In the present paper, the identification of the material parameters of a masonry lighthouse is discussed. A fully non-invasive method was selected, in which the material properties were determined via numerical model validation applied to the first pair of natural frequencies and their related mode shapes, determined experimentally. The exact structural model was built by means of the finite element method. To obtain experimental data for the inverse analysis, operational modal analysis was applied to the structure. Three methods were considered: peak picking (PP), eigensystem realization algorithm (ERA) and natural excitation technique with ERA (NExT-ERA). The acceleration’s responses to environmental excitations, enhanced in some periods of time by sheet piling hammering or by sudden interruptions like wind stroke, were assumed within the analysis input. Different combinations of the input were considered in the PP and NExT-ERA analysis to find the most reasonable modal forms. A number of time periods of a free-decay character were considered in the ERA technique to finally calculate the averaged modal forms. Finally, the elastic modulus, Poisson’s ratio and material density of brick, sandstone and granite masonry were determined. The obtained values supplement the state of the art database concerning historic building materials. In addition, the numerical model obtained in the analysis may be used in further cases of structural analysis.

The efficacy of modal curvature approach for damage localization is discussed in the paper in the context of input data. Three modal identification methods, i.e., Eigensystem Realization Algorithm (ERA), Natural Excitation Technique with ERA (NExT-ERA) and Covariance Driven Stochastic Subspace Identification (SSI-Cov), and four methods of determining baseline data, i.e., real measurement of the undamaged state, analytical function, Finite Element (FE) model and approximation of current experimental mode shape, are considered. Practical conclusions are formulated based on analysis of two cases. The first is a laboratory beam with a notch and the second is a stone-masonry historic lighthouse with modern restoration in its upper part. The analysis shows that NExT-ERA and SSI-Cov in combination with approximation of current mode shape provide high efficacy in damage localization alongside relatively straightforward determination of baseline data. It proves that the construction of advanced FE models of a structure can be replaced with a much simpler method of baseline data acquisition. Furthermore, the research shows the structural mode shapes identified with ERA may not always indicate the presence of damage.