The World’s Biggest Deforestation Project

1Background: the last great forest island

New Guinea is divided by a straight colonial border. The western half belongs to Indonesia and is now split into several Papuan provinces. The eastern half is the independent nation of Papua New Guinea. Together they hold one of the last great tropical forest wildernesses on Earth.

In 2020, 99 botanists from 19 countries published the first verified checklist of the island’s plants. They counted 13,634 species, 68% of them found nowhere else, making New Guinea the most floristically diverse island in the world, ahead of Madagascar and Borneo.^[20] Its forests and swamps are also home to birds-of-paradise, tree kangaroos and countless species that science has not yet described.

Merauke sits in the island’s far south-east, beside the border with Papua New Guinea. Its lowlands of forest, savanna and wetland are the ancestral lands of Indigenous Papuan clans whose food, medicine and identity are tied to them. It is here that the Indonesian state has chosen to build its largest agricultural project.

2From an old plan to a national priority

The idea is not new. A Merauke food and energy estate was launched under President Susilo Bambang Yudhoyono in 2010 and largely failed. President Joko Widodo revived it in 2023 and gave it National Strategic Project status, and land clearing for sugarcane began in 2024.^[1] Coverage at the time described it as the largest deforestation undertaking in the world: about 2 million hectares of sugarcane plantations and another 1 million hectares of new rice fields. Officials have given conflicting figures for the scale, with some citing 1 million hectares of rice and 500,000 of sugarcane, and others up to 2 million hectares for either crop.^[1]

Under President Prabowo Subianto, who took office in October 2024, the project has been placed at the centre of a drive for food and energy self-sufficiency. A presidential decree in August 2025 created a high-level coordination team under the coordinating minister for food, which reaffirmed the estate as a national strategic project and extended the designated area into neighbouring regencies including Mappi and Asmat.^[10] Reporting from the ground describes large-scale clearing under way with military backing, and a politically connected conglomerate, the Jhonlin Group, deploying some 2,000 excavators imported from China to build rice fields. The state company appointed to run the rice estate was still unfunded, according to the same reporting.^[9]

Indigenous Papuans in Merauke say they were not consulted before their land was cleared, and describe a heavy military presence that they see as aimed at silencing opposition.^[9] Investigative reporting has documented how sugar companies acquired Indigenous land with government and military backing.^[12] In February 2026, police dispersed Papuans protesting the food estate.^[11]

3The business model: state power and private capital

What makes the current period different is not one project but a model that links several levers of power. Each one lowers the cost or the risk of converting forest, and together they reinforce one another.

The state becomes the planter. Under a 2025 presidential regulation, a forest task force known as Satgas PKH has reclaimed control of millions of hectares of oil palm plantations that were operating illegally inside state forest zones. Rather than returning all of this land to forest, much of it has been transferred to a new state-owned company, PT Agrinas Palma Nusantara, which now manages around 4.1 million hectares of plantation land.^[5] In a single handover in May 2026, 2.37 million hectares were transferred to the company.^[13] Officials say conservation and protected forests within the seized land will be restored. Critics note that one study estimated about 614,235 hectares of the land handed over belonged to smallholder farmers and was wrongly classified.^[14] Cracking down on illegal plantations is a legitimate goal. Converting them into state assets also turns past illegal clearing into permanent state production, and makes the government the largest owner of the business it is supposed to regulate.

Demand is guaranteed by law. Since January 2025 Indonesia has required diesel to contain 40% palm-based biodiesel. In early 2025 the energy ministry said moving to a 50% blend would need about 2.3 million hectares of additional plantation land.^[15] The ministry later said the 50% blend could be supplied through replanting, while still higher blends would need new land.^[16] In January 2026 the 50% blend was postponed amid supply concerns.^[17] The direction of travel is clear even if the timing is not: every step up in the mandate pulls more palm oil out of the economy and into fuel tanks.

