I suspect that "geological record" quoted is only one source - the graph in the original link is many sources combined.

Also, note the scales are totally different, and that the original link at least states sigma (uncertainty.)