A GIS-statistical approach for assessing built environment energy use at urban scale

This paper illustrates the development of a geospatial bottom-up statistical model to estimate the energy consumption of a large number of residential building stocks for heating space, considering a wide range of variables. The proposed methodology is based on a 2D/3D- Geographic Information System (GIS) and Multiple Linear Regression (MLR), which provides location-based information for each single dwelling to identify correlations and assess the demand-side consumption at the urban scale. This framework was tested on a medium-sized Italian city, including around 3600 residential buildings. The results provided by the model are validated using residual analysis and cross-validation. Moreover, the spatial results provided by this study represent a useful tool to aid decision-makers in the urban planning process. These results can help to create future energy transition strategies, implementing energy efficiency and renewable energy technologies in the context of sustainable cities. This work is part of a national Smart City & Communities project, named “EEB- Zero Energy Buildings in Smart Urban Districts”; nonetheless, the methodology illustrated in this paper can be generalised and applied to other European urban contexts.