{"finding":{"findingId":"01a0f0e8-cc72-747e-b91d-e7658b36555d","status":"published","tier":1,"rule":{"name":"forest_loss","version":"1.0","params":{"days":90,"minAlerts":1530,"minConfidence":"high","minHa":153,"minValidPct":60,"ndviDrop":0.1}},"title":"Forest loss, Kazumba, Kasaï-Central (Democratic Republic of the Congo): 641.47 ha in 90 d","summary":"52415 GFW integrated deforestation alerts (≥ high confidence) covering 641.47 ha between 2026-07-02 and 2026-09-30; AOI alert density is 0.957× its 25 km neighbourhood. Awaiting optical confirmation.","geometry":{"type":"Polygon","coordinates":[[[21.4,-7.1],[22.49,-7.1],[22.49,-5.7],[21.4,-5.7],[21.4,-7.1]]]},"bbox":[21.4,-7.1,22.49,-5.7],"aoi":{"id":"cod-7-5","name":"Kazumba, Kasaï-Central (Democratic Republic of the Congo)","tags":["forest-hotspot","discovered","cod","democratic-republic-of-the-congo","kasai-central","kazumba","protected-or-indigenous"]},"context":{"enso":{"phase":"El Niño","oni":1.8},"events":[],"baseline":{"metric":"alert ha per deg²","ringKm":25,"aoiValue":420.36,"regionalValue":439.134,"ratio":0.957},"notes":["Living value (benefit transfer, order of magnitude): 641.47 ha of tropical forest ≈ $4,309,395/yr alive (band $1,265,620–$16,695,540, 2020 USD), ≈ $430,939,500 over 100 yr ($185,727,821 NPV @ 2 %). Source: Costanza R. et al. et al. — a floor for one kind of value, not a price."]},"blindSpots":["Coverage is 30°N–30°S only — zero alerts outside the tropics means 'not monitored', not 'no loss'.","Cannot distinguish legal clearing, fire, storm damage and selective logging from illegal clearing; attribution needs concession/protected-area layers (M3).","Alerts lag reality by ~1–2 weeks; a burst just before the sweep is undercounted.","Regrowth and plantation harvest cycles look like loss; the baseline ring only partly compensates.","Persistent cloud can delay optical confirmation for months (SAR-only RADD alerts stay 'candidate')."],"observedAt":"2026-09-30T00:00:00Z","createdAt":"2026-09-30T06:00:59.123Z","updatedAt":"2026-10-01T06:01:01.847Z","createdBy":"system:forest_loss@1.0","confirmed":{"at":"2026-10-01T00:01:01.398Z","independence":"sensor","signal":{"id":"s2-ndvi-median-cod-7-5-2026-05-18..2026-09-30","kind":"scene","source":"sentinel-2-l2a","datetime":"2026-09-30T00:00:00Z","method":{"name":"eo_compare","version":"1.0","params":{"bbox":[21.4,-7.1,22.49,-5.7],"composite":"median","dateA":"2026-05-18","dateB":"2026-09-30","index":"NDVI","width":256}},"summary":"Median-composite NDVI changed -0.123 (2026-05-18 → 2026-09-30); valid pixels 99% / 100%. Scenes: [object Object], [object Object], [object Object], [object Object], [object Object], [object Object]","values":{"deltaNdvi":-0.12293023357901012,"validPctA":99,"validPctB":100}}},"evidence":[{"id":"gfw-integrated-cod-7-5-2026-07-02..2026-09-30","kind":"alert","source":"gfw-integrated-alerts","datetime":"2026-09-30T00:00:00Z","href":"https://data-api.globalforestwatch.org/dataset/gfw_integrated_alerts","method":{"name":"forest_alerts","version":"1.0","params":{"bbox":[21.4,-7.1,22.49,-5.7],"days":90,"minConfidence":"high"}},"summary":"52415 alerts (≥ high confidence), 641.47 ha, 2026-07-02…2026-09-30.","values":{"alerts":52415,"ha":641.47,"high_alerts":52187,"high_ha":638.68,"highest_alerts":228,"highest_ha":2.79,"living_value_100y_npv2_usd":185727821,"living_value_100y_usd":430939500,"living_value_usd_yr":4309395}},{"id":"gfw-integrated-cod-7-5-2026-07-03..2026-10-01","kind":"alert","source":"gfw-integrated-alerts","datetime":"2026-10-01T00:00:00Z","href":"https://data-api.globalforestwatch.org/dataset/gfw_integrated_alerts","method":{"name":"forest_alerts","version":"1.0","params":{"bbox":[21.4,-7.1,22.49,-5.7],"days":90,"minConfidence":"high"}},"summary":"119338 