The y-axis is indeed height and it is labeled in KM AGL.

All of the images that we presently have on the DAAC are 6 hour
images with the start time indicated at the top next to the date.
The radar pointed vertically for 24 minutes out of every 30, so the
short regular gaps indicate time when the radar was scanning in
either elevation or azimuth.  Longer, irregular gaps indicate
periods when the radar was off.  If I remember correctly, we have
a six hour .gif image for any time period that had any data at all,
but the radar might not have been on for that entire period.  
If a .gif image is completely missing, it is because the radar 
was off for that entire 6 hour period.
 
The radar field that is displayed should be reflectivity.  There is
a color bar at the bottom of the figure which indicates what
color corresponds to what radar reflectivity in dBz.  Please note that
if you compare our radar images to PSU's images, the reflectivities
will NOT match up.  PSU had some calibration problems during FIRE II
so they could not calculate accurate reflectivities.  We worked with
them quite a bit on this matter, and agreed that the ETL reflectivity
values were the more correct ones.

Other products that you might be interested in:
 1) Similar time-height cross-sections of reflectivity for
    30 min periods, rather that 6 h periods which allows the examination of
    the cloud structure in far greater detail.
 2) ASCII files which contain time series of cloud boundary heights,
    and which keeps track of multiple layers.  (i.e. base1, base2, top1,top2, etc)
 3) ASCII files which contain profiles of microphysical data
    including particle size, particle concentration, and IWC which were derived
    using Dr. Sergei Matrosov's technique of combining radar and radiometer
    data. (For select periods during FIRE II)