The signal from the top. At a national planning meeting on 30 December 2024, President Prabowo called for more oil palm plantations and dismissed concerns about deforestation, arguing that oil palms absorb carbon because they are trees with leaves. Environmental groups called the remarks dangerous; the main industry association welcomed them.^[4] Weeks later the government said it was targeting 2.3 million hectares of protected forest for food and biofuel crops.^[18] Ecologists point out that a plantation of a single species cannot do what a forest does: it stores far less carbon, supports a fraction of the wildlife, and regulates water far less effectively.

4What the satellites show

For several years Indonesia was held up as a rare success story, with deforestation falling from its peaks of the 2010s. That trend has reversed. Auriga Nusantara counted 433,751 hectares of deforestation in 2025, up 66% from 261,575 hectares in 2024.^[2,3]

The increase is concentrated where the state has planned it. In 2025 the government allocated 20.6 million hectares of forest areas to food, energy and water programmes, and 43% of that area was natural forest. Auriga found that 78,213 hectares, or 18% of all deforestation in the country that year, happened inside these designated areas. Papua, the hub of the food estates, saw the largest increase of any region.^[3] Oil palm clearing in Papua also doubled, from about 3,510 hectares in 2024 to 7,333 hectares in 2025, the highest level since 2018.^[19]

In Merauke itself, researchers at the University of Maryland found that 5,315 hectares of forest were cleared inside sugarcane concessions between June 2024 and June 2025, releasing an estimated 4.8 million tonnes of carbon dioxide equivalent. Meeting the national bioethanol target would require converting 1.1 to 1.5 million hectares and release between about 630 and 835 million tonnes.^[7] An earlier estimate by the Indonesian think tank CELIOS put emissions from clearing 2 million hectares for sugarcane at 782 million tonnes, and warned it could push back Indonesia’s net-zero target by five to ten years.^[6] For comparison, that is more than many entire countries emit in a year.

5What is lost when forest becomes plantation

Indonesia’s earlier palm oil boom, centred on Sumatra and Borneo, shows what happens to wildlife when lowland forest becomes plantation.

Orangutans. A landmark study estimated that Borneo lost about 148,500 orangutans between 1999 and 2015, roughly half the population. The steepest declines were where forest was cleared for industrial agriculture, though hunting killed many orangutans in forests that were still standing. The authors projected that another 45,000 could be lost to habitat loss alone in the following decades.^[8]

Tigers. Only a few hundred Sumatran tigers survive in the wild. A study of Sumatra’s forest landscapes found that the island’s potential tiger population fell by 16.6% between 2000 and 2012 as forests were cleared, mainly for plantations, and that only two national parks hold populations large enough to be secure over the long term. The main threat has shifted from poaching to habitat loss.^[21]

Elephants. The Sumatran elephant was moved to Critically Endangered in 2012 after losing roughly 70% of its habitat and about half its population in a single generation. The lowland forest it depends on is exactly the land most sought after for oil palm.^[22]

New Guinea now faces the same pressure, with less known about what is at stake. With 68% of its plant species endemic,^[20] clearing in Papua risks losing species before they are ever described. A plantation is not a degraded forest; it is a different ecosystem, one with a single tree species planted in rows, regular spraying, and little room for the animals and plants that lived there before.

6Forest loss and disaster risk: the Sumatra floods

Indonesia’s deadliest recent disaster offers a test of how forest loss and hazard interact.

In late November 2025, extreme rainfall linked to Tropical Cyclone Senyar triggered floods and landslides across Aceh, North Sumatra and West Sumatra. By early December more than 800 people had died.^[24] The environmental group WALHI said around 1.4 million hectares of forest had been lost in the three provinces between 2016 and 2025, much of it inside company permits, and the forestry minister acknowledged to lawmakers that forest areas there had been converted to other uses.^[25] Analyses of the disaster traced its severity to decades of deforestation in the upper watersheds.^[23] Attribution is still debated, and the storm itself was exceptional. But the lesson for risk managers is consistent: forest loss in a catchment raises the cost of the next extreme event.

