In this paper we compare various methods of calibration that can be used in practice to improve the accuracy of reduced-order models based on Proper Orthogonal Decomposition. The bench mark configuration retained corresponds to a case of relatively simple dynamics: a two-dimensional flow around a cylinder for a Reynolds number of 200. We generalize to the first and second-order the method of calibration based on Tikhonov regularization recently used in [1]. Finally, we show that for this flow configuration this procedure is the most effective in terms of reduction of errors.