alerts (≥ high confidence), 1460.72 ha, 2026-07-03…2026-10-01.","values":{"alerts":119338,"ha":1460.72,"high_alerts":118993,"high_ha":1456.5,"highest_alerts":345,"highest_ha":4.22,"living_value_100y_npv2_usd":422929167,"living_value_100y_usd":981311700,"living_value_usd_yr":9813117}}],"attribution":null,"narration":{"v":1,"eventId":"01a0f50d-8422-7c45-9b02-16ae0907dc66","findingId":"01a0f0e8-cc72-747e-b91d-e7658b36555d","at":"2026-10-01T01:21:38.850Z","actor":"model:claude-opus-5","prev":"4d27aa78c46ecfeef1d10aaddf2b0cae691d3b93a1964eaa5285f238401a8363","kind":"narrated","text":"641 ha of forest loss flagged in Kazumba, DR Congo, over 90 days — at the regional pace\n\nBetween 2 July and 30 September 2026, satellite alert systems recorded 52,415 high-confidence deforestation alerts across Kazumba, Kasaï-Central, covering 641.47 hectares. An independent optical check using Sentinel-2 imagery supports it: median greenness (NDVI) fell by -0.123 between May and September, with 99% and 100% of pixels usable, so cloud is not hiding the picture. The area carries protected-or-indigenous tags.\n\nWhat this is not: it is not proof of illegal clearing. The alerts cannot tell apart legal clearing, fire, storm damage, selective logging or plantation harvest, and regrowth cycles can look like loss. The loss rate here is 0.957× the surrounding 25 km neighbourhood — slightly below the regional pace, not an outlier. An El Niño phase (ONI 1.8) can dry vegetation and push NDVI down on its own.\n\nWhat would change our mind: concession and protected-area layers showing authorised activity; burn-scar or fire data matching the alert footprint; or follow-up imagery showing green returning, which would point to seasonal stress or regrowth rather than permanent clearing.\n\nKey numbers:\n- High-confidence alerts (90 days): 52415 [gfw-integrated-cod-7-5-2026-07-02..2026-09-30]\n- Hectares covered by alerts: 641.47 [gfw-integrated-cod-7-5-2026-07-02..2026-09-30]\n- Change in median NDVI: -0.12293023357901012 [s2-ndvi-median-cod-7-5-2026-05-18..2026-09-30]\n- Valid pixels, later image (%): 100 [s2-ndvi-median-cod-7-5-2026-05-18..2026-09-30]\n\nConfidence: medium\nCaveats:\n- Alerts lag reality by 1–2 weeks; recent clearing may be undercounted.\n- Cause is unattributed — fire, storm, legal logging and harvest cycles look alike.\n- Alert density is 0.957× the 25 km ring, i.e. slightly below the regional norm.\n- El Niño (ONI 1.8) can depress NDVI independently of clearing.\n- Monetary 'living value' figures are order-of-magnitude benefit transfers, not prices, and are not cited as findings.","model":{"id":"claude-opus-5","provider":"anthropic"},"promptSha256":"794cff33a6e535447b4d017f5912bf3a262d6630bca953452bcabd7f8dd60b80","transcriptSha256":"b516c6349c612db9ed8e4d66ae4e14cebb12311430686403e45b90c1f154c269","evidenceRefs":["gfw-integrated-cod-7-5-2026-07-02..2026-09-30","s2-ndvi-median-cod-7-5-2026-05-18..2026-09-30"]},"reviews":[{"v":1,"eventId":"01a0f50d-bca3-78ca-a145-129a6f1a2cdd","findingId":"01a0f0e8-cc72-747e-b91d-e7658b36555d","at":"2026-10-01T01:21:53.315Z","actor":"model:claude-sonnet-5","prev":"67b445470818fbe453c1c284773ebbf8559eadd2568fb2cf739802a0292d17e8","kind":"reviewed","decision":"approve","tier":1,"note":"All key numbers (52415 alerts, 641.47 ha, NDVI delta -0.123, valid pct 99/100, baseline ratio 0.957) match the evidence exactly.; Narration correctly notes the AOI's alert density is slightly below rather than above the regional norm, avoiding overclaiming.; Blind spots are honestly surfaced: unattributed cause (fire/storm/logging/harvest), El Niño confound, alert lag, and the caveat that monetary 'living value' figures are not prices.; No individuals are named; AOI is not a control AOI, consistent with publishing a genuine finding.","verdict":"publish"}],"notifications":[],"replies":[],"retracted":null,"history":[{"at":"2026-09-30T06:00:59.123Z","kind":"created","status":"candidate","actor":"system:forest_loss@1.0"},{"at":"2026-10-01T00:01:01.398Z","kind":"confirmed","status":"confirmed","actor":"system:forest_loss@1.0"},{"at":"2026-10-01T01:21:38.850Z","kind":"narrated","status":"confirmed","actor":"model:claude-opus-5"},{"at":"2026-10-01T01:21:53.315Z","kind":"reviewed","status":"confirmed","actor":"model:claude-sonnet-5"},{"at":"2026-10-01T01:21:53.332Z","kind":"status_changed","status":"published","actor":"model:claude-sonnet-5"},{"at":"2026-10-01T06:01:01.847Z","kind":"evidence_added","status":"published","actor":"system:forest_loss@1.0"}],"eventCount":6,"lastEventHash":"91a69ebbd2bc0a61529fce442b34632945dd194808082b01e2fc7fc5e47cc08f"},"events":[{"v":1,"eventId":"01a0f0e8-cc72-7996-8ab9-7eabf5a81440","findingId":"01a0f0e8-cc72-747e-b91d-e7658b36555d","at":"2026-09-30T06:00:59.123Z","actor":"system:forest_loss@1.0","prev":null,"kind":"created","rule":{"name":"forest_loss","version":"1.0","params":{"days":90,"minAlerts":1530,"minConfidence":"high","minHa":153,"minValidPct":60,"ndviDrop":0.1}},"title":"Forest loss, Kazumba, Kasaï-Central (Democratic Republic of the Congo): 641.47 ha in 90 d","summary":"52415 GFW integrated deforestation alerts (≥ high confidence) covering 641.47 ha between 2026-07-02 and 2026-09-30; AOI alert density is 0.957× its 25 km neighbourhood. Awaiting optical confirmation.","tier":1,"geometry":{"type":"Polygon","coordinates":[[[21.4,-7.1],[22.49,-7.1],[22.49,-5.7],[21.4,-5.7],[21.4,-7.1]]]},"bbox":[21.4,-7.1,22.49,-5.7],"aoi":{"id":"cod-7-5","name":"Kazumba, Kasaï-Central (Democratic Republic of the Congo)","tags":["forest-hotspot","discovered","cod","democratic-republic-of-the-congo","kasai-central","kazumba","protected-or-indigenous"]},"observedAt":"2026-09-30T00:00:00Z","evidence":[{"id":"gfw-integrated-cod-7-5-2026-07-02..2026-09-30","kind":"alert","source":"gfw-integrated-alerts","datetime":"2026-09-30T00:00:00Z","href":"https://data-api.globalforestwatch.org/dataset/gfw_integrated_alerts","method":{"name":"forest_alerts","version":"1.0","params":{"bbox":[21.4,-7.1,22.49,-5.7],"days":90,"minConfidence":"high"}},"summary":"52415 alerts (≥ high confidence), 641.47 ha, 2026-07-02…2026-09-30.","values":{"alerts":52415,"ha":641.47,"high_alerts":52187,"high_ha":638.68,"highest_alerts":228,"highest_ha":2.79,"living_value_100y_npv2_usd":185727821,"living_value_100y_usd":430939500,"living_value_usd_yr":4309395}}],"context":{"enso":{"phase":"El Niño","oni":1.8},"events":[],"baseline":{"metric":"alert ha per deg²","ringKm":25,"aoiValue":420.36,"regionalValue":439.134,"ratio":0.957},"notes":["Living value (benefit transfer, order of magnitude): 641.47 ha of tropical forest ≈ $4,309,395/yr alive (band $1,265,620–$16,695,540, 2020 USD), ≈ $430,939,500 over 100 yr ($185,727,821 NPV @ 2 %). Source: Costanza R. et al. et al. — a floor for one kind of value, not a price."]