2000A&AS..142..325B

When planning new facilities, one is interested to know whether and by how much the new technique is superior to already existing ones. We describe a general approach permitting us to evaluate the relative merits of various techniques used in astrophysical observations, following the theory of model parameter estimation. It is applied to compare two-aperture optical long baseline interferometry to classical spectroscopy, both used to define the model parameters of the P Cyg wind. The wind modeling was done using an efficient approximation for computation of the line source function; it allowed us to analyse about 10⁵ points in the parametric space of P Cyg envelope models. It is shown that interferometry offers no real advantage if the object can be described by stationary spherically symmetric models with a priori fixed thermal structure. However, if the object must be described by a model with a large number of free parameters, e.g. when the thermal structure of the envelope is not fixed a priori, then the interferometric measurements can reduce the error in model parameters determination by an order of magnitude. In the case of P Cyg, the reduction of errors provided by interferometry is highest for the baseline lengths in the range 45-90m. This illustrates the capacity of the proposed method to be used for optimization of interferometric configurations. The simplifications adopted for this first trial application are critically analyzed, and future improvements are indicated.