The published record is permanent

Once an article is published and assigned a DOI, it is not removed, altered, or replaced. Other work may already cite it, and the scholarly record depends on a citation continuing to resolve to the thing that was cited.

When something is wrong, we correct it in the open, by publishing an additional record — never by quietly editing the original.

Corrections

A Correction is published when an error affects the interpretation or reuse of the data: misreported coordinates, a transposed value, an incorrect unit, a mislabelled figure, a misattributed method.

The Correction is a citable record with its own DOI, linked from the original article. The original remains available and unaltered, and carries a prominent notice linking to the Correction.

Anyone may report an error — an author, a reader, or the Editor. Write to scott@earthaccess.us with the article and the specific problem.

Data updates

Data are sometimes superseded rather than wrong: a sample is re-dated with a better method, a location is resurveyed with better instrumentation, a dataset is extended.

This is not an error and it is not a retraction. Where the underlying dataset lives in an external archive, the archive's own versioning governs, and the article will link to the current version. Where an update materially changes what the article says, we publish a Correction.

Retractions

A Retraction is published where the findings or data are unreliable or the work is unethical: fabrication, falsification, plagiarism, data that cannot be substantiated, or publication without the consent of a co-author or the data's owner.

A retracted article remains online, watermarked as retracted, with a linked Retraction notice stating who retracted it and why. It is never deleted — removing it would break every citation to it and conceal the record.

Expressions of Concern

Where a serious concern is raised but cannot yet be resolved, the Editor may publish an Expression of Concern linked to the article, and will resolve it as a Correction, a Retraction, or a withdrawal of the concern once the matter is settled.

Authorship and consent

Every author must have consented to being listed, and a faculty advisor must have attested to the work. A dispute over authorship or over the right to publish a dataset is grounds for an Expression of Concern while it is investigated.

Who decides

The Editor. Where the Editor has a conflict of interest — including any article the Editor co-authored — the decision is referred to a member of the Editorial Board who does not.