7Papua New Guinea: the other half of the island

Across the border, Papua New Guinea faces a different version of the same pressure. In the 2000s the government issued Special Agricultural and Business Leases covering over two million hectares, often in the name of oil palm development. A study of these leases concluded that most developers were clearing forest with no intention of planting oil palm: it was a land grab, and a way to export logs, under the guise of agriculture.^[26] Global Witness reported that the leases covered roughly 12% of the country’s land area, and that permits to log and clear forest under them continued to be issued even after the government decided to repeal the mechanism.^[27]

The problem has not gone away. A 2023 moratorium on new forest clearing authorities has had little effect on the timber trade, and Malaysian companies dominate the clearance permits that remain, according to a 2025 report.^[28] Papua New Guinea lost about 1.0 million hectares of humid primary forest between 2002 and 2025, and 66,000 hectares of natural forest in 2025 alone.^[29] Around 90% of its timber exports go to China.^[30]

8Implications for disaster risk management

Analysis. This section is Disaster Science’s interpretation of the evidence above. It is not drawn from the cited sources.

Treat standing forest as risk infrastructure. Forests hold soil on slopes, slow rainfall, store floodwater and keep peatlands wet enough not to burn. When they are cleared, the risk does not disappear; it moves downstream to the people who live there. Risk assessments for any land conversion should price the flood, fire and landslide protection that is lost, not only the crop that is gained.

Watch the water in Merauke. Converting wetland and swamp forest into rice and sugarcane requires draining and reshaping the land, which changes how water moves across a low-lying, flood-prone landscape. Large monocultures are also exposed to drought, pests and price shocks. Hydrological and flood-risk studies should precede clearing, not follow it.

Separate the regulator from the planter. When the state both enforces forest law and owns millions of hectares of plantations, enforcement and revenue pull in different directions. Independent monitoring and public reporting on what happens to seized land would show whether forest zones are restored or simply change owners.

Use degraded land first. Food and energy goals can be pursued on idle and degraded land, and through higher yields on existing plantations, before primary forest is opened. The order matters: a degraded field can be improved, but a cleared primary forest, and the species found only there, cannot be brought back.

Plan for failure. Earlier food estates in Indonesia, including Merauke’s first attempt, did not deliver what was promised. Exit plans, restoration bonds and clear performance milestones would limit the damage if the new estate falters.

Make consent part of risk reduction. Communities displaced from forests lose the food, water and income that make them resilient, and gain exposure to floods, fires and price shocks. Free, prior and informed consent, and a working grievance process, are not only rights protections; they reduce the social and security risks that stall large projects.

Recognize the role of outside demand. Palm oil is in a vast range of packaged foods and cosmetics, and in a growing share of fuel. Buyers, lenders and governments outside Indonesia shape the demand that makes conversion profitable, and they can ask where their supply comes from.

9Uncertainty and limitations

Figures disagree. Government deforestation figures are lower than independent counts, and the forestry ministry argues that temporary clearing and tree plantations inflate them.^[3] Reporting on the task force’s land totals also varies by date and agency.^[5][13]

The scale keeps changing. Officials have given conflicting figures for the size of the Merauke estate, and much of it has not yet been cleared.^[1] Its final footprint is unknown.

Emissions are projections. The 782 million tonne estimate assumes 2 million hectares of sugarcane is cleared.^[6] The University of Maryland range depends on bioethanol targets, while observed clearing so far is far smaller.^[7]

Attribution of the Sumatra floods is contested. The November 2025 rainfall was exceptional, and experts disagree on how much forest loss added to the death toll.^[23][25]

The policy goals are real. Indonesia imports sugar and diesel, and food and energy security are legitimate priorities for a country of 280 million people. This brief questions the means, not the goals.

Scope. This brief synthesizes peer-reviewed studies, independent monitoring, investigative journalism and official statements. It presents no original data and is an editorial synthesis that has not undergone peer review. Much of the reporting is recent, and figures may change as new data are published.