A complete, transparent explanation of how satellite data from NASA, USGS and ESA is combined into a single 0–100 risk score — and how every number in your report is derived from verified public datasets.
📡 6 data parameters⚖️ Weighted composite algorithm🇮🇳 IS 1893-2016 compliant🔓 Fully transparent methodology
01 — The Algorithm
How the 0–100 Risk Score Is Calculated
The GeoLens composite risk score is a weighted sum of six independent parameters, each derived from a different satellite or government data source. The weights reflect relative contribution to real-world land hazard, calibrated against Indian construction standards and disaster records.
The formula is intentionally simple and auditable — no black-box machine learning. Every point can be traced back to a specific data value.
Sum of all six components (capped at 100)= RISK SCORE / 100
What the Score Means
0 – 34
Low Risk
Relatively safer site. Standard construction with full local building code compliance recommended. Professional geotechnical consultation still advised.
35 – 64
Medium Risk
Moderate risk indicators detected. Engineering precautions advisable. Consult a licensed civil engineer before proceeding with construction.
65 – 100
High Risk
Significant risk factors present. On-site geotechnical investigation required before any investment decision. Do not proceed without expert assessment.
ℹ️
Important: This is probability, not certainty.
A Low score does not guarantee the land is safe — it means historical satellite data shows fewer risk indicators. A field survey by a licensed geotechnical engineer is always required before construction.
02 — Data Sources
The Six Datasets Behind Every Report
Every parameter in a GeoLens report is fetched in real-time from verified public APIs at the moment you generate it. No stored, cached, or synthetic data is used.
🌧️
Open-Meteo ERA5
open-meteo.com · ECMWF Copernicus
10-year hourly climate reanalysis from the European Centre for Medium-Range Weather Forecasts. Used for precipitation, temperature, wind speed and evapotranspiration. Covers 1940–present globally at 30km grid resolution.
✓ ±5% precipitation accuracy
⛰️
NASA SRTM 30m DEM
Open-Elevation · USGS distributed
Shuttle Radar Topography Mission digital elevation model captured by NASA in February 2000. Provides global elevation data at 30-metre horizontal resolution. Used to determine site elevation above sea level and terrain slope angle.
✓ ±10m vertical accuracy
🌍
USGS FDSNWS Earthquakes
earthquake.usgs.gov
US Geological Survey real-time earthquake catalog. GeoLens queries all M≥3.0 events within 100km radius of the site over the past 10 years. Data is verified by USGS seismologists and updated within minutes of any event worldwide.
✓ M0.1 magnitude accuracy
🏛️
BIS IS 1893-2016
Bureau of Indian Standards · USGS ShakeMap
Bureau of Indian Standards seismic zonation map. Classifies all of India into Zones II through V based on peak ground acceleration and historical seismicity. Zone V (Northeast India, Himalayas, Andaman) carries the highest structural risk requirement.
✓ Zone boundary resolution
🌱
NDVI Vegetation Proxy
Open-Meteo ET0 + Precipitation
Normalized Difference Vegetation Index proxy derived from evapotranspiration and precipitation data. Used to classify soil type (Deep Alluvial, Black Cotton, Laterite, Rocky/Loamy) and estimate soil moisture index. Direct Sentinel/Landsat NDVI requires paid API access.
✓ Statistical proxy estimate
🗺️
OpenStreetMap Overpass API
overpass-api.de · CC-BY-SA License
Community-maintained global geographic database. Used in Premium reports to find nearest classified road (for truck access), power substations, water works, and other infrastructure within 1.5km of the site. Data quality depends on community mapping activity in the region.
✓ <100m address accuracy
Data Accuracy Summary
Parameter
Source
Dataset
Accuracy
Precipitation / Climate
Open-Meteo
ERA5 Reanalysis (10yr)
±5%
Elevation Above Sea Level
NASA / USGS
SRTM 30m DEM
±10m
Earthquake Catalog
USGS FDSNWS
Real-time verified
M0.1
Seismic Zone
BIS IS 1893-2016
Zone boundary map
Zone-level
Geocoding / Address
OpenStreetMap/Nominatim
CC-BY-SA
<100m
Vegetation / NDVI
Open-Meteo ET0 proxy
Statistical estimate
Indicative
03 — Data Pipeline
Step-by-Step: From Coordinates to Report
When you submit a location, GeoLens runs six API calls in parallel on its server — before any payment is taken. If any critical data source fails, the report is blocked and you are not charged.
1
Geocoding — Coordinates → Address
Your entered address or map pin is reverse-geocoded via Nominatim / OpenStreetMap to confirm the exact location label used in the report. GPS coordinates are the primary input — the address is display-only.
2
Elevation Fetch — NASA SRTM
A single API call to open-elevation.com returns the site's metres above sea level and an estimated terrain slope. Slope is calculated by querying 4 nearby points and computing the gradient angle. Low elevation (<15m ASL) raises the flood and groundwater score.
