programs for calculating and displaying curvature of the earth
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.

Calculations of earth curvature

Although I'm firmly in the "spherical earth" camp, I have friends who range from "curious about" to "convinced by" flat or hollow earth (on which we live on the inside) hypotheses. This repository is my attempt to allow exploration of these viewpoints so that the inquistive mind can test which one holds up best to objective reality.

My eventual goal is to develop a site similar to Ulrich Deuschle's panorama generator but having choice of output to match more than one of these 3 "theories". That should give the most convincing proof of one "viewpoint" over the other, since what one sees from a mountaintop should be distinctly different between a "flat", "round", or "hollow" earth.

The Python script, based on the formulas shown at and (although its output differs from those sites by a small amount, probably due to my choice of the value of R [earth radius]), should help lay one claim to rest: that of "If there were 266 feet of curvature in 20 miles, I'd be able to see it". The script output shows that in 1/10 of a mile -- 528 feet -- there is only about 2mm of "drop" due to curvature. If you've ever shot at targets at 100 yards using an iron-sighted rifle, you'll know how small something looks at that distance, and 528 feet is closer to 2 football fields' distance. So if you can't see 2mm in 528 feet, you can't arguably see any curvature over 20 miles either. What you should be able to see, on a curved earth, is that things at a distance are most often (except during anomalies of atmospheric refraction) partially or fully hidden from view by that curve.

A problem with the dizzib and earthcurvature sites is that they don't allow specification of a refractive index, which adjusts the "drop" by accounting for the bending of light rays in the denser air typically found at earth's surface. Typical values are .25 or .125. You can pydoc earthcurvature to see how to use this in your calculations.

Flat-earth proponents will likely balk at the source of the SRTM3 data, their arch-enemy NASA. However, anyone is free to create their own SRTM3 files with elevation measurements more to their liking.

As of 2019-01-06, the panoramas generated by are flat-earth views, not the "dish" view centered on the North Pole typical of classical FE belief, but an equirectangular, which should show acceptable panoramas close to the equator, but will be increasingly distorted the farther away you go. Another caveat is that they are projected as if the viewer has a narrow vertical slit with which to see, and so he is turning in place to see the panorama. And one more severe limitation of the current state of the software is that it assumes a view over the ocean, and will not show anything below "eye height".

But despite all those shortcomings, it can be useful in comparison with Ulrich's panoramas at the link mentioned above, and seeing which more accurately reflects the skyline you see at your current location.

Getting Started

This assumes a typical Debian developer's system. Your mileage may vary.

sudo mkdir -p /usr/local/share/gis/hgt
sudo chown $USER.$USER /usr/local/share/gis/hgt /usr/src
mkdir -p /usr/src/jcomeauictx
cd /usr/src/jcomeauictx
git clone
cd curvature/
sudo apt-get install zip imagemagick
make N37W120.fetch
make showfile  # shows part of Northern California west of Millerton Lake
make N24W111.fetch N24W112.fetch N25W111.fetch N25W112.fetch N26W111.fetch N26W112.fetch
make panorama  # shows flat-earth horizon in La Paz, MX looking at Isla San Jose
make EARTH_RADIUS_MILES=GLOBE panorama  # same view but adjusted for curvature