In 2011, Andy Revkin wrote an article (archive) entitled “Straight Talk on Rising Seas in a Warming World” (among other articles on the topic), in which he optimistically sought guidance on the topic from a then recent study of U.S. East Coast sea level coauthored by Mann (Kemp et al, 2011). Joshua Willis told Revkin “that, using patterns in layered salt marsh sediment, [they] found a sharp recent uptick in the rate of sea-level rise after 2,000 years of fairly stable conditions — a pattern Willis refers to as a “sea-level hockey stick” — an allusion to the suite of studies finding a similar pattern for global surface temperatures (albeit a hockey stick with a warped shaft)”.
However, as so often, the supposed “hockey stick” appeared only after the data had been severely adjusted. The difference is shown at the figure at right. Unadjusted (raw) relative sea level (i.e. how sea level appears locally – the concern of state planners and policy-makers) in North Carolina increased steadily through the last two millennia, with somewhat of an upward inflection in the 19th century; it is only after heavy adjustment that a HS shape appears.
In this case, the relevant data for local and regional planners is the data prior to adjustment by climate warriors, as I’ll discuss below: this is not a hockey stick but an ongoing increase through the Holocene.
Relative Sea Level and Glacio-Isostatic Adjustments
Kemp et al’s adjustment subtracted estimated glacio-isostatic adjustment (GIA) to arrive at “sea-level change after correction for glacial isostatic adjustment“. This is not a figure that is of any real relevance to policy and unfortunately serves only to mislead earnest observers (like Revkin) seeking guidance from a distressingly ideological community.
Glacio-isostatic adjustment is a very important topic in Antarctic studies, since (as discussed at CA here), the estimate of Antarctic ice mass loss is approximately equal to the (poorly-constrained) Antarctic glacio-isostatic adjustment. IPCC AR5 promoted very extreme Antarctic ice mass loss estimates through their adoption of GIA estimates that were already discredited in the specialist community (see CA here).
Isostatic adjustment is a lingering effect of disappearance of continental ice sheets more than 10000 years ago. The direction of the effect is opposite in Canada versus the US. In Canada, land is rising relative to the ocean as the land rebounds upward due to the removal of the massive LGM ice sheet. In the U.S., most land, including all of the Eastern Coast except for Florida, is moving downwards relative to the ocean, to compensate isostatically for upward movement of land in Canada. The contour line showing the break between the two zones follows the Canada-US border to an astonishing approximation. Around Hudson Bay, uplift is approximately 10 mm/year, while relative subsidence on the mid-Atlantic coast is estimated at about 1.1 mm/year.
Paleoclimate Evidence of Relative Sea Level Rise on US East Coast
There is overwhelming paleoclimate evidence of relative sea level rise through the Holocene along the US East Coast, reported in numerous specialist articles, a few of which will be discussed below. I haven’t attempted to survey literature on the topic; I presume that there are many other references. However, the story is consistent in the references below.
Long Island, New York
In 1975, Sanders and Kumar (pdf) identified analogues to present-day barrier islands in shallow cores offshore Long Island. ‘They (plausibly) interpreted these paleo-barrier islands as evidence of former sea levels. They estimated relative sea level in the early Holocene (~9000 BP) to be ~24 meters lower. In addition to sea level rise, higher sea levels through the Holocene caused encroachment of the Atlantic Ocean onto the continental shelf: they estimated that the ocean had advanced by ~7 km (nearly 100 m/century) during the sea level rise:
when the sea level stood at -24 m [9000 BP], a chain of barrier islands, very similar in height and size to the modern barriers south of Long Island, existed approximately 7 km offshore. These barriers were approximately 8 to 10 m high.
Miller et al 2009 (New Jersey)
In 2009, Miller et al (2009) estimated relative sea level rise since 10000BP as ~18 m, as shown in the figure at right. In the last half of the Holocene (since 5000 BP), they estimated relative sea level rise of 9 meters (~1.8 mm/yr). They said that this was a substantial increase from then current New Jersey estimates of only 0.5 mm/yr rise since 2000 yrBP. Their estimate was obtained from offshore drowned salt marshes.
Miller et al (2009) allocated over 60% of the observed relative sea level rise (1.1 mm/year) since ~5000 BP to geoidal subsidence, with the balance (~0.7 mm/year) attributed to eustatic sea level rise. Citing Lambeck 2002, they speculatively attributed late Holocene sea level rise to ongoing erosion of the West Antarctic Ice Sheet (a topic of very great interest separately):
Lambeck (2002) suggested that west Antarctic ice sheet melting was the major contributor to mid–late Holocene rise.
Miller et al reported that their results were indistinguishable from other recent estimates from contiguous regions (Delaware, southern New England):
Ramsey and Baxter (1996) evaluated dates from nearby Delaware sites and provided a preferred relative sea-level curve that is indistinguishable from ours, with a mean rate of 1.7 mm/yr. A compilation of data from Southern New England (Donnelly et al., 2005) is also virtually indistinguishable from the New Jersey and Delaware records with a mean rate of 1.9 mm/yr. We conclude that from 5000 to ~500 yrBP relative sea level rose ~1.7-1.9 mm/yr for the region from Delaware to southern New England. This is about 1 mm/yr slower than regional rates of rise since 1900 AD (Fig. 1). Rates in Delaware and Southern New England were faster from 8000-5000 yrBP, though the rates in New Jersey are not well constrained for this interval.
