A summary of GV cabin pressure changes and what they are caused by
From Britt Stephens, NCAR
If I remember correctly, there are essentially 2 pressure control options. The first is to let the GV computer handle the pressure control, using an input of the destination field elevation in this process. The second is to manually set a cabin pressure altitude and leave it the whole flight. (There is a 3rd option of having the pilots manually adjust the pressure set point throughout the flight but this adds to their workload in unsafe ways and in practice during START-08 they were not always able to remember to make the adjustment which led to larger changes). In either of the options available to us, if you descend below the set point pressure, the plane can not maintain a negative pressure difference and the pressure in the cabin rapidly climbs to outside pressure, only recovering once the plane climbs above this height again. The only way to avoid this is to set a destination elevation or constant set elevation of sea-level, but I *think* someone (Fred maybe) can't handle sea-level pressure so requested it be set higher. I also think there may be some advantage of the former option (setting destination field altitude) in that the computer can try to anticipate this threshold and adjust the cabin pressure more slowly, but again, the problem of descending below the set point is still there.
We have several cabin pressure measurements (one from Shuka that is in the final data set, and one from me that is only available now in my online plots) that you can look at - in practice there are rapid but small level shifts (when on auto control), and larger spikes (when descending to sea-level) of various magnitudes. You can also see the same plots from the START campaign on my web page (http://www.eol.ucar.edu/~stephens/AO2), but note that many of these flights only descended once at the end so had very constant pressure. Again, as far as I know the NOAA systems are the only ones that are sensitive to cabin pressure so you should feel free to request whichever of the two control modes that works best for you.
From Fred Moore, NOAA/CIRES
UCATS and PANTHER pressure controllers do not work when the cabin pressure altitude goes below sea level (say 1100 mbar) which happened on START-08 and on HIPPO. They also have problems with fast jumps from one press to another (step functions).
Each time I asked about it I was informed that the pilots were trying out different things per Britt's request. My head set does not connect to the pilots so I do not know what conversations went on between Britt and the pilots. There was one very high press dive on HIPPO that also caught Dale's attention and Jim Elkins is mostly referring to this event. The pilots were aware of the problem and by the last set of hippo-1 flights I was of the impression that they had a better idea of how the controls work and that the cabin pressure was under control. Jim wants to make sure that those are the conditions we fly under during HIPPO-2.
For us the bottom line is, no pressurization above 1000 mbar and no step functions during the flights are preferred. Presumably we are not the only ones with this desire.