Introduction A witty saying proves nothing. - Voltaire This PhD thesis deals with the most common method of studying the interaction between the earth's upper atmosphere and magnetosphere-solar wind system: imaging of the aurora. A novel and elaborate analysis is presented which allows previously untapped information to be retrieved. What is unique about multi-station imaging is its capability of determining the position and motion of imaged objects with stereoscopic triangulation, and that it makes it possible to estimate the inner structure of transparent objects by tomographic inversion. Scientific multi-station imaging of the aurora dates back at least to the early 20th century when Störmer used two cameras to image auroral arcs simultaneously from two separate sites. From these images the altitude of the lower border of the aurora was determined for different types of aurora. The next step in optical multi-station measurements of aurora was the two-point measurements with scanning photometers made by Romick and Belon (1967), measurements they used to derive the altitude distribution of the auroral luminosity in a north-south slice. The third step for this type of observations was the use of general tomographic inversion, to obtain the spatial distribution of airglow from satellite photometer data first used by Solomon et al. (1984). Later Aso et al. (1990) determined the spatial distribution of auroral arcs from stereo observations with TV cameras. In 1991 Valance Jones et al. (1991) determined the spatial distribution of several emissions with data from three scanning photometers and rocket-borne photometers. The state-of-the-art system in this field of research is the Auroral Large Imaging System (ALIS) that will be described in some detail in the next section.