Technical Documentation

Satellite Environmental Intelligence Methodology

How Zolena Lab uses open satellite data and public environmental databases to produce audit-traceable ESG measurements and environmental risk intelligence — with academic traceability to the University of Calgary Schulich School of Engineering.

Home› Methodology
Core Principle

ESG Data Must Be Measured and Traceable

The foundation of Zolena Lab's approach is a single non-negotiable principle: ESG data must be measured and traceable — not self-reported by asset owners.

Factor-based ESG estimation methods (applying generic emission factors or biomass coefficients to activity data reported by the asset owner) carry uncertainty ranges of 38% to 150% per IPCC and GHG Protocol documentation. This uncertainty makes the data difficult to audit, easy to manipulate, and increasingly unacceptable to institutional investors, regulators, and rating agencies.

Satellite remote sensing produces data that is:

  • Independent — the satellite records what it observes regardless of what the asset owner reports
  • Reproducible — any party with technical capability can re-run the same analysis on the same open dataset and verify the result
  • Spatially precise — every data point is geolocated to a specific coordinate on Earth
  • Historically archived — Sentinel-2 data extends back to 2015, enabling multi-year trend analysis
  • Free — Copernicus data is openly available at zero cost, enabling low-cost delivery at scale
Primary Data Source

Copernicus Sentinel-2 Satellite

All Zolena Lab primary analyses use Copernicus Sentinel-2 Level-2A (L2A) imagery, produced by the European Space Agency (ESA) and made freely available through the Copernicus open data programme.

SpecificationValue
SatelliteESA Sentinel-2A and Sentinel-2B (constellation)
Processing LevelL2A — atmospherically corrected surface reflectance
Spatial Resolution10 metres per pixel (visible and NIR bands); 20 metres (red-edge and SWIR bands)
Temporal Resolution5-day revisit frequency at equator; up to daily at higher latitudes
Spectral Bands UsedB2 (Blue), B3 (Green), B4 (Red), B8 (NIR), B11 (SWIR1), B12 (SWIR2)
Data AccessFree and open — Copernicus Data Space Ecosystem (CDSE) and Google Earth Engine
Historical CoverageJune 2015 to present (Sentinel-2A); March 2017 to present (Sentinel-2B)
LicenceCopernicus open licence — free commercial use with attribution
Image Selection Protocol
Cloud cover filter: Maximum 20% cloud cover per scene. Higher cloud cover scenes are excluded from analysis.

Seasonal window: For high-latitude locations (Alberta, northern Canada), analysis is restricted to the May through September growing season window to exclude snow-covered periods. NDSI (Normalized Difference Snow Index) snow masking is applied with a threshold of 0.4 to exclude residual snow-covered pixels.

Mosaicking: Where multiple scenes are available within an analysis window, a median composite is produced to minimize the effect of remaining atmospheric noise and cloud shadows.

Same-season comparison: All year-over-year comparisons use same-season imagery (summer to summer, not summer to autumn) to eliminate phenological variation from change detection results.
Core Index

NDVI — Normalized Difference Vegetation Index

NDVI is the primary vegetation health indicator used across all Zolena Lab products. It is one of the most extensively validated remote sensing indices in the scientific literature, with over four decades of application in ecology, agriculture, forestry, and land monitoring.

NDVI = (NIR − Red) / (NIR + Red)

Using Sentinel-2 bands:
NDVI = (B8 − B4) / (B8 + B4)

Range: −1.0 to +1.0
Healthy vegetation: typically 0.4 to 0.9
Bare soil or stressed vegetation: typically 0.0 to 0.3
Water bodies and built surfaces: typically below 0.0

Interpretation in Zolena Lab products:

  • NDVI mean above 0.5: healthy, well-vegetated area
  • NDVI mean 0.3 to 0.5: moderate vegetation coverage; may include managed green space or mixed land use
  • NDVI mean below 0.3 in an area not classified as built surface: anomalous — warrants investigation
  • NDVI decline greater than 0.1 between years in the same season: significant change event — primary trigger for flagging in change detection analysis
  • NDVI decline greater than 0.15 sustained over 12 months in a forested area: deforestation signal in EUDR context
Supplementary Index

SWIR — Shortwave Infrared Composite

The SWIR composite (Sentinel-2 bands B11, B8, B4) extends analysis beyond what the human eye and standard RGB photography can detect. SWIR wavelengths penetrate vegetation canopy more deeply and are sensitive to soil moisture content, mineral composition, and certain organic compounds.

