The FarmBuiLD Model (Farm Building Landscape Design): a Sampling Methodology for the Parametric Analysis of Historical Typologies

Within a broader research, the authors have worked out the FarmBuiLD research model (Farm Building Landscape Design), aimed at defining farm building design criteria able to combine improved architectural and landscape quality with functional efficiency. Since historic rural buildings are broadly associated with widespread semiologic and aesthetic values, FarmBuiLD uses ad hoc analytical and metadesign tools to carry out a comparative typological analysis of historic and recent rural buildings. These tools allow to interpret traditional physiognomic characters from a contemporary point of view and in future perspective, both in terms of functions and building techniques and know-how. FarmBuiLD assumes that a proper relation between modern rural buildings and landscape can be reached through specific criteria based on designing consistently or consciously breaking with the historical typologies, according to architectural quality and functional requirements. This work focuses on the physiognomic characterization of historic rural buildings, through the implementation of a set of synthetic architectural parameters related to shape characters. This quantitative assessment approach calls for high-detail surveys of the main geometric building features to be systematically performed, thus often entailing considerable costs, also due to the high density of the rural built environment over wide regions in Europe. The general goal of this paper is to define and test a repeatable and flexible methodology, suitable to be adapted to various geographic contexts, aimed to maximize the efficiency of the above-mentioned analyses while allowing the results to meet the precision standards of the model. The specific goals are: -to develop a method aimed at defining suitable samples of rural buildings for the typological analyses, capable of considering the various characteristics of the different study areas; -to work out a suitable procedure for surveying the data about the architectural features of the sampled buildings. With reference to an Italian study area, varied in terms of agricultural systems and landscape characters, the authors defined a stratified random sampling design based on sound statistical inference techniques, which proved suitable to obtain results that can be extended to the whole building heritage under study, meeting predefined reliability standards. We defined the sampling frame and the most appropriate geographic and architectural variables for its stratification, based on the elaboration of base information about typological classification and location of rural buildings. The GIS implementation of such phases allowed to calibrate the method and define a pilot sample, suitable to carry out the physiognomical parametric analyses through photogrammetric surveys.