That statista pages cites this paper in 'nature sustainability': https://www.nature.com/articles/s41893-021-00842-z

Which does not seem credible, it has way too many guesstimates.