Stand up a Bangladesh national heat action plan before the next April-May heat-index 40+ season

Diagnosis

Heatwaves in Bangladesh are a fast-onset, tier-1 climate hazard concentrated in April and May, when the heat index climbs past 40 (per the curated note). This is no longer a tail risk: the note records that both 2023 and 2024 broke records, meaning two consecutive seasons set new extremes. That pattern matters now because the exposure is recurring and predictable in timing (Apr-May) even though severity is rising, which is exactly the profile a standing operational plan is built for.

The binding gap is institutional, not meteorological. There is a clear lead body, the Ministry of Disaster Management and Relief (MoDMR), but no current_state indicator is being tracked (it is null, and the data_status flags that a collector is still needed). A tier-1 hazard with a named owner but no live nowcast and no triggered response protocol means the system reacts after people are already harmed rather than ahead of the heat. The horizon here is a nowcast: the value of action comes from acting in the days and hours before and during an event, not from long-run projections.

Recommended actions

1. Declare heatwave a notified disaster with a standing national heat action plan. Owner: MoDMR. Mechanism: issue a Standing Order on Disaster (SOD) amendment that classifies heatwave alongside flood and cyclone, assigning roles to the Department of Disaster Management, city corporations, and district administrations. Observable signal: a published, dated heat action plan in force before the Apr-May window, with named focal officers at district level.
2. Set heat-index trigger thresholds that activate response automatically. Owner: MoDMR, using Bangladesh Meteorological Department feeds. Mechanism: a published circular tying graduated actions (advisory, warning, emergency) to heat-index 40+ and higher bands, so an alert mechanically opens cooling centres and shifts outdoor work hours. Observable signal: each season's first 40+ reading triggers a logged, time-stamped activation rather than an ad hoc decision.
3. Build the missing nowcast collector and a public alert channel. Owner: MoDMR (data) with the warning disseminated through existing disaster SMS and mosque/loudspeaker networks. Mechanism: the data_status "needs_collector" item becomes a live heat-index feed that populates a current_state indicator and pushes alerts. Observable signal: a populated current_state value replacing the present null, and alerts reaching target districts within hours of threshold breach.
4. Pre-position cooling and water relief through the existing relief machinery. Owner: MoDMR via district relief allocations and city corporations. Mechanism: designate cooling shelters (schools, community centres), water points, and ORS distribution funded from the disaster relief budget line, activated by the thresholds in action 2. Observable signal: a pre-season register of designated shelters and stocked supplies per district.
5. Issue an outdoor-work and school heat protocol. Owner: MoDMR coordinating with labour and education authorities. Mechanism: an advisory framework rescheduling outdoor labour and school hours during emergency-band days. Observable signal: documented hour adjustments triggered on emergency days, not discretionary closures.

Sequencing (first 12 months)

First, before the next Apr-May window, MoDMR issues the SOD amendment (action 1) and the threshold circular (action 2): these are administrative, low-cost, and unlock everything downstream because response cannot be automatic without defined triggers and assigned roles. In parallel, commission the nowcast collector (action 3) so the current_state indicator exists by the time thresholds matter. With triggers and data in place, pre-position relief (action 4) and publish the work-and-school protocol (action 5) ahead of April. The first season then becomes a live test, after which MoDMR reviews activation logs and refines thresholds before the following year.

Risks and constraints

The binding constraints are fiscal and coordination, not technical. Cooling shelters, water, and ORS draw on a relief budget that competes with flood and cyclone demands, so heat must be funded as a named line rather than reallocated mid-crisis. Coordination across MoDMR, the met department, city corporations, and labour and education authorities is the classic failure point: without the SOD amendment giving clear authority, alerts will not convert into action on the ground. Finally, the data gap (null current_state, collector pending) means the alert system is the long pole; if it slips, the thresholds in action 2 cannot fire automatically.

Bottom line

Bangladesh faces a recurring, record-breaking Apr-May heat hazard with a clear owner in MoDMR but no live nowcast and no triggered response protocol. The decisive move is administrative and cheap: classify heatwave under the Standing Order on Disaster, publish heat-index thresholds that automatically open relief, and build the missing collector before the next season so response leads the heat instead of trailing it.