Use of data derived by different sources for the seismic vulnerability assessment of current building stock in GIS environment: an application to the municipality of Bisceglie, Italy

The paper presents a methodology that, starting from the extraction, integration and elaboration of data from different sources, is able to construct a geo-referenced cartographic and descriptive database in a GIS environment and to perform the seismic vulnerability of current residential buildings at a large scale through indirect, empirical algorithms. An application has been developed for a case study in Puglia, Italy: the city of Bisceglie. The starting point is represented by existing cartography and geospatial database, from which the data about the building stock under investigation are collected, catalogued and implemented in the GIS environment. Two level of GIS entities have been defined: “census section” and “urban block”, to which a “standard” set of minimum data has been associated (according to this definition, such data are by default available for almost any building in Italy). Subsequently, the georeferenced database has been supplemented by more detailed information extracted by “CARTIS” catalogue1, which is a database containing information about the typological and structural features of homogenous urban districts. This catalogue has been developed in the framework of the RELUIS project 2014-2018 and in Puglia, collects data about 13 municipalities investigated by the research unit of Politecnico di Bari-Dicatech, among which Bisceglie. In this way, it has been possible to refine the data and to obtain more detailed information associated to each building. Data have been collected in alphanumeric format ad processed within the GIS environment in 2D and 3D representation obtaining an integrated, easily accessible georeferenced database. The format is particularly useful for the automatic implementation of vulnerability methods of indirect type, that provide a qualitative seismic vulnerability index at different scales (a whole urban district, an urban block, a single building). Different types of data have been processed at different scales of analysis, and the results obtained have been compared to evaluate the reliability of the data set and the level of approximation. Finally, the collected data and results have been validated by using detailed information gathered on a proper sample of buildings, which have been filed and analysed one-by one in detail. The procedure allowed to verify the reliability of different types of data, and to perform a rapid and automatic assessment of the seismic vulnerability at urban scale, with a low computational effort. Moreover, it has been possible to obtain 2D and 3D elaborations that are very effective in order to provide an immediate, meaningful “picture” of the current state of existing residential building stock.