Science Heresy - November 2011


The Variance of Sea Surface Temperature

The upper map shows the variance of sea surface temperatures calculated by a climate model. The lower map shows SST variances computed from observations.

Clearly there are major differences between the two maps; observed variances are generally larger than model variances. The effect of El Nino is evident in the model variances, N, but not in the observational variances, due to differences in the way the two data sets were processed. The effect of variations in western boundary currents to the east of Japan, Labrador, South Africa, New Zealand and Argentina is discernible in both figures, W, but is much more widespread in the observations. There are large observational variances in the Southern Ocean south of 40oS in the region of the Antarctic Circumpolar Current, A, not matched by the model variances which are quite small in this region. There are some large mid-ocean observational variances, M, which are well away from western boundary currents and the Antarctic Circumpolar Current but close to Mid-Ocean Ridges (MORs).


Circulation Forcing by Ocean Floor Heating

It has long been known that the ocean floor is heated from below tectonically as seawater penetrates the crust and comes into contact with hot magma. The gross heat flux at any one time is estimated to be of the order of 10 TW (1 terawatt = 1000 gigawatt = 1 million megawatt). This flux has been considered by oceanographers to be distributed uniformly in both space and time so that when divided by the total area of the oceans a uniform flux density of about 40 mW m-2 is calculated. This number has long been dismissed as negligibly small. However the discovery of hydrothermal vents (HTVs) in the 1970s has led to the realization that this heat flux is not uniformly distributed but concentrated in vent fields distributed along ribbon-like mid-ocean ridges (MORs) and volcanic arcs so that local heat flux densities can be considerably greater than this.

 The global heat flux emitted along the axes of MORs is estimated to be from 2 to 4 TW  compared with 2.4 TW for ocean dissipation of tidal friction and 0.88 TW for wind stress power.

The local effects of HTVs can be observed as the “black smoker” plumes which are observed to rise only 100m or so above the sea floor, where they dissipate by entrainment of cold water and are advected horizontally by currents. Their apparent confinement to the bottom 100m meant that they continued to be ignored as a major factor in ocean energetics.

However there are two matters which have yet to be taken into account, viz.: 

  1. HTVs are not uniformly distributed along the MORs and occur in much greater densities in places where the ridges are spreading more rapidly and
  1. tectonic activity under the ocean is likely to be as highly intermittent and sporadic as it is on land.

It is reasonable to assume that the incidence of submarine volcanoes is similarly distributed and is perhaps 4 times that of volcanoes on land because of the greater area of the ocean and thinner crust.

Plumes from larger sub-sea volcanic events will rise higher above the sea-floor than those emitted by the more common and relatively steady HTVs. Such plumes are called “megaplumes”.  These have been observed to reach a height of 1000m above the ocean floor. The height to which a plume ascends above the ocean floor is approximately proportional to the logarithm of the flux density at the ocean floor, each tenfold increase in flux density leading to a further rise of 550m.

Although the megaplumes observed by Baker et al (1989) were observed well below the thermocline, it is not difficult to conceive that, for a sufficiently large eruption a megaplume generated by a submarine volcano could penetrate the mixed layer and affect the SST on scales of tens to hundreds of km.

Complexities of plume behaviour imply that a volcanically generated plume need not necessarily lead only to increases in temperature at the surface. Entrained cold water from below the thermocline would have a cooling effect on SST. This would be particularly true of bubble plumes.

A further consideration is that, for sufficiently thick plumes from very large eruptions, the temperature near the centre of the plume may remain sufficiently high to exceed the local boiling point as the plume rises and pressure decreases with decreasing depth. Steam bubbles would form, greatly enhancing plume buoyancy and massively increasing the amount of kinetic energy and upward momentum being communicated to the surrounding water mass.

A large enough major eruption could thus lead to the overturning of an entire ocean basin which would continue until the extra kinetic energy is finally dissipated by bottom friction. Such basin overturning would lead to a significant drop in global atmospheric temperature like that recorded during the Younger Dryas for example.



A PDF of the complete paper on which this article is based may be downloaded here

November 2011