Solitary electrostatic waves (SWs) were frequently observed by the Viking satellite in the auroral plasma. Previously, M¿lkki et al. [1993> reported occurrence statistics and voltage data for these structures. In this paper, we make a detailed study of the measurement process in order to establish what further knowledge can be extracted from the data. The possibilities that SWs are spacecraft-generated pulses or signatures of rectification of hypothetical wave packets are investigated and rejected. We analyze Langmuir probe data from several SW events and show that densities are at least as low as previously reported (a few cm-3). In order to assess the quality of the SW observations in this very tenuous plasma, we use a semiempirical model of the complete probe-spacecraft-plasma system, including nonlinearities in the probe sheaths. The usefulness of the model is shown by its ability to model details in the measured fluctuations. It is found that the errors in the voltage (electric field) measurements of solitary waves are very small, while the density fluctuation estimates may be severely affected by spurious signals. Therefore details in the shape of the observed probe current signature cannot be used to study the detailed shape of the density depletion. This limits the possibility of identifying weak double layer (WDL) characteristics in SWs, as effects of time evolution result in signatures similar to those of net potential drops.¿ 1997 American Geophysical Union