analysis estimated about 4.6 million hectares, nearly double the official figure. Peatlands burned about 50 per cent more extensively than in the previous peak year of 1997, according to a study cited in the scientific literature. At the height of the crisis, daily carbon dioxide emissions from the fires exceeded those from all economic activity in the United States, and the fires were estimated to account for roughly half of Indonesia's total greenhouse-gas emissions that year.
The World Bank estimated the economic cost to Indonesia at US$16.1 billion, equivalent to 1.9 per cent of 2015 gross domestic product. More than 500,000 people suffered respiratory ailments, and at least 25 people were killed directly by the fires. A 2016 study by Harvard and Columbia University researchers estimated that more than 100,000 premature deaths across Southeast Asia could be attributed to the smoke. Singapore's costs were estimated at around S$700 million.
After the 2015 disaster, the Indonesian government introduced several reforms, but severe fires returned in 2019, a year affected by a positive Indian Ocean Dipole and a delayed monsoon. The Ministry of Environment and Forestry estimated 1.64 million hectares burned, while a satellite study using Sentinel-2 imagery estimated more than 3.11 million hectares, roughly double the official figure. KLHK data cited by Greenpeace gave carbon emissions from the 2019 fires at 624 million tonnes of CO2-equivalent.
The World Bank assessed total damage and economic losses of at least US$5.2 billion, about 0.5 per cent of GDP, comprising US$157 million in direct damage and US$5 billion in losses to affected economic activities, based on eight affected provinces between June and October. More than 900,000 people reported respiratory illness, twelve airports suspended operations, and hundreds of schools in Indonesia, Malaysia and Singapore temporarily closed.
The 2023 fire season coincided with another El Niño. Government data indicated roughly 1.16 million hectares burned, and the Directorate for Forest and Land Fire Control reported emissions of 183 million tonnes of CO2-equivalent, about 29 per cent of the 2019 figure. An independent analysis by Greenpeace Indonesia and TheTreeMap, however, estimated an indicative burned area of at least 2.13 million hectares. It found that about 28 per cent of this, around 599,000 hectares, burned in 211 peat hydrological units across seven provinces prioritised for restoration, releasing an estimated 553 million tonnes of CO2-equivalent. The gap between the two accounts illustrated a continuing debate over the reliability and transparency of official fire statistics.
A new and severe fire episode developed in 2026 under what Indonesian and international media described as a "super" El Niño that officially arrived in June. Hotspots multiplied through July and August. The monitoring platform FireWatch, developed by Nusantara Atlas, counted more than 96,000 hotspots as of 27 July and over 154,000 by mid-August, an increase of around 60 per cent. Spatial analysis by the NGO Madani Berkelanjutan found that area indicated as burned rose eightfold, from 15,474 hectares in June to 124,040 hectares in July, with about 50,558 hectares indicated as burned in West Kalimantan alone and around 22,200 hectares inside key biodiversity areas.
Government figures reported by international media put the area burned in the first half of the year at more than 200,000 hectares, with almost 95,000 hectares burning in July alone; reports differ slightly on the exact reporting period. By 14 October 2026, Nusantara Atlas had recorded more than 1.06 million hotspots, surpassing the 885,037 recorded in 2019, and estimated 895,657 hectares burned by August. Smoke pushed air quality to hazardous levels in parts of Malaysian Borneo, forced the cancellation of outdoor Independence Day events there, and led ASEAN to activate its highest, Level 3, transboundary haze alert for the southern part of the region. Reports in 2026 also described government crackdowns on companies whose concessions burned, and environmental groups criticised funding cuts affecting preparedness.
Smoke from vegetation and peat fires contains fine particulate matter (PM2.5), carbon monoxide and other pollutants. Exposure is associated with acute respiratory infections, aggravation of asthma and cardiovascular disease, and premature death. Studies of the 2015 haze estimated more than 100,000 premature deaths across Southeast Asia. A study published in GeoHealth projected about 36,000 premature deaths per year on average across Indonesia, Singapore and Malaysia over the following decades if current trends continued without comprehensive land-management strategies such as peatland restoration. Children, the elderly and people with pre-existing conditions are most vulnerable.
