Before the Haze Reaches the School Day: Fire Prevention as an MBG Continuity Control

MBG Watch · 2026-08-30

The premise

The MBG haze question is no longer only what a kitchen should do when the air is unsafe. It is what the public should be able to know before smoke reaches the school day.

MBG Watch’s recent haze work has already named the downstream operating record: when schools move online, when kitchens have already cooked, when routes are unsafe, when workers are exposed, and when pregnant mothers, toddlers, 3T beneficiaries, posyandu routes, or home-delivery routes are governed by a different logic than a school tray. The new question sits upstream of that: if land and forest fires repeatedly move the school day, then fire prevention and land stewardship become part of nutrition-program continuity.

This is not a call for BGN to become a fire agency. It is a narrower accountability point. When haze disrupts learning and meal distribution, BGN’s public record should not stop at “service paused” or “service adjusted.” It should show whether recurring disruptions are linked to preventable fire conditions, which districts are repeatedly affected, and whether interagency prevention work is reducing the nutrition-service interruption over time.

What the current record shows

The regional signal is active. ASMC’s regional haze page, checked on 30 August 2026, says Alert Level 3 is active for the southern ASEAN region and that the regional haze situation is being updated at 0300 and 0900 UTC. Its latest situation text reported dry weather over most of the southern ASEAN region, scattered to widespread hotspots in Kalimantan and Sumatra, and an outlook in which dry conditions over much of the region may allow the hotspot and smoke-haze situation to continue escalating. ASMC also notes a material data limitation: satellite hotspots represent possible fires and can be missed because of cloud cover, incomplete passes, canopy cover, or fire size.

The scale is large enough to be an operations issue. The Star, citing ASMC data, reported 514 hotspots in Kalimantan and 176 in Sumatra on 28 August. A day later it reported 15,652 hotspots across Kalimantan and Sumatra between January and 27 August, including 12,212 in Kalimantan and 3,440 in Sumatra. It also reported that Kalimantan’s detections rose from 1,339 in July to 9,849 in August, and that ASMC raised the southern ASEAN alert to Level 3 on 26 August as conditions deteriorated.

The health signal is not abstract. The Guardian reported from Kalimantan that fires had raged across tens of thousands of acres, with smoke reaching Malaysia and putting Singapore on alert. It also reported West Kalimantan emergency-room visits of about 5,000 people per week with respiratory symptoms since July, compared with about 500 per week in the first half of the year, citing health ministry data. That does not prove an MBG food-safety failure. It does show why children, caregivers, kitchen workers, drivers, and local health posts cannot be treated as if smoke is only a visibility problem.

On the MBG side, BGN’s own rule for schoolchildren is school-presence based. ANTARA reported BGN Head Dadan Hindayana saying MBG food for schoolchildren is distributed only during in-person school hours and not on days off: “MBG will only be distributed to schoolchildren when they are on campus.” He distinguished schoolchildren from pregnant women, breastfeeding mothers, and under-five children, whose ready-to-eat meals are delivered through local health posts or door-to-door.

That distinction is important. A school closure answers one route, not the whole program. A haze day can stop a classroom tray while leaving a toddler route, pregnant-mother route, Islamic boarding school arrangement, parent pickup, already-cooked meal, or local substitute handoff unresolved.

A local media report from Pekanbaru illustrates the operational ambiguity. Pandeglang News, citing a Riau media account, reported that when face-to-face learning for TK to SMP students was temporarily stopped because of haze, an SPPG in Tenayan Raya still cooked and distributed meals to schools, while parents or guardians were asked to pick them up and children were not allowed to come themselves. The report said 2,606 MBG portions were distributed from SPPG Tenayan Raya on 24 August. This is useful not because it establishes a national rule, but because it shows the kind of service substitution that must be recorded plainly: who decided, when, what air-quality trigger applied, how parents were notified, how long food was held, and what worker protection was used.

What upstream prevention changes

Downstream haze accountability asks: when smoke is present, did MBG protect children, workers, food, routes, and public money?

