4 ms·
In the first article, fitting that data with a parabola+trigonometric is a very big red flag. The error interval for the quadratic coefficient must be huge, but
by gus_massa 7y ago
In the first article, fitting that data with a parabola+trigonometric is a very big red flag. The error interval for the quadratic coefficient must be huge, but I can't find it.
It may be fit with a line+trigonometric to get some approximation for the next few years (10? 100?). With a line you get 50000 years to the 0 oxygen level, that is less alarmist than 3600 years.
Also, without a good model it's a bad idea to extend these projections too much. There are a lot of natural process that depend on the concentration of oxygen and will change. For example, under 15%-10% concentration of oxygen you can't burn wood so you solve the problem of forest fires https://en.wikipedia.org/wiki/Limiting_oxygen_concentration https://en.wikipedia.org/wiki/Limiting_oxygen_concentration