Skip to content
Cube coverageMMR-513,619 clusters × 368MH-362,64,583 clusters × 336Pan-India653 districts × 39audit artefacts →
Use case · climate risk

High-exposure cells across MH-36.

Surface cells where population sits in low-lying cropland / built / wetland LULC with thin canopy buffer. The proxy is morphology-only; the full catastrophe stack lives in the 336-col cube.

§1 The buyer's question

“Where in Maharashtra am I most exposed to flood/heat risk on populated cells with no canopy buffer? Sort by that, slice by district, attach lineage to every cell.”

§2 The cube columns it touches

scope    : mh36
cube_fp  : dcc1845460cbd47c…
weights  : pop_density_per_km2  +0.40   (WorldPop India v3-2020)
           built_fraction        +0.20   (GOB v3 ÷ R5)
           terrain_elev_m        -0.30   (Copernicus GLO-30 — low-lying penalised)
           ndvi_p50              -0.10   (low canopy buffer)
filter   : lulc_class_primary IN (40 cropland, 50 built-up, 90 wetland, 95 mangrove)
           min_witnesses ≥ 3

# IN the 45-dim morphology vector v2 (long-term climate dynamics, with per-cell GUM uncertainty):
#   Tier-Q (IMD 125-yr): rainfall / monsoon / 4-season trends (Theil-Sen + Mann-Kendall),
#   interannual CV, annual-max-1-day extreme trend, Oliver concentration index.
#   Tier-U (NOAA GSOD ~50-yr): annual-mean temperature + warming trend (Theil-Sen +
#   Mann-Kendall), summer-max / winter-min, IDW from stations.
#   Each trend carries an honest per-cell uncertainty (Tier-Q slope SE, Tier-U IDW SD)
#   so a noisy / sparsely-supported trend is down-weighted, never sold as exact.
# STILL cube-only (high-frequency / realtime, not yet vectorised):
#   Tier-H/H′ IMERG precip return periods, Tier-J CHIRPS daily extremes,
#   Tier-M CPCB AQI, Tier-N WRIS hydro, Tier-S WRIS-GW. All in the 336-col cube on S3.

§3 Worked example — live from the cube

Cells scored
177,712
Districts in scope
36
Top-K returned
10
Climate tiers held
H/H′/J/M/N/S
in full cube only
#AnchorDistrictLULCpop /km²elev mNDVIscore
1483:4830000240Mumbai Suburban50 built up56,184.4280.14515.94
2497:4970000807Thane50 built up48,459.4260.14715.63
3483:4830000257Mumbai Suburban50 built up55,204.5240.13815.61
4497:4970001375Thane50 built up48,317.990.12515.58
5483:4830000241Mumbai Suburban50 built up49,870.1420.15014.35
6483:4830000103Mumbai Suburban50 built up43,637.4270.14314.32
7483:4830000221Mumbai Suburban50 built up50,563.3200.20614.06
8483:4830000319Mumbai Suburban50 built up49,814.1190.15214.05
9483:4830000347Mumbai Suburban50 built up46,560.3170.17314.03
10483:4830000063Mumbai Suburban50 built up44,364.6100.09313.99
Top cell — full evidence record
Mumbai Suburban · 483:4830000240
lat 19.1750 · lon 72.8550
LULC 50 built_up
Built fraction
0.334
Building count
1104
Building height p50
5.73 m
Road length
28.03 km
Population density
56184 /km²
NDVI p50
0.145
Terrain elevation
27.8 m
Honest scope note

The 45-dim morphology vector v2 now carries the long-term climate-dynamics axes directly (IMD-125yr rainfall / monsoon / 4-season Theil–Sen trends, Mann–Kendall significance, interannual variability and the precipitation-concentration index via Tier-Q; plus NOAA-GSOD ~50-yr temperature-warming trends via Tier-U — each with an honest per-cell trend uncertainty). What it does not yet carry are the high-frequency catastrophe layers parametric insurance prices on directly — IMERG precip return periods, CHIRPS daily extremes, WRIS river-discharge percentiles, GLOFAS flood return curves. Those ship in the mh36-unified-v1.7 cube on S3 today and can be wired into a per-customer feature set on request. See audit artefacts for the per-tier coverage figures.

§4 The workflow in Explorer

Step 1
Inspect a coastal cluster

Open Explorer → Inspect and pick a Mumbai-Suburban (483) or Palghar (665) anchor.

Step 2
Twin against inland

Find Twins to surface inland analogues — the difference tells you the coastal exposure premium.

Step 3
Score for portfolio

Use the climate-risk weights above or tune your own to rank exposure across all 36 districts.

§5 Talk to us about a pilot

Parametric insurance customers get the full climate-tier feature set wired into a custom vector before pricing.

Get in touch