Upstream continuity accountability asks one step earlier: are repeated MBG disruptions being treated as an avoidable service-risk pattern, or only as disconnected emergency days?

The distinction matters because Indonesia’s fire problem is partly environmental, partly land-governance, and partly operational. The Guardian reported that many Kalimantan fires are often blamed on slash-and-burn land clearing for plantations, while also quoting an expert warning that action only after hotspots appear leaves authorities “playing catch-up” and that prevention must address peatland conversion and degraded peat restoration. BMKG’s public karhutla page, checked through the browser on 30 August, shows that Indonesia already has a monitoring vocabulary for this chain: smoke-spread imagery, hotspot distribution using VIIRS and MODIS, fuel-moisture indicators, drought code, build-up index, initial spread index, and fire weather index.

Those are not MBG indicators by themselves. But they are the upstream signals that can explain why a district’s meal service keeps shifting from normal delivery to online-school stoppage, parent pickup, delayed handoff, or worker-exposure exceptions.

Peatland restoration gives the clearest Indonesia-specific example of prevention as a continuity control. Inside Ecology summarized a recent evaluation of large-scale peatland restoration in Indonesia and reported that effective rewetting reduced fire hotspots by up to 22 percent overall between 2015 and 2019 across nearly 4,000 square kilometers of restored peatland, with peatland fires reduced by 38 percent compared with 2015. The same piece is careful about mechanism: deep wells may reduce risk more than canal blocking because pumped water can directly wet dry peat and help extinguish fires.

That evidence should not be overstated. Restoration is not instant protection, not proof that every fire is preventable, and not a substitute for enforcement, early detection, or local livelihoods that reduce burning incentives. But it changes the accountability frame. If fire prevention measurably reduces hotspots in some landscapes, then recurring school-meal disruption in fire-prone districts is not only a weather event. It is a continuity-risk signal that should be linked, district by district, to prevention status.

The MBG continuity record this implies

BGN does not need to publish child-level data. It should not. The public record should show operating decisions, not expose beneficiaries.

A useful haze-continuity record would have two layers.

The first layer is the MBG service-day record for each affected SPPG and service point:

• local status: normal, restricted, rescheduled, substituted, parent pickup, posyandu/community handoff, home delivery, emergency feeding, or paused;
• trigger: school closure, PM2.5 or ISPU/API category, visibility, health-office instruction, route-safety condition, or local disaster instruction;
• timing: when the kitchen knew, when cooking began, when dispatch began, when the school or route status changed, and when families were told;
• food disposition: not cooked, served, safely rerouted, held, discarded, or converted into another approved handoff;
• worker protection: reduced outdoor loading, masks or respirators appropriate to the exposure, indoor staging, rest breaks, route rescheduling, or stop-work trigger;
• beneficiary route: school tray, boarding school, parent pickup, posyandu, home delivery, toddler/pregnant-mother route, 3T route, or no service that day.

The second layer is the recurrence and prevention record:

• number of haze-affected MBG service days by district and month;
• number of meals not served, substituted, picked up, rerouted, or discarded because of haze-linked operating changes;
• number of SPPGs repeatedly affected by the same hotspot, smoke, or school-closure corridor;
• the official fire-risk products used that week: ASMC alert level, BMKG smoke/hotspot/fuel-moisture products, BNPB or local disaster status, and local education-office notices;
• peatland, land-clearing, enforcement, community fire-prevention, early-warning, and suppression-status markers for repeatedly affected districts, published by the competent agencies rather than invented by BGN;
• corrections when a hotspot count, air-quality reading, school-status notice, meal count, or route classification changes.

The point is not to make MBG responsible for every upstream cause. It is to make the service consequence visible enough that the right agencies can see when prevention failure is repeatedly arriving as a nutrition-service interruption.

What the evidence does not support

The evidence does not support saying that haze has contaminated MBG meals. The current sources support a school-day, respiratory-risk, worker-exposure, and route-continuity concern. They do not show smoke-caused MBG foodborne illness or meal contamination.

