About this measurement
- Method: Sun tracking with mobile app
- Tracker: Jon Coxon
- Location: Lowestoft - England (52.49º, 1.75º)
- Date and Timezone: 2017, Aug 28, GMT+1
Raw data and spreadsheet
Comparison with predictive models
Azimuth data seems to have an error margin too high to be taken seriously. One example is the azimuth of 216º at 12:54. We know that time to be very close to solar noon at that location (as predicted by both models). The sun is known to be due south (azimuth = 180º) at that time. An error as big as 36º means that dioptra is clueless for Azimuth.
Elevation data shows a very close fit to the globe model. The elevation at sun peak and the time of sun set seems to be accurately predicted by the blue curve. This weakens the flat-earther argument that refraction makes the sun set on a flat earth. Even if we assume that could be the case, it’s hard to swallow that refraction would just happen to make the sun elevation vary with time EXACTLY as predicted by the globe model.
Since the height of the sun is not known precisely on the flat model, we must also try to vary it to try to fit the data (the elevation prediction above is for height=4828Km). Let’s try a different value of 3500Km.
The shape of the red curve doesn’t change. It only comes down a bit (which makes sense) and the data fit gets worse. The best fit actually seems to be around H=4828Km. There’s no way to put the flat sun in a height that will make the prediction fit that data.
The world is looking a little rounder. But this was only the first experiment. Science is made with repeatable, measurable, observable and testable evidence from independent sources. We need more.