Direct losses include destroyed crops and plantations, timber and infrastructure. Indirect losses arise from disrupted aviation, tourism, trade and schooling. The World Bank's estimates, US$16.1 billion in 2015 and US$5.2 billion in 2019, exclude long-term health effects and reputational damage to Indonesia's palm-oil sector. Economic loss estimates for neighbouring countries are not included in Indonesia's totals.
Peat fires release carbon that has accumulated over thousands of years. Indonesia's fires in severe years add substantially to the country's greenhouse-gas emissions; in 2019, emissions were estimated to be almost double those from fires in the Brazilian Amazon that year. Fires also threaten habitat for endangered species, including orangutans in Kalimantan, and damage protected areas.
Haze regularly crosses borders. Alongside Singapore and Malaysia, the 2026 haze affected Brunei and the Philippines, according to media reports. Transboundary haze has been a longstanding source of diplomatic friction in ASEAN.
Indonesia's Law No. 32 of 2009 on Environmental Protection and Management (UUPPLH) prohibits land clearing by burning under Article 69 paragraph 1(h). Paragraph 2 contained an exception allowing burning with regard to local wisdom (kearifan lokal), which implementing explanations defined as burning of up to two hectares per household for local plant varieties, with firebreaks. Critics argued that the exception was difficult to monitor and could be abused to justify larger burning. Industry groups also challenged strict liability for fires on company land before Indonesia's Constitutional Court in 2017, and reporting in 2020 on the omnibus Job Creation legislation described the removal of the local-wisdom exemption and a softening of the strict-liability rule; the current legal text should be consulted for the exact position. Burning to clear land is a criminal offence, and prosecutions have also been brought against individual farmers under Article 69(1)(h) in conjunction with Article 108.
Following the 2015 disaster, the government created the Peatland Restoration Agency (Badan Restorasi Gambut, BRG) on 6 January 2016 by Presidential Regulation No. 1/2016, tasked with restoring around two million hectares of degraded peatland within five years in seven priority provinces. Government Regulation No. 57 of 2016 strengthened rules on peatland protection and management. A moratorium on new licences for primary forest and peatland, first introduced in 2011 under a partnership with Norway, was extended repeatedly and made permanent by Presidential Instruction No. 5 of 2019. Authorities have also relied on firefighting task forces, weather modification (cloud seeding) and water-bombing aircraft during fire seasons.
The ASEAN Agreement on Transboundary Haze Pollution was signed in 2002 and entered into force in November 2003. Indonesia was the last ASEAN member state to ratify it, in 2014, which analysts said weakened its effectiveness. A roadmap towards a haze-free ASEAN by 2020 was agreed in 2016. Singapore enacted a Transboundary Haze Pollution Act in 2014, allowing action against entities whose conduct contributes to haze in Singapore.
Indonesia's official fire statistics are among the most disputed in the region. Official burned-area figures from SiPongi, which rely on Landsat-based interpretation, have repeatedly been lower than independent estimates: 2.6 million versus about 4.6 million hectares for 2015, 1.64 million versus more than 3.11 million hectares for 2019, and 1.16 million versus 2.13 million hectares for 2023. Coarse-resolution global products tend to miss small fires, which contributes to underestimation. Differences in methodology, definitions of burned area, and timing of reporting make comparison across years and sources difficult. Hotspot data from satellites, as used by Nusantara Atlas and others, track fire intensity but are not a measure of area burned.
Observers identify several persistent obstacles. Law enforcement is uneven and prosecutions of large corporations are rare relative to the scale of burning. Land-tenure conflicts create incentives to burn. Peatland restoration has been slower and more expensive than planned, and decentralised governance leaves implementation largely to provincial and district governments. Following the 2019 fires, researchers noted that despite significant investment since 2015, severe burning returned, suggesting that policy gains were fragile. In 2026, environmental groups argued that budget constraints undermined preparedness even as a strong El Niño was forecast.
Supporters of government action point to periods of reduced fire, such as the relatively mild fire seasons in the years between the major episodes, as evidence that moratoria, restoration and enforcement can work. Whether those reductions reflect policy or favourable weather remains debated.
{{cite report}}: |access-date= requires |url= (help)