IMPORTANT INFORMATION:
This version (V3) of the dataset is based on aerial imagery from 2024 (month depends on department) and cadastral data from January 2025. It features:
updates for 27 departments: 01, 02, 04, 11, 17, 23, 24, 29, 30, 33, 34, 38, 40, 47, 56, 60, 62, 64, 66, 67, 68, 73, 80, 84, 87, 2A, 2B
7 new departments: 77, 78, 91, 92, 93, 94, 95
Data for other departments is not reuploaded in this version of the dataset. Please use the version selector and goes to V2 to access older data for these departments.
This dataset is the result of a convolutional neural network (CNN) trained for the detection of Rooftop Photovoltaic systems. It contains the geographical locations of the Rooftop PhotoVoltaic systems in France as predicted by the model.
This dataset is related to the scientific publication "Thebault, Nerot, Govehovitch, Ménézo - A comprehensive building-wise Residential Photovoltaic system detection in heterogeneous urban and rural areas: application to French territories" Applied Energy, 2025, doi.org/10.1016/j.apenergy.2025.125630
It contains the results of applying a CNN classification model to the buildings of each French department. The results are provided as geospatial vector data (in shapefile format: Shapefile). Among other information described in the Readme file, the 'Score' column/attribute corresponds to the predictive model's output for the corresponding building, representing the likelihood of a PV system being present. The score is a scalar value between 0 (the model is certain that there are no PV system) and 1 (the model is certain that a PV system is present). In the associated publication we consider a threshold of 0.5 (if Score 0.5, we consider that a PV system is present). As discussed in the associated publication, this threshold value can be adjusted depending on the application of the results.
Important information
1) A large part of the French metropolitan buildings are covered. In a further version all the French departments should be covered
2) Reliability of the results : It is important to refer to the associated publication to understand how reliable are the results. The 'Score' discuss above provide a quantification on the model's confidence in his classification. Additionnaly, the user can refer to Figure 18 and 19 of the associated publication to assess whether the models is in accordance with official registers of the energy provider.
sgis, 0.1
tensorflow, 2.9.3
