Mumbai Suburban — 363 witnessed locality cells
A live confidence-weighted rollup of Mumbai Suburban district over 363 witness-gated MH-36 Voronoi cells (LGD 483). Every regional figure is pooled with GUM confidence weighting and carries its three-way coverage — no standard-error, because spatially autocorrelated cells would make s/√n false precision. Missing is never zero.
The district, pooled honestly
Each figure is a confidence-weighted mean over the district's observed cells, with its descriptive spread and the share of cells that actually witnessed it. No confidence interval is shown — cube cells are spatially autocorrelated, so a standard error would be false precision.
Witnessed now, at the district centroid
A live valid-time read at the district centroid — each value stamped with the instant its witness window closed. Refused (never zero-filled) where the station/satellite lag means the window has not yet closed.
Drill into any cell's full dossier
Each of the 363 published cells has a complete witnessed dossier — hundreds of measures, multi-decade trajectories, a monthly satellite time-lapse, agreement, uncertainty and cryptographic lineage. A sample below; the full set is in the sitemap.
and 303 more — every published cell is listed in the sitemap.
Re-witness this rollup
Nothing here is asserted — every figure is re-checkable against the live engine with one call.
# Re-witness this district rollup — live MCP, anonymous & free.
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":"aggregate_region",
"arguments":{"feature":"built_frac","admin_codes":[483]}}}'Scope: the live MH-36 cube. The engine supplies no goodness, suitability or risk score (refused by doctrine), and reports no confidence interval for a pooled region (spatial autocorrelation would make it false precision).