High-resolution polygons of photovoltaic (PV) panels from orthophoto segmentation in Austria and electricity output modeling

Overview This dataset comprises a nationwide inventory of photovoltaic (PV) panels in Austria, automatically extracted from high-quality orthophotos (20 cm ground resolution) acquired between 2021 and 2023. It includes both rooftop and ground-mounted installations. Panel orientation was derived from a digital surface model. Hourly energy production for 2004–2023 was simulated using pvlib in Python. For a full methodological description, see the thesis Automatic segmentation of photovoltaic panels in Austria and modelling of its impact on landscapes and the energy system (Wiesenhofer, F., 2025): . Note: The results reported in the thesis (particularly the modeled outputs) differ from those in this dataset because the horizon shading was updated. Vector Polygons of existing PV Systems Resolution: Vector polygons based on high-resolution input data (ortho images and digital surface models) Projection: EPSG:31287 Extent: Austria Time period : 2021–2023 Format: GeoPackage (GPKG) Column description # Column Name Data Type Unit Description 1 fid Numerical (int) - Unique identifier for PV panel cluster (matches the CSV) 2 PV_type Categorical (str) - Type of PV installation (Rooftop PV or Ground Mounted PV) 3 PV_name Categorical (str) - Name of the PV installation for PV power plants > 1 ha 4 NUTS_2_eng Categorical (str) - Federal State of PV panel cluster in English 5 NUTS_2_ger Categorical (str) - Federal State of PV panel cluster in German 6 county_ger Categorical (str) - Name of the county in German 7 panel_area Numerical (int) m² Area of the PV panels 8 base_area Numerical (int) m² Base area of PV panels (polygon area) 9 landuse Categorical (str) - Type of land use (e.g., building, farmland, meadow, other ...) 10 biodiv Categorical (str) - Biodiversity rank (−1 to 3); higher values indicate greater impact on biodiversity 11 aspect_str Categorical (str) - Cardinal direction of the panel orientation (e.g., ESE, SSW) or ‘Flat’ for flat panels 12 aspect_deg Numerical (int) ° Aspect of the PV panel cluster in degrees (north = 0, clockwise); flat = 500 13 slope_str Categorical (str) - Slope category of PV panel cluster (Flat, Slight Tilt, Steep Tilt, Two Sided) 14 slope_deg Numerical (int) ° Panel slope in degrees; flat = 0 15 elevation Numerical (int) m Elevation above sea level 16 hor_shade Numerical (int) % Portion of sky that is blocked by terrain 17 Wp Numerical (int) Wp Estimated peak wattage of the PV installation 18 kWh_a Numerical (int) kWh/year Modeled mean yearly energy output in kilowatt-hours (10-year mean, 2014–2023) 19 Wh_a_Wp Numerical (int) Wh/year/Wp Yearly energy output in watt-hours per estimated peak wattage PV electricity modeling The CSV file contains the results of PV electricity modeling based on the vector polygons for a 20-year period (2004–2023). Each row represents a PV cluster and can be joined to the polygons via the fid column. The hourly model outputs are summarized into four statistical metrics for each month–hour combination: mean, median, 25th percentile, and 75th percentile (see table for further column details). The hours 1–2 and 21–24 are not included, as PV electricity production does not occur during these times throughout the year. The Timezone is UTC+1. Modeling period: 20 year (2004 - 2023) Format: Comma separated values (CSV) DimensionsStatistical values: 4 (mean, q25, median, q75)Months: 12Hours: 18 (3:00 to 20:00)Total number of columns: 864 Column description # Column Name Data Type Unit Description 1 fid Numerical (int) - Unique identifier for PV panel cluster (matches the gpkg) 2, 6, 10 ... 862 mean_m[01-12]_h[03-20] Numerical (int) W mean electrical output of the given PV panel cluster at a specific hour of the day and in a specific month 3, 7, 11 ... 863 q25_m[01-12]_h[03-20] Numerical (int) W 25th percentile of electrical output of the given PV panel cluster at a specific hour of the day and in a specific month 4, 8, 12 ... 864 median_m[01-12]_h[03-20] Numerical (int) W median electrical output of the given PV panel cluster at a specific hour of the day and in a specific month 5, 9, 13 ... 865 q75_m[01-12]_h[03-20] Numerical (int) W 75th percentile of electrical output of the given PV panel cluster at a specific hour of the day and in a specific month Data sources Automatic segmentation of PV panels based on orthophotos:Bundesamt für Eich- und Vermessungswesen. (2025a). Serie Digitales Orthophoto Farbe und Infrarot (DOP RGBI) Stichtag 15.04.2025. https://data.bev.gv.at/geonetwork/srv/metadata/a9df237a-ed22-46fc-bab5-a25312ce0917 Orientation classification and horizon shading with digital surface model:Bundesamt für Eich- und Vermessungswesen. (2024). Serie ALS DSM Höhenraster 1m Stichtag 15.09.2024. https://doi.org/10.48677/3B27C53B-C176-4D05-BDE5-ABE7FA3DFC0E Biodiversity rank on the basis of INVEKOS:Agrarmarkt Austria (AMA). (2015). INVEKOS Agricultural plots. https://inspire.lfrz.gv.at/009501/ds/inspire_schlaege_2015_polygon.gpkg.zip Classification of rooftp PV with OSM buildings:OpenStreetMap contributors. (2025, August). Geofabrik OSM Data Extract for Austria (buildings). https://download.geofabrik.de/ Climate Data for Modeling:Lehner, F., Maier, P., Klisho, T., & Formayer, H. (2025, February). FORSITE-Clim Europe: European-wide climate indicators for historical periods and climate projections at high resolution. https://doi.org/10.5281/ZENODO.10623853 Acknowledgement The computational results presented in this master thesis have been achieved using the Austrian Scientific Computing (ASC) under the project 72536: Photovoltaics, Humans and the Biosphere: A transdisciplinary approach fostering Alpine resilience.