},"blindSpots":["Coverage is 30°N–30°S only — zero alerts outside the tropics means 'not monitored', not 'no loss'.","Cannot distinguish legal clearing, fire, storm damage and selective logging from illegal clearing; attribution needs concession/protected-area layers (M3).","Alerts lag reality by ~1–2 weeks; a burst just before the sweep is undercounted.","Regrowth and plantation harvest cycles look like loss; the baseline ring only partly compensates.","Persistent cloud can delay optical confirmation for months (SAR-only RADD alerts stay 'candidate')."]},{"v":1,"eventId":"01a0f4c8-9dec-76c2-8063-e221236a1947","findingId":"01a0f0e8-cc72-747e-b91d-e7658b36555d","at":"2026-10-01T00:01:01.398Z","actor":"system:forest_loss@1.0","prev":"1b519fe4eb5016776724c417dba7f477d92a928ce13e3e84c78da47fe27a610d","kind":"confirmed","signal":{"id":"s2-ndvi-median-cod-7-5-2026-05-18..2026-09-30","kind":"scene","source":"sentinel-2-l2a","datetime":"2026-09-30T00:00:00Z","method":{"name":"eo_compare","version":"1.0","params":{"bbox":[21.4,-7.1,22.49,-5.7],"composite":"median","dateA":"2026-05-18","dateB":"2026-09-30","index":"NDVI","width":256}},"summary":"Median-composite NDVI changed -0.123 (2026-05-18 → 2026-09-30); valid pixels 99% / 100%. Scenes: [object Object], [object Object], [object Object], [object Object], [object Object], [object Object]","values":{"deltaNdvi":-0.12293023357901012,"validPctA":99,"validPctB":100}},"independence":"sensor"},{"v":1,"eventId":"01a0f50d-8422-7c45-9b02-16ae0907dc66","findingId":"01a0f0e8-cc72-747e-b91d-e7658b36555d","at":"2026-10-01T01:21:38.850Z","actor":"model:claude-opus-5","prev":"4d27aa78c46ecfeef1d10aaddf2b0cae691d3b93a1964eaa5285f238401a8363","kind":"narrated","text":"641 ha of forest loss flagged in Kazumba, DR Congo, over 90 days — at the regional pace\n\nBetween 2 July and 30 September 2026, satellite alert systems recorded 52,415 high-confidence deforestation alerts across Kazumba, Kasaï-Central, covering 641.47 hectares. An independent optical check using Sentinel-2 imagery supports it: median greenness (NDVI) fell by -0.123 between May and September, with 99% and 100% of pixels usable, so cloud is not hiding the picture. The area carries protected-or-indigenous tags.\n\nWhat this is not: it is not proof of illegal clearing. The alerts cannot tell apart legal clearing, fire, storm damage, selective logging or plantation harvest, and regrowth cycles can look like loss. The loss rate here is 0.957× the surrounding 25 km neighbourhood — slightly below the regional pace, not an outlier. An El Niño phase (ONI 1.8) can dry vegetation and push NDVI down on its own.\n\nWhat would change our mind: concession and protected-area layers showing authorised activity; burn-scar or fire data matching the alert footprint; or follow-up imagery showing green returning, which would point to seasonal stress or regrowth rather than permanent clearing.\n\nKey numbers:\n- High-confidence alerts (90 days): 52415 [gfw-integrated-cod-7-5-2026-07-02..2026-09-30]\n- Hectares covered by alerts: 641.47 [gfw-integrated-cod-7-5-2026-07-02..2026-09-30]\n- Change in median NDVI: -0.12293023357901012 [s2-ndvi-median-cod-7-5-2026-05-18..2026-09-30]\n- Valid pixels, later image (%): 100 [s2-ndvi-median-cod-7-5-2026-05-18..2026-09-30]\n\nConfidence: medium\nCaveats:\n- Alerts lag reality by 1–2 weeks; recent clearing may be undercounted.\n- Cause is unattributed — fire, storm, legal logging and harvest cycles look alike.