The Humidity of Lies: Why Your Data Should Be Waterproof (Or It's Not Science)

Acts of God as the Ultimate Get-Out-of-Jail-Free Card

The "dog ate my homework" excuse used to be the punchline of academic jokes. Today, it has been upgraded to a catastrophic environmental event. We are witnessing the rise of the force majeure defense in publishing ethics: if you can’t explain why your Western blots look like a poorly stitched quilt, just wait for a pipe to burst.

Let is be clear about the reality that the institutional gatekeepers avoid. A flood is a logistical tragedy, but it is never a scientific justification for keeping fraudulent records on the books. If the raw data is gone, the paper must be gone. Period. The fact that we are debating whether or not a flood actually happened is a distraction from the real systemic failure: the total lack of mandatory, decentralized data archiving.

The Ghost in the Gel: When Metadata Outlives the Lab

Look at the recent mess involving the American Journal of Physiology. It is a classic bit of theater. Figures were flagged on PubPeer for weird vertical stretches and duplicated lab dishes, and suddenly, they are orphans. The University of Kansas Medical Center pointed to a 2012 flood, which is a very convenient shield for missing evidence. But as Retraction Watch noted, the water damage does not explain the stuff we can already see. You do not need a soggy lab notebook to spot a digital original that has been cooked.

Data integrity sleuths like Kevin Patrick (aka Actinopolyspora biskrensis) are the only reason these cases come to light. They aren't looking at the physical notebooks submerged in a basement; they are looking at the digital ghosts left behind in the PDF. When an image appears three times with different brightness levels, it doesn’t matter if the original hard drive is at the bottom of the ocean, the artifact itself is the confession.

The Infrastructure of Irresponsibility

Our obsession with physical lab presence as a tool for validation is a 20th century antique. It is a gift to every sloppy researcher in the world. We have allowed a culture where losing your data is a valid excuse rather than a professional disqualification. If a bank loses your records in a fire, they do not get to keep your cash. So, why do we let authors keep their citations when the floor of the study has washed away? It is an absurd double standard.

Radical Structural Reform: The 'Cold Storage' Mandate

To end the era of the "Waterproof Lie," we need two radical shifts in how we define a published work:

The Immutable Ledger Requirement: Journals must stop accepting "the lab was flooded" as a reason for missing data. If the raw, unedited TIFF files and metadata are not deposited in a third-party, silver-standard repository (like Figshare or Zenodo) at the time of submission, the paper should be considered a preprint, not a publication. Permanent status requires permanent data.

The Retraction-by-Default Clause: We need to flip the burden of proof. If a paper has credible signs of trickery and the author cannot produce the goods (whether they blame a flood, a fire, or bugs), it gets pulled. A retraction is not always an accusation of faking it. Often, it is just a clear statement that the work is no longer verifiable. It is dead weight.

We must stop treating scientific records as private property kept in a desk drawer. Science is a public trust. If your data isn't waterproof, it isn't science, it’s just a story that got wet.