Attached are plots of CI-D and CI-A vs. altitude, without any filtering done on them in terms of latitude or water vapour vmr. The black diamonds are
daytime measurements and then red are nighttime. Both show data for the whole day of 15 August 2003.
The CI-A plot behaves in a manner that we expect, flagging cloud (ie. CI-A <=1.8) at lowish altitudes (ie. 5 - 20 km) and not flagging cloud in higher
regions where there shouldn't be cloud in any case (ie. 20 - 30 km). Above 30 km, the definition of cloud index breaks down anyways due to noise, so it
doesn't matter that the CI-A values reported for such higher altitudes have low values.
The CI-D plot is less convincing. "Dodginess" and stuff. CI-D values seem to remain at fairly low values even in the range of 20-30km where there should be
no cloud registered. Very high altitude flagging is explainable by PSC's, but seems as if,overall, CI-D is less reliable in registering cloud presence due
to all of the positive flagging where there should be none.
Some profiles show very low (1e-10) water or ozone values. Shown not to be correlated with the cloud flag in either channel A or D. Looking at latitude / temperature dependance of D band cloud flag.
Using the [CFI] software to calculate MIPAS viewing geometries. Aim to design observation mode to sample many local solar times - at present MIPAS always samples the same latitude at the same local solar time. (i.e. its orbit has a fixed equator crossing LST.) Figure [here].
Aim at the moment is to investigate the range of LSTs that can be sampled
with the various MIPAS azimuth scan positions.
