KUON Hatanaka
Compress / decompress RINEX observations with the Hatanaka format (.crx ⇄ .obs): free, no account.
Processing uses RNXCMP (GSI, Y. Hatanaka / crx2rnx · rnx2crx). As required for embedded use, we cite Hatanaka, Y. (2008), Bull. GSI, 55, 21-30. Lossless, but the decompressed file may not be byte-identical to the original (trailing-whitespace normalisation is per RNXCMP; the observation values are identical). Your data is auto-deleted after 24 hours. No account required.
RINEX conversion ・ RINEX viewer ・ all tools
Why RINEX gets compressed
An hour of raw observation runs to tens of megabytes, and an agency archiving hundreds of stations every day cannot move that. So Yuki Hatanaka's method was adopted, and it is now what the archives actually distribute.
It is not a general-purpose zip. It understands what is inside a RINEX observation file and stores the differences between epochs, which is where the redundancy lives.
The file names
.crx and .??d are compressed. The d in the old naming, where o means observation, marks the compressed form.
.obs and .??o are expanded. This is what analysis software expects.
You will often meet a double wrapper: .crx.gz or .??d.Z. Undo the outer archive first, then expand the Hatanaka layer.
When you need this
You downloaded observation data from a national or international archive and your software will not open it.
You are sending a long session to someone and it is too large to attach.
You are keeping years of your own base-station data and want the archive to stay a reasonable size.
What to know
The expansion is exact. Nothing is approximated and nothing is lost, so the file you get back is the file that went in.
Only observation files can be compressed this way. Navigation files hold orbits, not epoch-by-epoch measurements, so there is nothing for the method to work on.