To the Editors:

Professor Freeman Dyson’s review [NYR, May 15] of The Earth’s Biosphere: Evolution, Dynamics, and Change by Vaclav Smil discusses the acceleration of sea-level rise. He says that “a slow rise of sea level… could become catastrophic if it continues to accelerate” and “we observe a steady rise from 1800 to the present, with an acceleration during the last fifty years.”

Acceleration means that the rate of change is itself changing. An acceleration in sea-level rise is unsupported by the majority of data. Sea-level data come from tide gauges, few of whose records predate the twentieth century. (See the Web at www .html for an authoritative international compilation.) Of US tide gauges, there are fourteen that are located at or near the ocean coast and that have relatively complete records with annual sea levels at quarter points (1925, 1950, 1975, 2000) in the twentieth century. These fourteen gauges are, clockwise around the US portion of North America, Portland, Boston, New York, Atlantic City, Charleston, Key West, Pensacola, Galveston, San Diego, La Jolla, Los Angeles, San Francisco, Ketchikan, and Honolulu. If acceleration occurred “during the last fifty years” as claimed by Professor Dyson, we should expect sea level to have risen significantly more in the fourth quarter (1975–2000) than in the second quarter (1925–1950). All fourteen of these gauges should show that, but they do not.

The fourteen tide gauge records show clear distinctions between the eight gauges on the Atlantic and Gulf coasts and the six gauges in the Pacific. For the seventy-five years 1925–2000, the Atlantic and Gulf tide gauges show a cumulative rise of 2,053 mm (257 mm per gauge), but only 10.7 percent of that rise came in the last twenty-five years (1975–2000). Sea level at each of the eight Atlantic and Gulf gauges rose less during 1975–2000 than during 1925–1950. These data suggest a strong deceleration, rather than any acceleration in sea-level rise.

The changes at the six Pacific gauges differ from the Atlantic and Gulf gauges both in magnitude and in timing. The cumulative rise at the six Pacific gauges was only 370 mm (62 mm per gauge). For the seventy-five-year interval, this 62 mm per gauge on the Pacific is less than a quarter of the 257 mm per gauge shown by the eight Atlantic and Gulf coast gauges.

But the six Pacific gauges support Professor Dyson’s claim for an acceleration, because 70 percent of this smaller change is in the last quarter, 1975–2000. In 1975–2000, the eight Atlantic and Gulf gauges rose an average 27.5 mm/gauge and the six Pacific gauges rose an average 43.3 mm/gauge (about one inch vs. an inch and three quarters in that quarter-century). The differences between gauges on the east and west coasts cannot be due to real sea level actually rising at different rates on either side of the continent. Water will not run uphill. The differences reflect compaction of sediment under the gauges, geological uplift, El Niño, and other natural factors that must be separated out. The data, taken as a whole, do not support a recent acceleration in sea-level rise.

Cyril Galvin
Coastal Engineer
Springfield, Virginia

Freeman Dyson replies:

I am grateful to Cyril Galvin for correcting my mistake. My mistake was to write, “We have accurate measurements of sea level going back two hundred years.” In fact we have measurements, but their accuracy is questionable, for the reasons described by Galvin. Apart from this sentence, the statements made in my review are correct, and are not in conflict with the facts stated by Galvin. The older measurements showing a rise in sea level mostly come from Europe and not from North America. The new measurements showing an acceleration come from satellite observations of the ocean surface.

Cyril Galvin is an expert on this subject and I am not. I take my information from the paper “Twentieth Century Sea Level: An Enigma,” by Walter Munk, Proceedings of the National Academy of Sciences, Vol. 99, pp. 6550–6555 (May 14, 2002). Munk is a well-known oceanographer at the Scripps Institution of Oceanography, University of California, San Diego. The enigma discussed in his paper is a large discrepancy between the directly observed sea-level rise and the rise deduced from other evidence. This discrepancy, like the discrepancies reported by Galvin, demonstrates that both the past and the future of sea-level rise are poorly understood. The effect of Galvin’s comment is to strengthen rather than weaken the main point of my review, which says that all our theories of the Earth’s biosphere are incomplete and uncertain. This is also one of the main points emphasized by Smil in his book.

This Issue

August 14, 2003