3
10-Year Climate Archive — ERA5
Open-Meteo's ERA5 endpoint is queried for 10 years of daily precipitation data. Days with rainfall exceeding 80mm/day — the IMD threshold for "Very Heavy Rainfall" — are counted per year. This flood-day total drives 40% of the risk score.
4
Earthquake Catalog — USGS FDSNWS
USGS is queried for all events M≥3.0 within a 100km radius over the past 10 years. The maximum recorded magnitude drives the seismic-history component. Zero events in a Zone III area still carries a base zone penalty.
5
Seismic Zone Classification — IS 1893
A deterministic lookup against the IS 1893-2016 zone boundaries, implemented as coordinate-range logic. Zone II = 2 pts, Zone III = 5 pts, Zone IV = 10 pts, Zone V = 15 pts. Northeast India and Himalayan foothills typically resolve to Zone IV or V.
6
NDVI Proxy + Soil Classification
Annual average precipitation and evapotranspiration from ERA5 are used to compute a vegetation proxy score. Geographic coordinates determine the soil type from five regional classes: Alluvial (Indo-Gangetic Belt), Black Cotton (Deccan), Laterite (South India coast), Rocky/Loamy (Himalayas), Mixed Alluvial (others).
7
Composite Score + Report Generation
All six parameters are combined using the weighted formula. The score, risk level (LOW / MEDIUM / HIGH), and all raw data are returned to the browser. No payment is taken until data is confirmed available. The PDF is built client-side from this data using jsPDF.
04 — Future Projection
How the 5-Year Rainfall Projection Works
The Standard and Premium reports include a 5-year projected annual rainfall outlook. This is calculated using Ordinary Least Squares linear regression on the 10-year ERA5 archive — not a meteorological model.
The regression formula applied to annual rainfall data:
ŷ = intercept + (slope × year_index)
Where slope = (n·Σxy − Σx·Σy) / (n·Σx² − (Σx)²) — computed over the available ERA5 years. The slope determines trend direction.
📈
Increasing
slope > +50 mm/yr
➡️
Stable
−50 to +50 mm/yr
📉
Decreasing
slope < −50 mm/yr
⚠️
This is a trend projection, not a meteorological forecast.
ERA5 reanalysis covers approximately 10 years of archive. True multi-decade analysis (1990–2025) requires extended Copernicus Climate Data Store API access. Projections are indicative signals — not substitutes for IMD station data for investment-grade decisions.
05 — Project Intelligence
How Project Type Shapes Your Report
The composite risk score is always location-based and never changes with project type. What changes is everything else: the engineering recommendations, construction advisory, compliance checklist and the sections that are emphasised. Here is exactly what each project type focuses on:
Elevated floor area, critical systems placement, tenant disclosure requirements
Electrical panels on first floor minimum, flood NOC from DDMA, commercial flood insurance
Lease disclosureFlood NOC mandatory
🏫School / InstitutionalLife Safety — Zero Tolerance
Seismic safety above all else, emergency evacuation routes, safe school standards
IS 1893 Zone V compliance mandatory, soft-storey analysis, NDMA safe school protocol
NDMA Safe SchoolEvacuation plan
☀️Solar PlantWind, Soiling & Sun Hours
Wind load on panels, dust/soiling from bare soil, annual solar irradiance potential
Wind-resistant panel mounting (>25 km/h sites), anti-soiling coatings for alluvial soil, monsoon flood clearance height for inverters
ERA5 wind dataIrradiance proxy
06 — Seismic Classification
Understanding IS 1893-2016 Seismic Zones
India is divided into four seismic zones under the Bureau of Indian Standards code IS 1893:2016. This classification determines mandatory structural design requirements for all construction in India — not following it is illegal.
ZONE II — LOW DAMAGE RISK
Parts of South India, interior Maharashtra, Rajasthan desert regions. Basic seismic-resistant construction recommended. Adds 2 points to risk score.
ZONE III — MODERATE DAMAGE RISK
Central and northern India including UP, Bihar, Madhya Pradesh. Standard seismic precautions required. Adds 5 points to risk score.
ZONE IV — HIGH DAMAGE RISK
Delhi NCR, J&K, parts of West Bengal, Andaman Islands. Earthquake-resistant construction required per IS 1893. Adds 10 points to risk score.
ZONE V — VERY HIGH DAMAGE RISK
Northeast India (Assam, Manipur, Meghalaya), Himalayan foothills, Andaman & Nicobar. Strongest earthquake-resistant design mandatory. Adds 15 points to risk score.
📌
Northeast India note: Assam, Arunachal Pradesh, Manipur, Nagaland, Meghalaya, Mizoram, Tripura and Sikkim all fall under Zone IV or Zone V — the two highest risk categories. This region also has some of the highest flood probability in India due to Brahmaputra basin dynamics. Both factors combine to produce higher GeoLens scores in this region, which is why site-specific data matters especially here.
