In this study we use a numerical simulation of an artificial coronal mass
ejection (CME) to validate a method for calculating propagation directions and
kinematical profiles of interplanetary CMEs (ICMEs). In this method
observations from heliospheric images are constrained with in-situ plasma and
field data at 1 AU. These data are used to convert measured ICME elongations
into distance by applying the Harmonic Mean approach that assumes a spherical
shape of the ICME front. We use synthetic white-light images, similar as
observed by STEREO-A/HI, for three different separation angles between remote
and in-situ spacecraft, of 30{\deg}, 60{\deg}, and 90{\deg}. To validate the
results of the method they are compared to the apex speed profile of the
modeled ICME, as obtained from a top view. This profile reflects the "true"
apex kinematics since it is not affected by scattering or projection effects.
In this way it is possible to determine the accuracy of the method for
revealing ICME propagation directions and kinematics. We found that the
direction obtained by the constrained Harmonic Mean method is not very
sensitive to the separation angle (30{\deg} sep: \phi = W7; 60{\deg} sep: \phi
= W12; 90{\deg} sep: \phi = W15; true dir.: E0/W0). For all three cases the
derived kinematics are in a relatively good agreement with the real kinematics.
The best consistency is obtained for the 30{\deg} case, while with growing
separation angle the ICME speed at 1 AU is increasingly overestimated (30{\deg}
sep: \Delta V_arr ~-50 km/s, 60{\deg} sep: \Delta V_arr ~+75 km/s, 90{\deg}
sep: \Delta V_arr ~+125 km/s). Especially for future L4/L5 missions the
60{\deg} separation case is highly interesting in order to improve space
weather forecasts.
