|| Checking for direct PDF access through Ovid
We operated identical Low-Light level TV cameras to observe the Leonid 1999 meteor storm, both from a ground-based station in Southern Spain and from the ARIA airplane of the Leonid Multi-Aircraft Campaign. The ground-based camera was pointed to a fixed position about 50° from the zenith, the airborne camera was pointed through a window at 75° from the zenith. During the peak of the Leonid storm, the two cameras were located between 10° and 20° apart in geographical longitude. The recorded meteor numbers differed by a factor 5.3 ± 0.4, the airborne camera recording the higher rates. This is much more than what could be expected from the geographical separation of the cameras. The different elevation angles and altitudes of the cameras can explain this. Pointing the camera low to the horizon results in a much larger volume in the atmosphere which is observed, resulting in higher meteor count rates. However, the meteors are on average much further away than when observing high above the horizon. The atmospheric extinction reduces the brightness of the meteors, effectively reducing the count rates. For two ground-based cameras, these two effects are expected to compensate. Due to the high altitude of the airborne camera, the increasing effect of the count rate dominates. We set up an atmospheric extinction model taking into account Rayleigh scattering that quantitatively explains the number difference. Using the same model, we predict number differences for cameras observing from the same location, but pointed at different elevation angles. For typical observing conditions, neglecting this effect can result in differences up to a factor of 10 in the derived meteor numbers.