Every claim in this course about reading the data comes down to one distinction: a tracker tells you what public sources currently say, not what is currently true. The two align most of the time, and the gap between them is where careful use of the data actually lives.
What a tracker genuinely is
A structured compilation of publicly available information — filings, permits, company statements, industry reporting — organised into a consistent schema and updated on a periodic cycle. Global Energy Monitor’s Global Gas Infrastructure Tracker, the source behind this site, does this work carefully and transparently, publishing its methodology and revising figures as better sources become available.
That is a genuinely valuable thing to have, and it is a different thing from direct measurement of the physical world. A tracker figure is always, at bottom, a claim about what a document said, filtered through a compiler’s judgement about which document to trust.
What follows from that
Coverage depends on disclosure. A project that generates no public documentation the compiler can find will not appear, or will appear only when someone reports on it. This affects small facilities, very early proposals, and projects in jurisdictions with limited public reporting requirements disproportionately, not because they matter less but because less is published about them.
Timeliness depends on the release cycle. Between one tracker release and the next, the world keeps moving — permits are granted, decisions are reached, capacity changes — and none of it appears until the next compilation catches up. A tracker snapshot is accurate as of when it was assembled, not as of the moment you read it.
Accuracy depends on the primary sources being right, and the compiler reading them correctly. Both usually hold, and occasionally one does not: a permit application overstates eventual capacity, a press release describes an aspiration as a commitment, a compiler carries forward an older figure past the point a newer one existed. This is rare, not structural, but it is not zero, which is why an implausible figure is worth checking against its source rather than dismissed or trusted automatically.
A status is a classification, not an observation. Reading the status field correctly means treating “operating” or “shelved” as the compiler’s best current judgement about where a project sits, informed by public evidence, rather than a directly witnessed fact — the compiler did not visit the site to check, and in the great majority of cases has no need to, because the disclosed evidence is normally sufficient. It is a judgement built on good evidence almost all of the time, which is different from certainty.
What a tracker is unusually good at, precisely because of this structure
A single well-maintained tracker, revised consistently against a stated methodology, is one of the few instruments that can show what changed and when — which is exactly what this site’s change log is built to surface. That capability comes directly from the periodic, sourced, disclosed nature of the compilation: a live sensor network would tell you the current state more precisely, but it would not tell you, against a comparable and citable prior state, what moved and in which direction. The trade a tracker makes — precision about the present for a citable history of change — is a genuine advantage for exactly the questions this course keeps returning to: not “what is the number today” but “what happened, and when did the sources say so.”
What this means in practice
Treat every figure on this site the way you would treat a well-sourced but dated newspaper report: almost certainly reflecting the real situation as of a known date, compiled with real editorial care, and not a live feed. When a number matters for a decision, check its release date, check what field basis is stated for it, and where the stakes are high, look for whether a more recent primary source exists that the tracker has not yet incorporated. That is not distrust of the source — it is the same discipline any careful reader of a periodic compilation would apply, and it is the discipline reading a capacity figure and reading ownership have already asked you to bring to the individual fields. This module asks you to bring it to the tracker as a whole.