For Professionals
Citation-Grade Due Diligence for Real Estate
A due-diligence report that calls a neighbourhood "clean and quiet" is an opinion. One that states a PM2.5 annual mean against the WHO 2005 guideline, with the reanalysis product and retrieval date named, is evidence — the kind that survives a client's counsel asking "who says so?" Buyer's agents, relocation consultants, surveyors, and funds all write reports that get challenged. This is how to build the environmental section so it holds.
At a glance
- A defensible due-diligence report names the body, the publication, and the date for every environmental figure — never "our data says."
- Public primary sources already cover the terrain: USGS, EPA, WHO, EU/Euratom, NASA/Copernicus, PHMSA, FEMA, ICNIRP.
- The repeatable checklist runs to roughly eight categories, each one mapped to a named field and a named source body.
- "Green, quiet neighbourhood" is a marketing claim; "PM2.5 annual mean 4.1 µg/m³, WHO AQG 2021 threshold 5 µg/m³" is a citable one.
- Export figures with their source attribution attached — the citation has to travel with the number, not sit in a separate methodology page nobody reads.
The claim that doesn't survive a challenge
Every environmental section of a due-diligence report eventually meets a sceptical reader — opposing counsel in a dispute, a fund's risk committee, a client who lost money on the last purchase and is done taking things on faith. The question is always the same: who says so? If the answer is "the platform we used," the claim is an opinion with a UI. If the answer is "WHO, Global Air Quality Guidelines, 2021, retrieved 14 June 2026," the claim is evidence, and evidence is what a report is for.
Real estate due diligence has run this way for title, structural, and legal matters for decades — a deed reference, a survey stamp, a building-code section number are never presented unsourced. Environmental exposure is the one category that, until recently, ran on adjectives: peaceful, leafy, up-and-coming. That gap exists because granular public environmental data is a relatively new availability, not because the underlying science was ever opinion-shaped. The professionals who close the gap first are the ones whose reports hold up.
Why traceability is the deliverable
The output of an environmental due-diligence workup is not a number. It is a number plus a chain — the claim made in the report, the public primary source that produced the underlying figure, and a citation string specific enough that a third party can retrieve the same figure independently. Drop any link and the chain stops being evidence. A radon reading with no source is a guess. A source with no publication date is unverifiable once the standard body revises its guidance. A citation with no retrieval date can't be defended against "that changed six months ago."
A number with no named source is an opinion wearing a lab coat.
Due diligence has always cited primary sources for title, structural, and legal matters — a deed reference, a survey stamp, a building-code clause. Environmental exposure is the one section that historically ran on adjectives, not because the science was opinion but because granular public data wasn't yet available at parcel scale. That data now exists. The work is extending a citation habit the profession already has, one category further.
The repeatable per-property checklist
A checklist earns its keep only if it produces the same eight-or-so categories on every file, each one tied to a named public body rather than a general impression. This is the version worth keeping on the desk:
- Air quality (PM2.5 / PM10) — grid field
pm25/ pm10 — WHO Global Air Quality Guidelines, 2021; regional reanalysis via Copernicus CAMS. - Radon potential — grid field
radon— EU JRC European Atlas of Natural Radiation / Directive 2013/59/Euratom; USGS–EPA national radon zones in the US. - Water hardness / quality — grid field
hardness— USGS National Water-Quality Assessment programme. - Flood exposure — FEMA National Flood Hazard Layer (US); EU Floods Directive 2007/60/EC hazard mapping.
- Night noise (aircraft / rail) — grid fields
noise/flights— WHO Environmental Noise Guidelines for the European Region, L_night 40 dB. - RF / EMF density — grid field
g5_count— ICNIRP 2020 Guidelines for Limiting Exposure to Electromagnetic Fields. - Pipeline / industrial proximity — PHMSA National Pipeline Mapping System (US); national equivalents elsewhere.
- Land cover / vegetation trend — grid field
forest_cover— NASA / ESA Copernicus Landsat and Sentinel-2 land-cover time series.
