ESG Screening

Blog

How to reconcile self-reported ESG site data against satellite imagery

A forested hillside partly obscured by low cloud and mist.

A sustainability report says a mine's tailings footprint hasn't changed in two years. A CSRD filing says zero hectares of forest loss in the concession area. Your stewardship file has no way to check either claim except to ask the company again and wait for the next report. That's the gap reconciliation is meant to close: comparing what a holding tells you against what's physically there.

This isn't about catching every company in a lie. Most discrepancies come from lagging data, rounding at the site level, or a subsidiary that didn't make it into the parent's disclosure. But a few discrepancies are the ones you need to find before a shareholder vote, an engagement call, or a disclosure review, and self-reported data alone won't surface them.

What reconciling actually means at the site level

Reconciliation means picking the handful of claims that are both material and checkable against overhead imagery, then comparing them on a cadence that matches your review cycle rather than the company's reporting calendar.

The claims worth checking usually sit in four categories:

  • Deforestation and land clearing near a concession boundary, where disclosed figures can lag actual clearing by a full reporting cycle.
  • Tailings dam footprint, where a "no material change" statement is either true or it isn't, and the dam's outline is visible from above.
  • Gas flaring, where thermal signatures show activity that doesn't always match what's logged in operational reports.
  • Encroachment on buffer zones or protected areas adjacent to a site, which tends to show up as gradual change rather than a single event.

None of these require field access. All four show up as visible change in medium-resolution multispectral imagery, the kind that comes from routine satellite passes rather than a one-off commissioned flyover.

Building a repeatable screen instead of a one-off check

A single comparison tells you about one quarter. A portfolio-wide screen run on the same cadence every quarter tells you whether a holding's trajectory matches what it's reporting, and whether that trajectory is getting better or worse between disclosure cycles.

A standing record like that holds up during an engagement call or in a voting rationale in a way a one-off check doesn't. A quarterly map-layer screen across your portfolio-company sites, flagging deforestation, tailings-dam footprint change, flaring and encroachment against a per-holding risk map, is what ESG Screening is built to deliver, as a check that runs alongside the company's own disclosure process and shows whether that process is keeping up with what's on the ground.

What to do when the numbers don't match

A mismatch is a starting point for a question, and the first thing to check is dates. Imagery timestamps and report reference periods rarely line up exactly, and a few weeks' difference can explain a lot. The useful next move after that is a direct, specific question to the company's investor relations or sustainability team about that site and that change.

Where the gap holds up after that conversation, it becomes part of the engagement record. A clearing event that shows up in imagery six months before it shows up in a disclosure is worth noting even if the company eventually reports it accurately. Pattern matters more than any single quarter, and a pattern is only visible if you've been checking consistently rather than reactively.

None of this requires in-house remote sensing expertise. It requires a standing process that runs on its own schedule, flags change against a baseline, and leaves you with a record rather than a one-time answer. If you're building that process for your portfolio, start by seeing what a quarterly screen would surface for your current holdings.

← Back to the blog