07 — Engineering Advisory
How Engineering Recommendations Are Generated
Every recommendation in a GeoLens report is derived from Indian Standards and NDMA guidelines — not from generic AI text. The system selects recommendations based on the combination of your site's actual data values and your chosen project type.
🏗️
Foundation Type → Soil Classification + Elevation
Deep Alluvial soil at low elevation (<20m ASL) triggers pile foundation recommendation (min 10m depth). Rocky/Loamy soil at higher elevation allows spread footing. Black Cotton Soil always triggers under-reamed pile or geo-grid raft recommendation due to high expansive shrink-swell properties.
🌊
Flood Mitigation → Flood Level + Return Period
Sites with HIGH flood probability (total extreme rain days >40 over 10 years) receive mandatory recommendations for 1.5–2.0m elevated plinth levels, perimeter drainage channels, flood-proof door frames and vent covers — based on NDMA flood-resilient construction guidelines.
🌍
Seismic Compliance → IS 1893 Zone
Zone III and above triggers IS 1893-2016 compliant RCC frame construction recommendation. Zone IV and V additionally require ductile detailing per IS:13920, and explicitly prohibit unreinforced masonry load-bearing construction — matching the mandatory requirements under Indian law.
💨
Wind & Temperature → ERA5 Annual Averages
Sites with average wind speed above 25 km/h receive recommendations for wind-resistant roof design. Sites with average maximum temperature above 36°C trigger heat stress warnings and cooling system advisory — important for factory and solar installations where thermal management affects equipment life.
⚖️
Legal status of recommendations: GeoLens recommendations are informational guides based on Indian Standards, not engineering certificates. They help you ask better questions and prepare briefs for licensed engineers — but they cannot replace a structural engineer's stamp or a geotechnical report for regulatory submissions.
08 — Common Questions
Methodology FAQ
Detailed technical questions about how GeoLens works — different from the product FAQ on the home page.
The score you see on screen and the score in any downloaded PDF are computed from the same server-side data returned by the API. If you notice a very small difference, it is likely due to rounding of floating-point intermediate values in different display contexts. The authoritative score is the one returned by the server at the time of report generation and shown in the PDF header. All three plan tiers (Basic, Standard, Premium) receive the same risk score calculation — the difference between plans is in the depth of data analysis and number of report pages, not the score itself.
The India Meteorological Department (IMD) classifies a day with rainfall exceeding 80mm as a "Very Heavy Rainfall" event — the category most strongly associated with urban flooding, agricultural damage and infrastructure disruption in Indian conditions. GeoLens counts ERA5 days where the 24-hour precipitation total exceeds 80mm at the site coordinates. More than 40 such days over 10 years indicates HIGH flood probability; fewer than 15 indicates LOW. This threshold is used consistently in NDMA flood hazard assessments for India.
GeoLens provides an indicative safe bearing capacity (SBC) range — not a measured value. The estimate is derived from three inputs: (1) the geographic soil type classification (Alluvial, Black Cotton, Laterite, Rocky/Loamy), (2) the site's elevation above sea level, and (3) historical rainfall. For example, Deep Alluvial soil at elevation below 20m ASL is assigned a conservative SBC range of 60–100 kN/m² with a pile foundation recommendation — consistent with published ranges in IS:1888 and IS:2911 for similar conditions. An actual measured SBC requires plate load tests or standard penetration tests by a licensed geotechnical engineer.
Direct NDVI from Sentinel-2 or Landsat imagery requires authenticated access to ESA's Copernicus Data Space or NASA's Earthdata API, both of which have usage costs and rate limits that make per-report real-time fetching impractical at GeoLens pricing. The ERA5-based proxy uses the relationship between annual evapotranspiration, precipitation, and vegetation cover — a well-established correlation in climate science — to produce a vegetation density estimate that is sufficient for soil type classification purposes. It cannot detect fine-scale changes like crop rotation or recent deforestation.
GeoLens does not store your location data beyond the duration of your session. The Cloudflare Worker that processes API calls does not log coordinates to any database. Location data is sent to third-party APIs (Open-Meteo, USGS, Open-Elevation, Nominatim) as required to fetch your report data — these are public APIs whose privacy policies govern that use. No location data is sold or shared for advertising purposes. See the GeoLens Privacy Policy for full details.
GeoLens fetches all six data parameters in parallel before displaying the payment screen. If any critical API fails (Open-Meteo, USGS, or Open-Elevation), the report generation is blocked and you receive an error message — no payment is taken. This is by design: we believe you should only pay for a report when all data has been successfully retrieved. Minor failures in non-critical parameters (like the OpenStreetMap infrastructure scan) are handled gracefully with "data unavailable" placeholders in the affected section.
READY TO ANALYSE YOUR LAND?
Now That You Know How It Works
Enter any address or GPS coordinates across India. Get a full satellite risk report in under 60 seconds.