The Sea Also Rises RTFR Version

There has been much comment on Hay, Morrow, Kopp and Mitrovica's "Probabilistic reanalysis of twentieth century sea level rise", much of it centered about the finding that sea level rise was slower than thought in the first nine decades of the twentieth century and has really take off since.

Now some, not Eli to be sure, have been mislead about this, and Eli for one would place a reasonable amount of blame for some of this on the authors, but, of course, others have pitched right in.  For starters the abstract and the paper push the conclusion

Here we revisit estimates of twentieth-century GMSL rise using probabilistic techniques^{9, 10} and find a rate of GMSL rise from 1901 to 1990 of 1.2 ± 0.2 millimetres per year (90% confidence interval). Based on individual contributions tabulated in the Fifth Assessment Report⁷ of the Intergovernmental Panel on Climate Change, this estimate closes the twentieth-century sea-level budget. Our analysis, which combines tide gauge records with physics-based and model-derived geometries of the various contributing signals, also indicates that GMSL rose at a rate of 3.0 ± 0.7 millimetres per year between 1993 and 2010, consistent with prior estimates from tide gauge records⁴. The increase in rate relative to the 1901–90 trend is accordingly larger than previously thought; this revision may affect some projections¹¹ of future sea-level rise.

 Of course, any such calculation depends on a better reading of the intermittant, spotty and just plain strange tidal guage record and other some such before 1990 when satellite measurements became available, both for themselves and for calibration of sea level gauges.

Eli has some experience in this sort of stuff having suffered through the late John Daly's Isle of the Dead tidal sage whose echoes can be found even today.  Daly and his fellow rejectionists succeeded in stirring up enough of a furor that it provoked an official investigation, which, of course, found that Daly was, shall Eli say, indulging in ahistorical fiction.

Make no mistake about it, Hay, et al. have made a major advance in methods

In this Letter, we revisit the analysis of GMSL since the start of the twentieth century using Kalman smoothing. This statistical technique naturally accommodates spatially sparse and temporally incomplete sampling of a global sea-level field, provides a rigorous, probabilistic framework for uncertainty propagation, and can correct for a distribution of GIA and ocean models.We applied the approach to analyse annual records from 622 tide gauges included in the Permanent Service for Mean Sea Level (PSMSL) Revised Local Reference database and reconstruct the global field of sea-level change for each year from 1900 to 2010.

and, of course, they can check their tidal gauge record against  satellite measurement in the most recent parts of the period of study as well as making many consistency checks.  Finally their results appear to close the balance

This estimate closes the sea-level budget for 1901–90 estimated in AR5 (ref. 7) without appealing to an underestimation of individual contributions from ocean thermal expansion, glacier melting, or ice sheet mass balance. Moreover, it may contribute to the ultimate resolution of Munk’s sea-level enigma (defined by the argument that Earth rotation measurements and bounds on ocean warming are inconsistentwith a rate of sea-level rise beginning in the late nineteenth century of 1.5–2.0 mm/yr), since it may lower the signal of twentieth century ice melting in Earth rotation measurements.

But, knew there was going to be a but, dinna you?  The take home for Eli which he has not seen much commented on is Figure 4.

 It is not that sea level rise from tidal gauges was a constant ~1 mm/yr or that the rate suddenly accelerated in the 1990s, but it is more complicated than that which is Eli's point.  The impression left by much of the to and fro has been that there was a constant amount of sea level rise until 1990 when it jumped.

Not the case.  The rate varied with a period of ~ 40 years or so (dangerous because the smoothing is over a 15 year period, before starting to grow systematically in the late 1960s (remember the 15 year average).  The insert shows that the most recent times have been the periods of largest sea level rise.  The satellite era tidal gauge record from Hay et al, matches well on the satellite measurements, while the tidal gauge measurements before 1990 are lower than previously thought but show the same sort of variation. (from SKS)

Added on afterthought: The rise per year is a rate, the slope of the line is the acceleration.  From this plot bunnies can say that the acceleration has been monotonically postive since ~1960 and not far off from lines, implying polynomially growth in sea level.