Field data are always, even if ground truth is available, subject to ambiguity and consequently a matter of interpretation. Synthetic models can more objectively be used to compare different software, their potential and drawbacks. Here we can collect such benchmark models to test our algorithms. In the best case they are published but can also just be well described here. Apart from the synthetic data and some Figure and Table about the underlying model, this should also be saved for numerical comparisons.