Bhārata Strata · Maharashtra · Ratnagiri · morphological twins
Places like cell 4920004824
The MH-36 locality cells most similar to this Ratnagiri cell across terrain, built form, land cover, climate dynamics and connectivity — ranked by a GUM-confidence-weighted (R24) distance, with the temporal block down-weighted. Each twin is a real cell with its own witnessed dossier.
Twins found
18
across 2 distinct districts
Index confidence
0.9
GUM mean confidence of the anchor (R24)
Dimensions compared
45
morphology-v2 feature dimensions
Closest twin distance
0.255
lower = more alike (weighted)
The twins, closest first
#1 · Sindhudurgcell 4950000461
distance 0.255 · cross-district
#2 · Ratnagiricell 4920005494
distance 0.265
#3 · Ratnagiricell 4920007081
distance 0.288
#4 · Ratnagiricell 4920006063
distance 0.29
#5 · Ratnagiricell 4920004249
distance 0.303
#6 · Ratnagiricell 4920006209
distance 0.304
#7 · Ratnagiricell 4920004589
distance 0.311
#8 · Ratnagiricell 4920006672
distance 0.314
#9 · Ratnagiricell 4920006009
distance 0.317
#10 · Sindhudurgcell 4950000351
distance 0.318 · cross-district
#11 · Ratnagiricell 4920006152
distance 0.322
#12 · Ratnagiricell 4920005495
distance 0.324
#13 · Ratnagiricell 4920005806
distance 0.326
#14 · Ratnagiricell 4920004613
distance 0.335
#15 · Ratnagiricell 4920004183
distance 0.337
#16 · Ratnagiricell 4920004643
distance 0.344
#17 · Sindhudurgcell 4950000747
distance 0.345 · cross-district
#18 · Ratnagiricell 4920006070
distance 0.347
witness: find_twins — morphology-twin-index-v2 (R24, GUM-confidence-weighted, magnitude-aware, temporal block down-weighted). Distance is the weighted feature distance; lower is more alike. Cross-district twins are the discriminative signal.
Re-witness these twins
curl -s https://api.gridrock.ai/mcp -H 'Content-Type: application/json' \
-H 'Accept: application/json, text/event-stream' \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/call","params":{"name":"find_twins",
"arguments":{"admin_code":492,"cluster_id":4920004824}}}'Scope: the live MH-36 cube. Twins are computed from the sha-verified morphology-v2 vector; the engine supplies no goodness/suitability ranking — only morphological similarity.