This is a reblog from Eli Rabett, one of the post “On Records”, with additional comments and material from the author-moderator of this blog, 667-per-cm.net:
A distinguishing mark of a new record in a time series is that it exceeds all previous values another is that the first value in a time series is always a record.
Given a stationary situation with nothing except chance, aka natural variation, the number of new records should decline to zero, or pretty close, as the series extends in time.
.
.
.
(Click on figure to get a larger image, and use browser Back Button to return to blog.)
The above is from:
S. I. Seneviratne, R. Wartenburger, B. P. Guillod, A. L. Hirsch, M. M. Vogel, V. Brovkin, D. P. van Vuuren, N. Schaller, L. Boysen, K. V.
Calvin, J.n Doelman, P. Greve, P. Havlik, F. Humpenöder, T. Krisztin, D. Mitchell, A.r Popp, K. Riahi, J. Rogelj, C.-F.h Schleussner, J. Sillmann, E. Stehfest, ``Climate extremes, land–climate feedbacks and land-use forcing at 1.5°C'', Philosophical Transactions of the Royal Society A, Mathematical, Physical and Engineering Sciences, 2nd April 2018, 
While Seneviratne, et al approach the estimates with a deeper and far more nuanced analysis, it’s been known that as the globe warms, land will warm faster:
And the travesty and tragedy are we’ve known about this a damn long time and have done nothing:
That’s from 1958.
See also:
H. D. Matthews, K. Zickfeld, R. Knutti, M. R. Allen, ``Focus on cumulative emissions, global carbon budgets and the implications for climate mitigation targets'', Environmental Research Letters, January 2018, 13(1).
ecoquant
