A tide gauge doesn’t measure the height of the sea surface; rather it measures the difference between the height of the sea and the that of the land. Thus is it sensitive to changes in sea surface height (SSH), and to vertical land movement (VLM) as well.
In the previous post I took sea level data from tide gauges on the east coast of the U.S., and computed a composite sea level record for each of four regions: New England, the northern mid-Atlantic, the southern mid-Atlantic, and Florida. Now that PSMSL (the Permanent Service for Mean Sea Level) has updated their database to include the latest complete year I recomputed my composites, using the data as is (not removing those periodic fluctuations I discussed previously, but yes removing the yearly cycle of the seasons by using anomaly values). And here are the regional time series, offset to make them easier to see:
It’s easy to see that sea level rise if faster along both mid-Atlantic regions than it is in New England and Florida. This is further suggested by directly comparing the smoothed series, which gave me this (not offsetting them, but letting them plot on top of each other):
Clearly sea level rise is faster along the mid-Atlantic, north and south both, and slower for both New England and Florida.
Now the twist: for each series I estimated a “reference rate” of sea level rise using the time span 1950-1990 (January 1950 through December 1989 actually). Then I subtracted that trend from the data series and looked at what was left over — which I decided to call by the imaginative name “excess” (above the trend line 1950-1990 extended future and past). The reference trends (1950-1990 values), in order of decreasing rate, were southern mid-Atlantic at 2.6 ± 1 mm/yr, northern mid-Atlantic at 2.5 ± 1 mm/yr, Florida at 2.0 ± 0.8 mm/yr, and New England at 1.4 ± 0.6 mm/yr.
What makes the whole exercise interesting is how it affects comparison of different regions when you de-trend them by different amounts (amounts suited to each region). It brings the two southern regions (southern mid-Atlantic and Florida) into surprisingly good agreement, and does so for the two northern regions (northern mid-Atlantic and New England) as well:
The difference of the trends requiring removal is most likely due to different rates of land uplift/subsidence when averaged over each region. Sea level rise at Florida (New England) is indeed slower than along the southern (northern) mid-Atlantic coast, but part of the reason is the different average rates of land uplift/subsidence in the two regions.
The encouraging aspect is that much of the regional-average difference is because of glacial isostatic adjustment (GIA), which is indeed very slow and believed to be genuinely linear, at least as closely as we can tell on time scales we’re working with (a few centuries at most). When we subtract a linear approximation of GIA, we get all of it because GIA really is linear.
Of course we also subtract a linear approximation of other stuff too. Still, the result supports the notion that along the southern mid-Atlantic and Florida coasts SSH is mainly moving in lock-step — almost all of the difference between these two regional averages is VLM. But the northern vs southern comparison reveals significant differences in how SSH, the actual height of the sea, has changed over the last century+ due to ocean dynamics.
Probably the most interesting results are the recent upticks for all four regions, showing the acceleration starting about 1990 in the two more northerly regions but more recently (and more dramatically) in the more southerly regions. My intuition says that the high speed of southern sea level won’t be maintained for very much longer, rather it will return to rates comparable to the northern regions. But they too are rising faster than before. My intuition also says that the Gulf Stream System slowdown is somehow involved in the recent southern uptick. Alas, I lack the knowledge base to evaluate that idea more rigorously.
This blog is made possible by readers like you; join others by donating at My Wee Dragon.