\n- Alert density is 0.957× the 25 km ring, i.e. slightly below the regional norm.\n- El Niño (ONI 1.8) can depress NDVI independently of clearing.\n- Monetary 'living value' figures are order-of-magnitude benefit transfers, not prices, and are not cited as findings.","model":{"id":"claude-opus-5","provider":"anthropic"},"promptSha256":"794cff33a6e535447b4d017f5912bf3a262d6630bca953452bcabd7f8dd60b80","transcriptSha256":"b516c6349c612db9ed8e4d66ae4e14cebb12311430686403e45b90c1f154c269","evidenceRefs":["gfw-integrated-cod-7-5-2026-07-02..2026-09-30","s2-ndvi-median-cod-7-5-2026-05-18..2026-09-30"]},{"v":1,"eventId":"01a0f50d-bca3-78ca-a145-129a6f1a2cdd","findingId":"01a0f0e8-cc72-747e-b91d-e7658b36555d","at":"2026-10-01T01:21:53.315Z","actor":"model:claude-sonnet-5","prev":"67b445470818fbe453c1c284773ebbf8559eadd2568fb2cf739802a0292d17e8","kind":"reviewed","decision":"approve","tier":1,"note":"All key numbers (52415 alerts, 641.47 ha, NDVI delta -0.123, valid pct 99/100, baseline ratio 0.957) match the evidence exactly.; Narration correctly notes the AOI's alert density is slightly below rather than above the regional norm, avoiding overclaiming.; Blind spots are honestly surfaced: unattributed cause (fire/storm/logging/harvest), El Niño confound, alert lag, and the caveat that monetary 'living value' figures are not prices.; No individuals are named; AOI is not a control AOI, consistent with publishing a genuine finding.","verdict":"publish"},{"v":1,"eventId":"01a0f50d-bcb4-7110-9da9-554172049a82","findingId":"01a0f0e8-cc72-747e-b91d-e7658b36555d","at":"2026-10-01T01:21:53.332Z","actor":"model:claude-sonnet-5","prev":"ad5f57d68a84e410e4a92f29a40af1afcd08c749ff61622f4fcce44d27d2e3a4","kind":"status_changed","from":"confirmed","to":"published","reason":"autonomous publish: narrated by model:claude-opus-5, reviewed by model:claude-sonnet-5 (policy 2026-09-26-autonomous)","gates":{"narratedBy":"model:claude-opus-5","reviewedBy":["model:claude-sonnet-5"],"policy":"2026-09-26-autonomous"}},{"v":1,"eventId":"01a0f60f-1d99-7c1d-b6a9-7f2e941fe37e","findingId":"01a0f0e8-cc72-747e-b91d-e7658b36555d","at":"2026-10-01T06:01:01.847Z","actor":"system:forest_loss@1.0","prev":"6b1d92060044a9dc0cae11815ab579ecbb5f38d15bea5a1ce276b36223359063","kind":"evidence_added","evidence":[{"id":"gfw-integrated-cod-7-5-2026-07-03..2026-10-01","kind":"alert","source":"gfw-integrated-alerts","datetime":"2026-10-01T00:00:00Z","href":"https://data-api.globalforestwatch.org/dataset/gfw_integrated_alerts","method":{"name":"forest_alerts","version":"1.0","params":{"bbox":[21.4,-7.1,22.49,-5.7],"days":90,"minConfidence":"high"}},"summary":"119338 alerts (≥ high confidence), 1460.72 ha, 2026-07-03…2026-10-01.","values":{"alerts":119338,"ha":1460.72,"high_alerts":118993,"high_ha":1456.5,"highest_alerts":345,"highest_ha":4.22,"living_value_100y_npv2_usd":422929167,"living_value_100y_usd":981311700,"living_value_usd_yr":9813117}}]}],"inclusion":{"index":1094,"leafHash":"0b30baa977f824b600bb87aa7bffd0191feab0f33d068c127bfdb28af5e384fe","proof":["88513b39f9c383d5ab1ca11bc62acd39e32983cbf32e4e3ac7fc4a664b87a2fe","8db4379abb0db900a5d972c3ca73855d12c04d644b096c2edc453ed7408ebc2f","611d5a427539aa6b2cd822d216b801219676e5fc85bb957be344d57f76e1b409","f70b81a02441d6121c5ad48f6f4c3ddb21c1430b914c3b4ed38e548c75286ee7","faf94375e7054adbd22fa0ead48c4d4d3a748de7b2b9cd5efbc8cf21e012ab03","8c9c126ddde5463be4c66b848b31cefa5e2e8b034ab906e8158680a170325566"],"size":1104,"root":"2d46ab865c7813b8991ed0d713906b15bef1321a022b8d382a1069dac9d5686a"}}