In risk analyses, two components are taken into account: (1) hazard analysis, including susceptibility and temporal occurrence, and (2) consequence analysis, including characterization of elements at risk (EaRs) and their vulnerability. This study focused on characterization of EaRs, in which items are spatially displayed and impacted by natural events. Several methods can assess these EaRs through expert or engineering approaches. Among the expert approaches, the multicriteria method is more flexible and allows integration of a wide range of information in order to characterize and discretize different EaRs. Traditionally, the mapping and criteria accuracy of EaRs is the same at all spatial analysis scales, while the hazard accuracy changes according to the spatial scales. Therefore, we propose an approach based on the selection of different information/criteria among several private or open access multiple geographical databases to adapt the mapping and criteria accuracy of each EaR according to the hazard analysis spatial scale. After harmonizing the different databases and merging them under GIS, a single database per work scale is created through a specific procedure with interoperability of results between scales. Thus, the number of criteria used to describe these EaRs will depend on the scale of work and the spatial scale of the analysis. To develop and test the transposability of this method, three experimental coastal study sites subject to several hazards (multirisk) have been selected in Normandy (France) with error estimations ranging between 10% and 20%. Subsequently, these data can be integrated into risk and multirisk analyses.