Physical climate risk screening for ESG reporting

What is a physical-risk screening step?

A screening step is an organized way to decide what to investigate next. Begin with a facility register and a clear question: which operating locations need more evidence about flood, heat or other hazards? Keep the supplied facility identity intact, then attach area context in columns that can be checked independently. A category is useful when the reader knows which hazard and geography it describes.

Agree on how findings move into a research queue before reading the labels. A county flag might prompt a map check, an operational interview or a request for building information. Record the reason for that action rather than presenting the flag as a final judgment. This makes the screening useful to the person who must collect the next piece of evidence.

For work discussed under IFRS S2, ESRS E1, TCFD or ISO, check the framework text your report follows. County indicators can help identify hazards for initial research; they do not complete scenario analysis or a reporting assessment. Retain your team's interpretation and approval separately from the county extraction.

Where does county data fit?

County context fits alongside a facility register as an area evidence layer. Match each location to a verified county identifier, preserve that identifier as text and keep the original facility ID. Facilities in the same county may legitimately share a county record. Their operating conditions and vulnerability can still differ, so repeated county values should remain visibly county values.

The table maps downloadable columns to suggested register destinations. The destination names are working suggestions, not a prescribed reporting template. Add source release, source URL, checked-on date and review status to your own register. Do not silently rename county annualized loss as facility financial exposure. Keep monetary units in the column label when a value travels into another workbook.

Total county expected annual loss is $2,217,702,383 per year.

Use the sample comparison to test the join before extending the register. Inspect an unmatched location as carefully as a matched one. A missing county match is a coverage problem that needs an owner, not a reason to remove the facility from the review population.

Which facility details require separate evidence?

County grain does not show where a building sits within an area or how that building operates. A facility near a river and another on higher ground can inherit the same county rating. The county record does not distinguish their floor elevations, drainage, equipment placement or continuity arrangements. Keep those questions attached to the facility rather than interpreting a shared label as shared site conditions.

The same issue appears in portfolio concentration. Several locations attached to a county total do not represent several independent estimates of loss. Avoid adding the repeated county total across the facility rows. Summarize where facilities cluster and retain the county figure once as area context. Asset-level valuation belongs in another field with its own provenance.

Which hazards does FEMA NRI cover?

FEMA National Risk Index is the source of the hazard fields used here. The reviewed sample includes coastal flooding, riverine flooding, wildfire, heat wave, hurricane, earthquake and tornado ratings. This is the sample's reviewed field selection; it should not be mistaken for an exhaustive inventory of every physical climate issue relevant to a business.

Earthquake belongs to natural-hazard context and should not be described as a climate-driven hazard. Keep that distinction when writing the report narrative. Retain coastal and riverine flooding separately, and preserve the publisher's labels even when the output contains No Rating or Not Applicable. Neither label supplies a substitute low-risk finding.

A useful screening note identifies which requested hazards have evidence and which remain outside the reviewed extract. Supplier dependencies, water availability and operational sensitivity may prompt additional research. Describe those as open questions unless you have separate, traceable evidence for them.

What evidence should I retain for review?

Retain the original field name, displayed value, unit, geographic unit, county identifier, source release and source agency. Record the date your team checked the value separately from the period described by the source. Attach the question the evidence supports and a short statement of its limit. A reviewer should be able to follow the trail without guessing what a renamed spreadsheet heading means.

Keep an exceptions log for malformed identifiers, unreviewed counties, missing evidence and disputed matches. Assign a review owner and a concrete next action. Preserve the original input alongside any correction so the reason for a changed join remains visible. A corrected county code can change the attached evidence even when the supplied address text looks unchanged.

Before a narrative is approved, select a facility and trace it back through the register to the source field. Check that the statement says county context and that unresolved items remain visible. Save the reviewed export with its input version, then record who accepted the wording. A reproducible extraction is useful evidence of process, while the report conclusion still depends on the wider assessment.

Which decisions require additional evidence?

An evidence register documents the screening process; it does not demonstrate that a reporting requirement has been met. Review the chosen framework, reporting boundary and decision horizon separately. Assign unresolved source and location questions to a reviewer, and retain the evidence needed to explain how each screening finding was reached.

How should I cite the evidence and review date?

Last reviewed: 2026-10-23. Source: FEMA, FEMA National Risk Index. The county sample supplies the worked values. Keep the release and field notes from the methodology with exported figures. The review date records when this guide was reviewed; the underlying observations have their own periods.