Uncertainty over weakening circulation

March 2007, page 14

Barbara Goss Levi's Search and Discovery story (PHYSICS TODAY, April 2006, page 26) discusses evidence of weakening ocean circulation and its possible connection to global warming. The Atlantic Ocean circulation across 25° N latitude has been used as a benchmark for characterizing the mass and heat transport from the tropics to the northern latitudes. The upper portion of this transport includes the Gulf Stream, which is at least partially responsible for a moderate climate in Europe. A weakening of the Atlantic meridional overturning circulation and of the Gulf Stream might have the unpleasant consequence of cooling Europe's climate.

The PHYSICS TODAY piece is based on analysis of work by Harry Bryden, Hannah Longworth, and Stuart Cunningham,¹ which concluded that the Atlantic meridional overturning circulation slowed by about 30% between 1957 and 2004. Their work inspired speculations that the anthropogenic increase in carbon dioxide may be responsible for the weakening of heat transport from the tropics, and that such an effect has now been detected.

The conclusion that the Atlantic meridional overturning circulation has decreased by 30% does not follow from the data presented by Bryden and coauthors, but is based on an incorrect treatment of measurement errors.

According to Bryden and coauthors, the 1957 transport in a layer shallower than 1000 m was 22.9 ± 6 Sverdrups (1 Sv = 10⁶ m³/s) compared with the transport of 14.8 ± 6 Sv in 2004. The ± 6 Sv represents an uncorrelated error of each measurement. Bryden subtracts the two quantities and presents the results as 8.1 ± 6 Sv (instead of 8.1 ± 12 Sv or ± 8.5 Sv, depending on the character of errors), which is an incorrect result. It is a mystery how such an error was missed by Levi and by the editors and reviewers of the original paper. The observed change of 8.1 Sv is well within the uncertainty of the measurement. The correct conclusion from the data presented in Bryden's paper should have been that no statistically significant change in Atlantic meridional overturning circulation at 25° N between 1957 and 2004 has been detected. Such a conclusion is in agreement with the earlier analysis of essentially the same data (between 1957 and 1999) by Alexandre Ganachaud and Carl Wunsch.²

Research also failed to detect any slowing,^3,4 and one of the relevant papers⁴ concludes that "there is no sign of any Meridional Overturning Circulation slowdown trend over the past decade, contrary to some recent suggestions."¹

In defense of Bryden and his coauthors, I must share a comment from a personal communication I had with Bryden shortly after his Nature paper was published. Bryden's paper as submitted for publication to Nature included a question mark at the end of the title, suggesting only a possibility that the circulation might be slowing down. On the editor's insistence, the question mark was removed, and the title was changed into a positive statement that caused a considerable stir.