Illustration: a satellite above North America with layers of imagery, terrain, temperature and cloud stacked beside the continent.

Methodology

How ICI works with evidence

From satellites, weather stations and the sky above to the materials and records of one building: how ICI checks what it uses, labels what kind of information it is, and says plainly what it cannot tell you.

Illustration

Where each method stands

What ICI studies, and how far each method has come

ICI now connects Earth, space, past events, materials and buildings. Some of that is published and checked. Some is research that is not ready to give you a number. This list says which is which, and it changes only when the work does.

  • Published

    Earth and Space Almanac calculations

    Positions of the Sun, Moon and planets, sunrise and sunset, and the day and night line on the globe, calculated from published astronomy formulas and checked against worked examples and NASA JPL Horizons.

  • Published

    Past environmental events

    Event stories and history layers rebuilt from held records, each labelled as observed, reported or modelled, with what the record cannot show.

  • Research

    Infrastructure Condition Index: historical event assessment

    How significant a past weather event was, and what it may mean for buildings and infrastructure. Four things are kept apart: how severe the event was, which kinds of infrastructure it can affect, what damage or disruption was actually documented, and how complete the evidence is. No assessments are published yet.

  • Unchanged

    The ICI Score for building records (Method 2.4)

    What a building's documented records support about its main systems. Its definition is unchanged by the new event research; every existing result keeps the version that produced it. Pre-ratification: not independently validated or certified. How the ICI Score works

  • Published

    Exposure and what it can mean for buildings

    Plain explanations of how heat, cold, wind, rain, snow, smoke, coasts and sunlight act on buildings, linked to the layers they describe. Explanations, not scores.

  • Research

    Environmental opportunities

    The good side of the natural world: daylight, sunshine, snow seasons, dark skies, breezes. Each will be its own measured condition; you choose what you value. There will be no single best-place score.

  • In development

    Community Resilience Index

    Exposure, vulnerability, readiness and capacity, kept separate from how often disasters were recorded. No scores are published until the method is validated. About the Index

  • Research prototype

    Materials and emissions (Built World Footprint)

    Life-cycle stages A1 to D under the standard module names. Only stages with held, licensed source data are calculated; the rest are shown as not counted. Biogenic carbon and later benefits are never subtracted. See the lumber example

  • Always

    Forecasts are not observations

    Measured, modelled, estimated, forecast, projected and visitor-reported information are labelled and never passed off as each other.

  • Always

    Unknown stays unknown

    Missing information is shown as missing. It is never counted as low risk, good condition or zero.

  • Always

    Coverage is stated

    Built and tested for every Canadian province and territory and every U.S. state, D.C. and supported territory, and worldwide where a source covers it. Where a source stops, the page says so.

  • Always

    Versions you can reproduce

    Every method has a number. A change means a new version with its date, never a silent update, and older results keep their original definition.

Historical event registry: the counts so far

The first research run reads ten years of NOAA Storm Events records (2016 to 2025) and classifies every wind report with a gust of 50 knots or more: the strength the National Weather Service treats as damaging. These are research counts, read straight from the registry file. A classified record is not a published ICI finding: only assessments that pass the instrument check and keep the same severity class are published, under method ICI-HEA-1.0-wind, and you can explore them on the Resilience page. The first check compared the strongest gusts that the NWS filed as measured at an airport weather station with that station's own record: most agree, some do not, and a few cannot be checked. Disagreement is not proof that either record is wrong; it means the report is not counted as validated.

Loading the registry counts.

From observation to decision

Four steps, every time

  1. Gather

    Take information from the public body that published it, with its edition, date and licence recorded.

  2. Check

    Confirm it is complete, in the right place and the right units, and that it covers the area and period it claims.

  3. Find

    Compare like with like: the same measure, the same period, a nearby and similar station. Never mix sources silently.

  4. Explain

    Say what the evidence shows for a place, how sure we can be, and what it does not show, in plain language.

Follow one number

Watch an observation become a finding

Pick any province, territory or state to see one example. ICI takes a single published climate station near its largest city and walks one figure, the annual heating degree days, through every step: where it came from, what was checked, what it is compared with, and what can and cannot be said. One station describes its own location only. It does not stand for the whole province, territory or state, where conditions differ from place to place.

