4. Conclusions
A numerical model, used to predict the thermal behaviour of stratospheric balloon experiments, can be updated through a limited number of ground tests. When the environmental conditions are well described, but contain a number of uncertain parameter values, the updating routine delivers a major accuracy improvement by a factor of 2.2 to 7.6. This method can also be used for other well-known complex environment estimations as a replacement for expensive, long duration and/or dangerous experiments. The boundary conditions of the method are that the uncertain parameter values are independent of the experiment environment. Moreover, the methodology can be used to validate experimental designs for space applications and determine the failure risk of different components or design versions in a fast and cost-efficient way.