[laserbeam]

I don't think 0 makes sense here. Nor does "50% lower in preindustrial times" makes sense either. The correlation between temperature and co2 exists, but the absolute values involved are not intuitive.

"The co2 level is N standard deviations out of the ordinary compared to preindustrial times" feels more useful here, but reading that requires basic statistics (which general audiences probably lacks).

[timsco]

While I see what you're saying, the issue is that when you eyeball the graph, you think, "wow, it's 4x what it used to be".

[laserbeam]

Here's the thing... It might actually be 4x standard deviations higher, if not more. "Almost 2x absolute value" conveys less information here as the absolute minimum co2 value in the entire known history if the earth is non zero. And it is non zero. If I recall correctly it's closer to 80% of the value of recent preindustrial values (one should check this).

Would be cool to see that value on the graph, also a median/average/max preindustrial value.

[ad404b8a372f2b9]

It does not convey less information, it conveys accurate information. At the scale of the x-axis, the viewer can see that it was always at 280 so no information is lost. No information is added by breaking the scale, if you want to convey standard deviations, add a secondary axis.