The station is chosen by one fixed rule for every jurisdiction: the climate-normals station with a published heating-degree-day value nearest the largest city in ICI's atlas (for Guam, the U.S. Virgin Islands, American Samoa and the Northern Mariana Islands, the station there with the longest complete record). The station is an illustrative example for this walk-through, not a measurement of the jurisdiction as a whole. Figures are shown in the publisher's own unit and base, for the 1991–2020 normals period. Sources: Environment and Climate Change Canada, Canadian Climate Normals 1991–2020; NOAA National Centers for Environmental Information, U.S. Climate Normals 1991–2020.

Know what you are looking at

Six kinds of information, always labelled

A number means different things depending on where it came from. ICI labels every figure with one of these, and never presents one kind as another.

  1. ObservedMeasured by an instrument or recorded by an agency: a station reading, a satellite image, a documented flood.
  2. ModelledCalculated from observations by a scientific model, such as a weather analysis or a population grid.
  3. EstimatedWorked out from other facts, such as a system's likely age from its installation year.
  4. ForecastAn official prediction for the coming hours or days, shown with its issuing agency and valid time.
  5. ProjectedA long-range scenario for future decades. A possibility, not a prediction for one year.
  6. ReportedWhat a visitor tells ICI about their own building, kept as their answer.

Same question, different kinds of answer

How cold is the heating season here?

Four honest answers to one question, each a different kind of information. They are not interchangeable, and ICI never passes one off as another.

Observed
A long-run average of measurements at one weather station. Real, but it describes the station's own location.
Estimated
The same station value applied to a building some distance away. Useful, but distance and elevation make it an estimate, so ICI names both.
Modelled
A value calculated for the exact spot from many stations, terrain and physics. ICI shows no modelled degree-day value yet, so no number is given.
Forecast
An official prediction for the coming days from the national weather service. A 30-year average is not a forecast for this winter.

The evidence framework

What ICI draws on, and its limits

  • Earth observation

    NASA Blue Marble satellite imagery (November 2004, about 500 m per pixel), national terrain and elevation, lakes, rivers and coastlines. A view of the setting, not current conditions.

  • Climate records

    1991 to 2020 climate normals from Environment and Climate Change Canada and NOAA. A station describes its own location; ICI uses the nearest suitable station and names its distance.

  • Weather and warnings

    Current official observations and warnings come from their issuing agencies and are shown as theirs, kept separate from any ICI explanation.

  • Geography

    Postal areas, ZIP codes and place names from Statistics Canada, the Canadian Geographical Names Database and the U.S. Census Bureau, so a search lands in the right place.

  • Hazard history

    Documented events such as Natural Resources Canada's historical flood records. History, not a prediction, and no record nearby is not proof that nothing happened.

  • Population and development

    Population, settlement and built-up area datasets are in scientific review. They appear on public pages only after their findings are validated.

  • Atmosphere and emissions

    Greenhouse gas and atmospheric research datasets are in scientific review, used for research until a finding is checked and approved.

  • Building records

    Equipment and recall notices from Health Canada and the U.S. Consumer Product Safety Commission, and the owner's own records of a building's systems.

Each page names the exact source, edition, period and licence of what it shows. Where a source covers only some provinces, territories or states, the page says so.

Published methods

Each public measure has its own method page

  • Building records

    The ICI Score

    What the documented records of a building's main systems support about their condition. Method 2.4, pre-ratification: not yet independently validated or certified. Unchanged by the new event-assessment research.

    How the ICI Score works
  • An area's climate

    Local Environmental Conditions Index

    How demanding an area's recorded climate may be for buildings, dimension by dimension, with the geography and sources shown.

    Read the method
  • A community

    Resilience profile

    Documented local hazard history and conditions for a community, with every gap named.

    Read the method
  • In development

    Community Resilience Index

    Documented hazard history and published community readiness across every state, province and territory. No scores are published yet: they will be, with the full method, only after validation.

    About the Index

Open where it matters

Transparent enough to check

Anyone should be able to see how a published figure or classification was reached: its sources, period, geography, assumptions, confidence and limits. That is published with every measure.

ICI's software, private systems and unpublished research stay protected. A finding reaches a public page only after it is checked against its sources and approved; until then it stays in research.

Correlation is reported as correlation. ICI does not claim that one thing caused another unless the evidence shows it.

Illustration: warm and cool ocean currents and storms around the Americas, with ports and coastal cities.
Illustration

See the evidence for a place

Search a postal area, ZIP code or city in Living Earth and see what has been measured nearby, and from where.