Germany's PV Atlas

Best estimate: 86,010 MWp, with a 90% range of 76,182–97,943 MWp. The range covers two measured sources of uncertainty, not everything that could move this number. Below: what is inside the range, what is not, and how it compares with other sources.

Inside the range:

• The two area-to-power constants. Panel area on a roof converts at about 0.18 kWp per m², the rating of a typical module; the outline of a ground-mounted park converts at about 0.05 kWp per m², measured on parks whose capacity is known. Each is varied over its own range (90%: 0.15–0.21 and 0.035–0.075 kWp per m²).
• How reliable the segmentation model's detections are. How often its detections turn out to be real, and how many mapped arrays it misses, were measured size class by size class against mapped installations. Those measurements have their own sampling error, which is drawn into the range.

Outside the range, and not something more arithmetic can add back:

• Getting the national total right is not evidence the geography is right. Cross-validated, this method's German national figure lands within a few percent of the register while the median municipality is out by roughly half, because over-prediction of the densest tenth of municipalities cancels under-prediction of the other ninety percent. Three different models whose municipal errors spanned 38–64% all produced nearly the same national number. What a grid model is consumed for is the geography, not the total.
• In Germany, multiplying roof area by a constant does as well as the classifier. Against a per-municipality roof-area baseline, the classifier's median error (48.4%) is worse than the baseline's (37.8%). Where rooftop PV is close to ubiquitous, capacity is nearly proportional to roof area by construction. The classifier earns its keep where PV is rare -- in Pakistan it cuts the same error from 88.9% to 26.5% -- which is the regime it was built for and Germany is not.

Treat this as an early estimate from an active research pipeline, not a census. What it offers is a reproducible way to estimate distributed solar from free satellite imagery and open-source models where official statistics are thin or missing. Expect the figures to move as the evidence behind them grows.

There are no hand-mapped ground-truth areas behind this page, and for the small-panel half there do not need to be. Registration in Germany's Marktstammdatenregister is legally mandatory rather than voluntary, and it publishes totals per municipality -- which is exactly the quantity the small-panel estimator produces. The conversion from flagged roof area to kilowatts was therefore fitted against the register across 9,674 fully-covered municipalities, at 0.2793 kWp per m² of credited roof area, and applied countrywide as a genuine out-of-sample test. What a register cannot substitute for is a precision check on small installations: below 30 kWp it publishes no coordinates at all, so there is no way to ask whether a particular flagged roof really carries panels.

A complete national register puts the Best estimate above too high. The Marktstammdatenregister records roughly 74.8 GWp of PV in Germany, below the figure this page reports, so read these tiers as geography -- where capacity sits -- rather than as a national total. The hand-mapped OpenStreetMap component was reconciled against the register in September 2026 and both placements now sit within about 10% of it; what remains unreconciled is the model's own detections and the small-panel half. Every other country in this project would need a register of this kind to make the same statement, and does not have one.

Layers. "Show" picks the quantity the cells are coloured by: the Best estimate, the OSM-mapped capacity alone, the satellite detections above 400 m², rooftop PV below it, farm ground PV below it (arrays beside fields, where the farm detector has run), or the gap between the Best estimate and what is already mapped, which is where mapping effort pays off most. Each cell shows its total in MWp. The colour scale is logarithmic and is refitted to whichever quantity is shown; the legend always states its range.

Your own data. "Layers" → "Add a layer" overlays a GeoJSON file, a CSV with latitude/longitude columns, a GeoJSON URL, an XYZ raster tile URL or a WMS (use {bbox-epsg-3857} in its GetMap URL). Files are read in your browser and are not uploaded anywhere. The live OpenStreetMap overlay queries the public Overpass API for the area in view when you zoom in, so it shows mapping done after this page was built.

Exports. Every figure downloads as PNG (for slides), SVG (vector, for papers) and CSV (the numbers behind it); the map as PNG with its legend and attribution, and the grid as GeoJSON or CSV. Basemap imagery and tiles keep their own providers' terms: OpenStreetMap data is © OpenStreetMap contributors (ODbL), the grey canvas and the satellite imagery © Esri and its partners.