michaelkoopmans asked: After analyzing Waveglider Ben's significant wave height estimations vs. the Half Moon Bay buoy no. 46012 during 11/17 18:00 - 11/19 22:30, I found a data set ~3 weeks old was much worse than one I downloaded yesterday. What kind of modifications were made to the data set this long after recording and why?

The datasets for that time period haven’t changed (I hope!) The did get regenerated a few times, but the conversion code for those instruments is the same.  During the time that the craft were in Monterey Bay, the manufacturers of the Datawell sensor sent us new firmware that fixed a calibration problem affecting wave direction only.  That was installed on December 7th.  Does this sound like what you’re seeing?