What the rewetting actually moved
Salmayenti, Spracklen and Holden set restored blocks of peatland in Sumatra and Kalimantan against similar unrestored ground and followed both from 2015 to 2019. Across nearly 4,000 square kilometres where rewetting had taken hold, fire hotspots fell by up to 22 per cent; during the 2019 El Niño, peatland fires ran 38 per cent below the 2015 season — the season in which 2.6 million hectares burned and losses passed 16.1 billion dollars.[1]
The mechanism is plain water management: peat does not burn while it stays wet. Canal blocking holds water in place so it does not drain away; it involves no pumping and affects only land within a few hundred metres. A deep well pumps water up during the dry season, wets dry peat directly, reaches a wider area, lasts longer as infrastructure, and the same water can be used to put out fires. Where suitable locations are chosen and enough canals are blocked, ideally in clusters, canal blocking can be more effective than deep wells.[1]
The number that governs the result
Indonesia had restored around 1.6 million hectares by 2024. That total measures work commissioned; the current condition of the hardware is a separate figure, and it reads worse. A recent survey found 70 per cent of canal-blocking units damaged, and the authors note that only blocks around a year old genuinely keep peat wet. In restored ground the water table still sits below the 40 centimetre depth treated as safe, and hotspot density there runs up to ten times that of pristine peat forest. Restoration bounds the fire risk on degraded peat and leaves it short of the condition of intact forest.[1]
There is a reading of that gap that does not run through maintenance. The two methods differ in reach and durability, so a programme weighted toward canal blocking in poorly chosen sites would underperform even with every unit intact. The siting and clustering condition the authors describe supports that second reading; the age of the blocks supports the first. Whichever dominates, the ledger comes out in the same place: restored hectares are an input, the water table is the output, and the output is what puts out a fire.[1]
Two numbers to watch as the El Niño peaks
A strong El Niño is due to peak later in 2026. Indonesia had already recorded more than 100,000 hectares burned between January and June, and burned area inside peat forest regions has stayed consistently above 500,000 hectares. The restoration agency's mandate ended in 2024 and its work was distributed across several ministries, which is where responsibility for repairing canal blocks now sits. If the maintenance backlog is not cleared before the peak, expect hotspot density in restored blocks to converge on the unrestored comparison, and expect peat-region burned area to hold above 500,000 hectares through the 2026-2027 season. Both numbers are published often enough to settle the question without waiting for the next paper.[1]