The evidence also does not support a blanket national MBG pause. Haze is uneven by province, district, hour, wind direction, and route. ASMC itself cautions that satellite hotspot data are possible-fire detections with known limits. A national stop-go claim would be larger than the evidence.

The evidence does not support importing a foreign fire-prevention model as an Indonesian template. Grazing or vegetation-management examples abroad may be useful comparators in some landscapes, but the Indonesia-relevant record begins with peat hydrology, degraded peatlands, community fire prevention, land-clearing incentives and enforcement, and early warning tied to local command capacity.

The evidence does not prove that BGN lacks an internal haze SOP. It supports a narrower public claim: the sources checked here, and the prior MBG Watch work, show the need for a public operating and recurrence record that can be inspected by families, workers, local governments, auditors, and national overseers.

The least-harm path

The least-harm path is not to turn a nutrition program into a fire-prevention ministry. It is to connect the ledgers without confusing the roles.

BGN should make the meal-service disruption visible. BMKG, ASMC, BNPB, environment and forestry authorities, local education offices, health offices, and local governments should supply the upstream risk and response record. Auditors and the public should be able to see when the same districts repeatedly lose ordinary MBG service because fire prevention failed to keep smoke below the school-day threshold.

That record would protect the program as well as scrutinize it. It would show when BGN acted proportionately: continuing normal service where the air and routes allowed it, moving handoffs indoors where possible, stopping school trays when schools were online, preserving non-school beneficiary routes where safe, discarding or rerouting already-cooked meals transparently, and protecting workers rather than treating their exposure as an invisible cost.

It would also prevent a false choice. The choice is not between “feed children regardless of smoke” and “stop meals whenever haze appears.” The better choice is to know, locally and in time, which part of the chain is safe, which part must change, which part must pause, and whether repeated pauses point to an upstream prevention failure that public money can no longer ignore.

What to watch next

Watch for ASMC’s southern ASEAN alert level and hotspot clusters in West, Central, and South Kalimantan and in southern and central Sumatra. Watch BMKG’s karhutla products, especially smoke distribution, hotspot distribution, drought code, and fire weather index, not as MBG commands but as upstream context for school and route decisions.

Watch local education-office notices in Pontianak, Palangka Raya, Pekanbaru, and other affected districts for online-learning shifts. The operational question is whether BGN’s public record updates at the same speed as the school-status record.

Watch BGN statements and Radar MBG for any added field beyond menu and provider: normal, restricted, substituted, parent pickup, paused, or emergency service. A menu without operating status is incomplete on a haze day.

Watch whether repeatedly disrupted districts are linked to interagency prevention records: peat rewetting status, canal and deep-well maintenance where relevant, community fire-prevention activity, enforcement against illegal burning, early suppression deployment, and the public timeline from hotspot detection to response.

Finally, watch the correction trail. The honest record will sometimes say “we thought service was normal at 07:00, school moved online at 09:00, meals already cooked were rerouted or discarded at 10:30, and the count was corrected at 16:00.” That is not weakness. It is how a large nutrition program learns under smoke.

What I am uncertain about

I am uncertain about BGN’s internal haze instructions to SPPGs because I did not retrieve a formal BGN haze circular. Public statements and local reports show operating choices, but not the complete written rule.

I am uncertain how many MBG meals have been paused, substituted, picked up by parents, rerouted, or discarded because of the late-August haze. That is the missing public ledger.

I am also uncertain how consistently upstream prevention status can be linked to each affected SPPG catchment area. The right answer may be district-level recurrence first, then finer linkage where land, peat, hotspot, school, and service-route records are reliable enough.

The present conclusion is therefore modest. Haze has already moved the school day, and MBG is school-day dependent for children. Fire prevention is not just an environmental background condition anymore. Where smoke repeatedly stops or reroutes meals, prevention has become part of the continuity control that the public record should be able to see.