WHO halved its PM2.5 annual guideline from 10 µg/m³ to 5 µg/m³ in the 2021 revision of the Global Air Quality Guidelines. A citation that names "WHO" without a publication year is already ambiguous — a reading that cleared the old threshold may not clear the current one. Date every standard, not just every reading.
Anatomy of a citation-grade claim
Strip it down and a defensible environmental claim needs exactly four parts, every time. The source body — the named agency or standards organisation, not a category ("a government source"). The publication and version — the specific document or dataset and its edition year, because guidance revises. The value or threshold — the actual reported figure, in the source's own units, not converted without saying so. The retrieval date — when the figure was pulled, because public datasets update on their own schedule and a report has to say which snapshot it used. Miss any one of the four and the claim reverts to an assertion.
Any mapping platform — CleanZone included — is a retrieval and formatting layer over public data, not the authority behind a figure. A report that cites "the mapping tool" as the source for a claim has a citation that dissolves the moment it's questioned. Cite the issuing body and the publication; treat the platform as the mechanism and the timestamp, nothing more.
Marketing claim
"Peaceful, green neighbourhood with clean air and a quiet setting."
No body, no publication, no date. It cannot be checked, and it cannot be cited in a report that has to survive a challenge.
Sourced figure
"PM2.5 annual mean 4.1 µg/m³ (Copernicus CAMS reanalysis, retrieved 2026-06), against the WHO 2005 Air Quality Guideline of 5 µg/m³."
Named source, named standard, retrieval date. A reviewing party can go and check it independently.
Exporting per-cell data without losing the trail
The failure mode is rarely a missing figure — it's a figure that got copied into a report and left its attribution behind. A field is pulled from the grid, the number goes into an appendix table, and by the time it reaches the report body it reads as an unsourced fact. The fix is procedural, not technical: export figures with the source body and publication attached at the field level, footnote each cited number in the report text rather than bundling a single methodology page at the back, and keep a versioned export so that if a standard is revised or a reading is challenged eighteen months later, you can show exactly which snapshot and which publication version the report relied on.
This matters most for the categories most likely to be argued over: flood zone, radon potential, and industrial proximity carry direct insurance, mortgage, or disclosure consequences, and are exactly the figures a counterparty's counsel will ask to see sourced. A citable figure ends that conversation in one line. An unsourced one starts a longer one.
| Environmental factor | Source body | Public standard / publication | Why it matters for DD |
|---|---|---|---|
| Air quality — PM2.5 | WHO | Global Air Quality Guidelines, 2021 — 5 µg/m³ annual mean | Health-based benchmark increasingly cited in disclosure and insurance underwriting |
| Radon potential | Euratom / EU JRC; USGS-EPA (US) | Directive 2013/59/Euratom, 300 Bq/m³ reference level; EPA 4 pCi/L action level | Determines whether a long-term detector test is routine or a contractual condition |
| Flood exposure | FEMA (US); EU Floods Directive | National Flood Hazard Layer panels; Directive 2007/60/EC hazard maps | Drives insurance eligibility and, in the US, federally-backed mortgage requirements |
| Night noise | WHO | Environmental Noise Guidelines for the European Region — L_night 40 dB | The defensible basis for challenging a listing's "quiet street" description |
| RF / EMF exposure | ICNIRP | 2020 Guidelines for Limiting Exposure to Electromagnetic Fields | The same reference framework regulators and carriers cite for installation compliance |
| Pipeline / industrial proximity | PHMSA (US) | National Pipeline Mapping System | Public register of hazardous-liquid and gas transmission corridors near a parcel |
| Land cover change | NASA / ESA Copernicus | Landsat and Sentinel-2 land-cover time series | Evidences long-term site trend — clearance, development, vegetation loss — with dated imagery |
None of this asks a professional to trust a platform's judgement. It asks for the opposite: name the field (pm25, radon, hardness, noise, flights, g5_count, seismic_pga, forest_cover), name the public body behind it, and let the citation do the defending. The professional and research tiers on CleanZone are built around exactly this export-with-attribution workflow.