The middle, and higher atmosphere is a very interesting place from the point of view of atmospheric ionization processes, but unfortunately, due to the difficult accessibility of that altitude range, there are just a few measurements carried out so far. One of the most successful application tools for this purpose is the Gerdien condenser, which is capable to distinguish the different ion groups by the applied voltage. Such investigations are especially interesting at high latitudes, where galactic cosmic rays participate in greater degree in ionization processes.
The aim of the present thesis is to give a brief description of all of the tasks, related to every aspect of the evaluation of the altitude measurements with the above mentioned Gerdien condenser, starting from the noises caused by the environmental effects, all the way to a complete simulated test. The instrument was built for a specific experiment, and it was actually used in a sounding rocket. The aim of the experiment is to determine several ion mobility spectra with the concentration of each ion groups at different altitude ranges. The paper discusses the physical background and the limitations of the measurement as well. The main challenge is that the only way such a measurement can be carried out at the range 25-80 km, is to measure on a sounding rocket, or a research balloon, where the latter has a limited altitude range. Therefore, not only the measurement time will be limited, but possibly a lot of noise will distort the measured data. The primary objective of the thesis is to explore these factors regarding the distortions of the instrument as well. Such a definition is very important to be able to define a measurement model, which is essential to handle the measured data. Another problem is that we can not validate the defined model accurately, so apart from a few exceptions we have to rely on probability. Therefore, probability theory gets large role in this paper. Also other commonly used methods will be described to determine the uncertainties of measurements as well, such as GUM, Maximum-Likelihood method, or Monte Carlo simulation. After that a possible processing algorithm will be outlined, that can perform the whole measurement evaluation process automatically. To demonstrate the usability of the algorithm a test will be performed on a simulated measurement.
By the end of the paper most of the possible disturbing effects will be described which enables to determine a full model of the measurement. After that the uncertainties will be defined and a confidence interval can be constructed. Finally, we use these results and a simulated data to see the capability of the evaluating process.