UPDATE:

I am now providing two different data sets: one that is calculated witha 25 degree minimum user terminal angle and one with a 35 degree minimum user terminal angle. The more recent SpaceX filings show that they are capable of using the 25 degree minimum but that they eventually want to increase that. The previous datasets were calculated with a 35 degree minimum as that was what I had found on file at the time. With the lower minimum angle and about 100 more satellites at operational altitudes the 25 degree dataset now shows bands with continuous coverage.

Furthermore, I plan to calculate what I call the satellite multiplicity, the number of satellites overhead at any given time on average, since the number of overlapping satellites is relevant to some calculations for routing and bandwith.

New Features: You can now double-click a cell to toggle showing more granular data. Cells with 100% uptime now display a gold star. Faster parsing, reduced battery drain, hopefully.

This renders the percentage of the day that a portion of the earth is covered by a Starlink satellite. (Red is no coverage, blue is all day coverage).

Click on a cell to see how many minutes in a day a cell is covered by at least one satellite. Note that currently for performance reasons the cells are much larger than the resolution the data was simulated at. As such the error within a cell can be quite large, especially near the poles where the coverage drops off rapidly.

In my opinion anything less than 1440 minutes (aka continuous coverage) means that SpaceX should not provide service in that area.

For more info about how it's calculated: https://github.com/sebsebmc/starlink-coverage

Changelog:

Star image: Yakiv Gluck / CC BY-SA (https://creativecommons.org/licenses/by-sa/3.0)