Kemp et al 2011
Kemp et al 2011 thus arrived in a milieu where long-term relative sea level rise along the US East Coast was well known. They reported on two rather short North Carolina cores (Tump Point and Sand Point), which were dated to the past two millennia, and did not discuss the longer Holocene context. They observed that both cores showed a consistent increase in relative sea level of about 1.2 mm/year over the past two millennia, somewhat less than than the corresponding rate in New Jersey (which has higher GIA).
From these amounts, Kemp et al deducted 0.9-1.0 mm/yr for subsidence to obtain the “climate-related rate of sea level rise”:
To extract climate-related rates of sea-level rise (Fig. 2C), we applied corrections for crustal movements associated with spatially variable and ongoing glacial isostatic adjustment (GIA). A constant rate of subsidence (with no error) was subtracted from the Sand Point (1.0 mm/y) and Tump Point (0.9 mm/y) records.
This resulted in the “sea level hockey stick” (SKS version shown at right), which was characterized by Revkin as follows:
“Among other things, he [Josh Willis] speaks about the recent study [Kemp et al. 2011] that, using patterns in layered salt marsh sediment, found a sharp recent uptick in the rate of sea-level rise after 2,000 years of fairly stable conditions — a pattern Willis refers to as a “sea-level hockey stick” — an allusion to the suite of studies finding a similar pattern for global surface temperatures (albeit a hockey stick with a warped shaft).”
But, for an observer in North Carolina concerned about ocean encroachment, the relevant figure is the history of relative sea level, which has been consistently increasing over the entire Holocene, not the Mannian hockey stick.
North Carolina relative sea level rise previously discussed at CA here.
Kemp et al 2013 (New Jersey)
Two years later, Kemp et al 2013 reported on two cores from New Jersey, which, collectively (see their Figure 7b below) showed relative sea level increase of ~1.6 mm/yr (4 meters over 2500 years), slightly less than Miller et al’s average Holocene RSL increase of 1.8 mm/yr, with, once again, an inflection upwards starting in the 19th century.
This time, they allocated all of the relative sea level rise in New Jersey since 4000BP to glacio-isostatic adjustment (GIA) and other land changes (compaction), i.e.that there had been no eustatic sea level rise since 4000 BP
Eustatic input ceases at 4000 yrs BP in this model, since then all RSL changes (1.1 mm/yr) are attributed to GIA and associated processes such as redistribution of water in response to changes in the geoid.
Projections in a Holocene Context
To place recent projections of relative sea level rise e.g. Rutgers projections of ~42.5 cm (1.4 ft) by 2050 and 76-103 cm (2.3-3.4 ft) by 2100 (both from 2000 levels) in New Jersey (blue + signs), I’ve superimposed Miller’s estimated 1.8 mm/year Holocene increase (dashed line) to 2100AD on Kemp et al. The increase measured by tide gauges in the 20th century (the short red squiggle) is merely an episode in a a continuing sea level increase over thousands of years. While the increase in the 20th century and projected increase in the 21st century are more rapid than the long-term average, even average late Holocene encroachment of the ocean onto the land along the mid-Atlantic coast would be relevant to development along the coast. As an editorial comment, it seems plausible to me that the Little Ice Age in the second half of the 2nd millennium could easily have contributed to a relative slowdown in eustatic sea level rise.
Discussion
Although Revkin was clearly hoping for and expecting “Straight Talk” about sea level rise from Joshua Willis, he got anything but. Willis’ message was about as counterproductive to the science requirements of an earnest planner as one could imagine. The base case for a planner is that relative sea level rise along the US East Coast has been going on for thousands of years. Rather than East Coast relative sea level rise being represented by a “sea level hockey stick”, it has been an ongoing increase for thousands of years. Yes, there has been a recent inflection in relative sea level rise, but the planners have to deal with relative sea level regardless of the inflection. The ocean has been encroaching on the land in the U.S. mid-Atlantic for the past 10,000 years and will continue to do so for the next 5,000 years.
I noted above that Lambeck (2002) had suggested that West Antarctic ice sheet melting was the major contributor to mid–late Holocene eustatic sea level rise. The future of the West Antarctic ice sheet has been the subject of many recent alarmist screeds, none of which have fully discussed and analysed its past. In a future post, I’ll discuss West Antarctic Ice Sheet retreat in more detail. As a lead, it is my understanding that the West Antarctic Ice Sheet has been retreating for the past 20,000 years and will continue to retreat for the next 5000 years. Sea level rise seems to me to be intrinsically connected to being in the Holocene interglacial and the only thing that seems likely to stop continuing sea level rise is a reversion to Ice Age conditions – a cure obviously worse than the disease. As a matter of historical contingency, our civilization developed settlements along the ocean at a time when the West Antarctic Ice Sheet was only part way through its retreat and we might as well recognize that.
On balance, our civilization effortlessly accommodated sea level rise in the 19th and 20th centuries. However, barrier islands off the US East Coast have been gradually submerged. Present-day barrier islands will also be gradually submerged in their turn, regardless of the Paris Treaty. Attempts by climate scientists to portray this as a hockey stick issue with an easy control knob are false science.