SWIR colour interpretation in Zolena Lab reports:

  • Bright green: healthy vegetation with adequate soil moisture
  • Magenta or pink: built surfaces, rooftops, or hardscaped areas
  • Yellow-orange: low soil moisture, bare soil, or mild surface disturbance
  • Dark red or brown: significant soil disturbance, or areas of elevated surface temperature
  • Black or very dark: water bodies, deep shadow

SWIR is particularly diagnostic for identifying hydrocarbon residues on soil surfaces, which produce a distinctive spectral signature in the B11 and B12 bands. This makes it a useful screening indicator for potential contamination in industrial land contexts, though it cannot substitute for ground-truth soil testing.

Carbon Estimation

Carbon Sequestration Methodology

Carbon sequestration estimates in Zolena Lab ESG reports use a two-step approach combining NDVI-derived biomass proxy with IPCC standard biomass factors.

Step 1: NDVI to Biomass Proxy
NDVI is used as a proxy for above-ground biomass density. Higher NDVI values indicate greater photosynthetically active vegetation mass. The relationship between NDVI and biomass is well-established in the remote sensing literature for temperate vegetation types.

Biomass proxy (tonnes per hectare) = f(NDVI mean, vegetation class, region)

Vegetation class is determined through land cover classification using multi-band Sentinel-2 analysis. Classes include: tree canopy, shrub and grassland, managed lawn, agricultural crop, and impervious surface.
Step 2: Biomass to Carbon Sequestration
Standard applied: IPCC 2006 Guidelines for National Greenhouse Gas Inventories, Volume 4 (Agriculture, Forestry and Other Land Use), Chapter 4, Table 4.9 — biomass factors for temperate and boreal forest zones.

Carbon stock (tC/ha) = Biomass proxy × Carbon fraction (0.47 for temperate forest, IPCC default)

Annual sequestration estimate = Change in carbon stock per year (where multi-year data is available) or steady-state carbon stock proxy for single-year analysis.

Uncertainty disclosure: All carbon estimates are expressed with the IPCC-specified uncertainty range for the relevant biomass factor (typically ±30% to ±50%). This uncertainty is explicitly disclosed in all reports. Zolena Lab does not present carbon estimates as precise measurements — they are scientifically grounded estimates with disclosed uncertainty, which is more defensible than factor-based estimates that do not disclose uncertainty at all.
Citation: IPCC Carbon Standard IPCC (2006). 2006 IPCC Guidelines for National Greenhouse Gas Inventories. Volume 4: Agriculture, Forestry and Other Land Use. Chapter 4: Forest Land. Table 4.9: Default biomass factors for use with Tier 1 methods. Intergovernmental Panel on Climate Change, Geneva.
Climate and Hazard Data

Secondary Data Sources for Hazard Assessment

DatabaseProviderApplication in Zolena ReportsCoverage
NASA FIRMS
Fire Information for Resource Management System		 NASA Earth Science Division Historical wildfire point detection within 20km of analysis area. MODIS and VIIRS active fire pixels at 375m resolution. Global · 2000–present · Free open data
JRC Global Surface Water v1.4 European Commission Joint Research Centre Historical permanent and seasonal water coverage within parcel boundary. Maximum extent and occurrence frequency layers. Global · 1984–2021 · Free open data
Copernicus C3S Climate Data Store European Centre for Medium-Range Weather Forecasts (ECMWF) Historical drought indices (SPI, SPEI), freeze-thaw cycle frequency, precipitation anomaly records. Future climate projections in Pro Edition. Global · Historical and projected · Free open data
CanESM + IPCC AR6 Environment and Climate Change Canada + IPCC 50-year regional climate projections (2030, 2040, 2050 time slices) for Pro Edition reports. Temperature, precipitation, and extreme event frequency projections. Canada and global · Pro Edition only
SRTM / ALOS Digital Elevation Model NASA / JAXA Watershed delineation for upstream contamination pathway analysis. Terrain slope and flow direction modelling. Global · Free open data
Scoring Frameworks

Zolena Proprietary Scoring Systems

Zolena Lab has developed two distinct scoring frameworks for its product lines. These frameworks are completely separate and must not be confused or combined:

Zolena Land Risk Score™ · Products 04 (Land Risk Screening)
Purpose: Risk assessment for pre-purchase land due diligence. Lower scores indicate higher risk.

