The knowledge of temperature and emissivity of the receiver are both critical for a solar tower power plant, in order to guarantee an efficient operation of the thermal receiver on the one hand, while monitoring any degradation of the receiver coating on the other hand. To make these measurements, a new thermographic system is currently being developed, using a multispectral camera working in the short wavelength infrared spectrum. This system applies the principle of ratio thermography, using a couple of narrow bandpass filters centered on atmospheric water absorption bands, at 1.4 and 1.9 µm, to reduce the influence of solar reflections on the measurement signal, making it sensitive to atmospheric conditions. In this thesis, a batch simulation approach is used to identify boundary atmospheric and operating conditions necessary to achieve temperature errors below 2 %, minimizing the influence of solar reflection. Furthermore the influence of atmospheric parameters on the sensitivity of ratio thermography is analyzed, in particular the validity of the gray body assumption. It is shown that the atmosphere has a critical influence on the measurement accuracy. A humid atmosphere and/or high zenith angle is necessary for making accurate measurements. Furthermore only receiver temperatures above 450◦C could be measured for the current system configuration, regardless of atmospheric conditions. Assuming negligible solar reflections, the validity of the gray body assumption is shown to be sensitive to the precipitable water vapor. A model based atmospheric compensation is therefore required to further improve the accuracy of ratio thermography.