Rooftop solar data for 673 North American cities

Every figure on this site comes from the same measured dataset: Google Solar API rooftop samples across 673 cities in the US and Canada. This page publishes the whole thing, including the CSV. Free to use with attribution.

Download the full dataset (CSV) → 673 rows, 17 columns. Updated whenever the site rebuilds.

What this measures, and what it doesn't

Each city's figure is the median of roughly 16 individual rooftops sampled in a ring across the city's suburban area — not one reading from the city centre. That distinction is the whole reason this dataset exists: a single downtown sample produced results so distorted that Charlotte appeared to get less sun than Vancouver. Sampling a spread of real suburban roofs and taking the median fixed it.

The sunlight figure is usable sunlight. It already has shading from trees, chimneys and neighbouring buildings subtracted, which is why these numbers are lower than the "hours of daylight" figures published elsewhere. The unshaded column is what an unobstructed roof in the same city reaches, and the gap between the two is the shading loss.

What it is not: it is not a measurement of your roof, and it is not a forecast. Payback figures are modelled from the measured sunlight plus regional average electricity rates, using 400W panels at $2.85/W. Local incentives, your actual usage and your utility's export rules will all move the answer.

The headline finding: sunlight is not what decides this

Palmdale tops the sunlight table at 2,115 hours and Fairbanks sits last at 954 — a 2.2× spread. But the fastest payback in the dataset belongs to Honolulu, HI at 3.1 years, on 1,787 sun hours — well short of the sunniest city.

Electricity price moves payback far more than sunlight does. A dim city with expensive power beats a sunny city with cheap power, consistently, across the whole table. That is the single most useful thing in this dataset and it is the opposite of how solar is usually marketed.

Fastest payback

Most sunlight

Least sunlight

Shading

State College, PA loses the most sunlight to obstruction of any city measured — 20.5% below what an unshaded roof there would receive. Shading loss is the column most likely to differ from whatever a national calculator assumed about your roof.

Using this data

The CSV is free to use, including commercially. Attribution to solarpanelfit.com is appreciated and is all that's asked. If you cite it:

PanelFit, Rooftop solar data for 673 North American cities, derived from the Google Solar API. https://solarpanelfit.com/solar-data

The underlying rooftop measurements come from Google's Solar API and remain subject to Google's terms. What's published here is our sampling, aggregation and modelling of it.

Generated 2026-08-22. Estimates only — not a quote and not financial advice.