Output: Risk Grade (L1 Low Risk → L4 High Risk) + Investment Recommendation Grade (A Proceed → D Defer)

Scoring dimensions:
· Current environmental state (50% weight) — Layer 01 current NDVI, SWIR, RGB + Layer 03 500m buffer pressure
· Historical change trajectory (30% weight) — Layer 02 10-year NDVI trend + Layer 04 climate hazard history
· Future climate trend (20% weight) — Pro Edition only, C3S/CanESM/IPCC AR6 projections

Design principle: Scores current state most heavily because land buyers are primarily concerned with present condition. Historical trajectory provides context; future trend supports long-term investment planning.
Zolena Green Space Score™ · Product 01 (Corporate Green Space ESG)
Purpose: Value quantification for corporate ESG disclosure. Higher scores indicate greater green asset quality.

Output: Composite score (0–100) benchmarked against comparable green spaces

Scoring dimensions:
· Vegetation health — NDVI mean and distribution (35% weight)
· Green coverage density — vegetated area fraction (25% weight)
· Carbon sequestration capacity — NDVI-biomass-IPCC proxy (25% weight)
· Community accessibility — population within 500m radius (15% weight)

Design principle: Weighted toward ecological indicators (NDVI, coverage, carbon) with a community benefit component reflecting GRI 413-1 Local Communities standard.
Academic Reference

University of Calgary Schulich School of Engineering

Zolena Lab's data processing methodology references open-access research on environmental sensing and machine learning calibration from the University of Calgary Schulich School of Engineering.

Algorithm Reference Du, K. et al. University of Calgary, Schulich School of Engineering. Environmental sensing and machine learning calibration research. Published under Creative Commons Attribution 4.0 International Licence (CC BY 4.0).

This reference is cited in the appendix of all Zolena Lab reports in the format: "Du, K. et al. University of Calgary, Schulich School of Engineering. CC BY 4.0"

Licence scope: CC BY 4.0 permits free commercial use of the research content with attribution. This reference does not imply endorsement or partnership with the University of Calgary or the Schulich School of Engineering.
Limitations and Disclosure

What Satellite Data Cannot Do

Zolena Lab is committed to transparent disclosure of the limitations of satellite-based environmental analysis. The following limitations apply to all our products:

  • Spatial resolution: Sentinel-2 has a 10-metre spatial resolution. Features smaller than 10 metres may not be detectable. This includes small contamination plumes, individual tree canopy health, and sub-parcel land use details.
  • Sub-surface conditions: Satellite remote sensing observes surface and near-surface conditions only. Sub-surface contamination, groundwater conditions, and buried structures cannot be assessed through satellite analysis alone.
  • Carbon precision: Carbon sequestration estimates carry uncertainty ranges of ±30% to ±50% per IPCC biomass factor specifications. These are estimates, not precise measurements.
  • Cloud cover: Optical satellite imagery cannot penetrate cloud cover. In regions with persistent cloud cover, suitable imagery may not be available for the desired analysis period.
  • Self-reported data in B-Track reports: Client-provided domestic supply chain data included in EUDR B-Track reports has not been independently verified by Zolena Lab. This limitation is explicitly stated in every B-Track report.
  • Not a substitute for Phase I ESA: Land risk screening reports do not constitute a Phase I or Phase II Environmental Site Assessment as defined under ASTM E1527 or CSA Z768. They are preliminary screening tools intended to inform the decision of whether a formal ESA is warranted.
Standards Alignment

Reporting Standards Referenced

StandardFull NameApplication in Zolena Products
GRI 413-1GRI 2021 Standard: Local CommunitiesGreen space ESG reports — community impact indicators
TCFDTask Force on Climate-related Financial DisclosuresPhysical climate risk overlay in all products
CSRDCorporate Sustainability Reporting Directive (EU)EUDR B-Track and property ESG international version
EUDREU Regulation 2023/1115 on Deforestation-free ProductsEUDR supply chain compliance verification product
GRESBGlobal Real Estate Sustainability BenchmarkProperty ESG value report — indicator alignment
TNFDTaskforce on Nature-related Financial DisclosuresBiodiversity proxy indicators in property ESG reports
IPCC 2006IPCC Guidelines for GHG Inventories Vol.4 Ch.4 Table 4.9Carbon sequestration estimation in all vegetation products
ASTM E1527Standard Practice for Environmental Site AssessmentsReferenced to define scope boundary of land risk screening
CSA Z768Phase I Environmental Site Assessment (Canada)Referenced to define scope boundary of land risk screening
Questions About Our Methodology?
We welcome technical questions from ESG auditors, environmental professionals, and institutional investors.
All methodology documentation is available for review upon request.
Contact: [email protected] View Land Risk Product →