The Meteorology, Climatology, and Geophysics Agency (BMKG) has predicted that 2026 will be the most severe drought year to hit Indonesia since 1950. However, this prediction is not merely a technical projection on a weather forecast sheet. This figure forms the backdrop of a much more fundamental and thought provoking question: if the government already possesses moratorium instruments, designates conservation areas, devises rehabilitation plans, and binds permitted concessions with management obligations, why do forest and land fires (karhutla) continue to occur across all these categories simultaneously?

Based on the Indicative Burned Area (AIT) modeling data compiled by Madani Berkelanjutan, the accumulated fires throughout the January–May 2026 period recorded an alarming figure of 88 thousand hectares. Out of this total area, 82.5 thousand hectares constitute non-recurring areas. This figure provides a solid foundation for analyzing layer by layer the effectiveness of the land protection instruments we currently have.

Moratoria and Peatland Ecosystems Breached by Fire

The Indicative Map for the Moratorium on the Issuance of New Permits (PIPPIB) policy is essentially the government's primary instrument to halt new permit issuances in primary peatlands and natural forests. In reality, however, the existence of this protection status has been unable to prevent fire from breaching areas that legally should not be exploited.

• A Moratorium Breached: Out of the total 82.5 thousand hectares of burned area, 43.31% or equivalent to 35.7 thousand hectares was actually located within the PIPPIB permit moratorium area.

• Massive Emission Releases: The largest portion of the burned moratorium area was within the Peatland PIPPIB zone, reaching 26.23 thousand hectares. Peatlands are not only fragile ecosystems but also colossal carbon reservoirs. When peat burns, the emissions released into the atmosphere far exceed the proportion of the burned land area.

AIT data also shows a highly ironic pattern: 60% of the total fires (equivalent to 49.56 thousand hectares) were concentrated within Peatland Ecosystem Functions (FEG). The most extensively affected area was actually the peatland ecosystem protection function, reaching 35.81 thousand hectares, while the peatland ecosystem cultivation function recorded 13.7 thousand hectares during the January–March 2026 period.

Inverting the Baseline Assumption: This pattern completely upends long standing land management assumptions. Zones designated for protection functions rather than production functions actually experienced more extensive fires. This is a strong indication that peatland restoration programs, despite running for years, have not yet sufficiently altered the biophysical conditions on the ground to withstand fires when extreme droughts strike.

The same pattern holds true within forest zones, where 52.72% or 43.49 thousand hectares of the total AIT fell inside forest estates. The majority of these fires (33.38 thousand hectares) occurred in production forests. Meanwhile, protection forests recorded a fire peak of 2.2 thousand hectares in March 2026, and conservation forests recorded 2.3 thousand hectares in January 2026. This implies that variations in zonation functions whether production, protection, or conservation do not correlate with distinct differences in actual fire risks on the ground.

Permitted Concessions: Managed Land, Yet Fire Still Roams Free

More than half of the fires that occurred, specifically 53.06% or equivalent to 43.7 thousand hectares, overlapped directly with corporate permit and concession areas. Concession areas are not unmanaged empty spaces; each concession has a legal entity fully accountable for its land governance.

Our environmental legal framework ranging from the Plantations Law, the Forestry Law, to their derivative regulations explicitly obligates permit holders to prevent fires within their concession areas. However, AIT data demonstrates that these legal obligations have either not been adequately translated into real prevention capacities on the ground or are simply not being taken seriously.

The following is the breakdown of permit and concession sectors whose areas were impacted by fires from January to May 2026:

The dominance of palm oil permits in these fire figures is not a new finding. The close relationship between plantation expansion and land fires has been clearly documented since the catastrophic fires of 2015. What sets it apart now is the climatic context: the extreme 2026 drought, tracking far above historical averages, places immense additional pressure on lands that are already ecologically vulnerable and poorly monitored at the field level.

The Vicious Cycle of Land Restoration and Threats to Biodiversity

The impacts of this fire season go beyond destroying commodities; they severely set back environmental recovery efforts and high-value conservation zones:

• A Destructive Cycle in Rehabilitation Areas (RURHL): Around 13.01% or roughly 10.7 thousand hectares of the burned areas were located within General Forest and Land Rehabilitation Plan (RURHL) zones. Fire figures in these zones crept upward from 2.63 thousand hectares in January 2026 to 4.1 thousand hectares in March 2026, before dropping to 449 hectares in May 2026. When recovery areas burn, the cycle of degradation spins once more: degraded land burns again before it can even recover, forcing future rehabilitation budgets to recalculate from a significantly worse baseline.

• Threats to Key Biodiversity Areas (KBAs): Approximately 10.94% or around 9 thousand hectares of the fires swept through Key Biodiversity Areas (KBAs). KBAs are not typical administrative boundaries; these territories are defined based on concentrations of endemic and globally endangered species. Fires in KBAs trigger direct, localized disruptions to wildlife habitats that cannot simply be replaced by standard rehabilitation programs.

The Carbon Target Irony: When Sinks Shift to Emission Sources

The data from January to May 2026 shows a distribution pattern far too consistent to be dismissed as a passing seasonal anomaly. The trend is clear: moratorium zones are breached, protection areas suffer larger burns than production areas, permit holders fail to secure their concessions, and lands undergoing recovery are once again scorched.

The most ironic impact can be seen at the operational level of the FOLU Net Sink the flagship program serving as the primary backbone for Indonesia to meet its Nationally Determined Contribution (NDC) emission reduction targets. Notably, 49.86% or equivalent to 41 thousand hectares of the burned areas were located directly within the FOLU Net Sink program's operational blueprint.

Climate Irony: Half of the land area explicitly reserved and managed to sequester carbon instead turned into a source releasing carbon emissions into the atmosphere due to raging fires.

Geographically, West Kalimantan Province emerged as the region with the most extensive fire level throughout this period, reaching 24.8 thousand hectares. Meanwhile, at the regency level, Bengkalis in Riau served as the primary hotspot with an area reaching 11.86 thousand hectares. Both regions have repeatedly appeared in historical fire report track records from previous years. This happens not due to a lack of public attention, but because that attention has not been fully translated into adequate prevention capacity on the ground.

Indonesia inherently possesses numerous policy instruments to protect its landscapes. However, the 2026 fire data proves that the gap between what is written on policy maps and the reality of protection in the field is still far too wide for us to ignore.

(Data Source: Analysis utilizing the Indicative Burned Area (AIT) model by Madani Berkelanjutan for the January–May 2026 period, BMKG, SIPONGI Ministry of Forestry, KLHK, and KBA with a correlation coefficient level between AIT and SIPONGI data of 94.38% and an $R^2$ value reaching 89.07%).