Estimating past sea-surface temperatures from corals

E. Linacre and B. Geerts


What are corals?

Corals are animals of the phylum Coelenterata, to which also the jellyfish and sea anemones belong. Some species of corals have stony skeletons, consisting almost entirely of calcium carbonate (CaCO3), and the term coral is often applied to the skeletons themselves. Most atolls are built on thick layers of coral residue that accumulated over millions of years on top of submerging volcanoes. There are three kinds of this skeletal material, i.e. plate-like, branching, and ‘massive’. The last is rounded and bulky and proves to be useful for estimating past sea-surface temperatures (SST) in tropical regions (1).

Tropical coral reefs grow mainly where the SST exceeds 18°C, and optimal growth occurs at about 25°C. At 25°C the growth rate of massive corals is about 50% faster than at 18°C. The dependence on temperature means a variation of growth rate with season. In the past there has been confusion in interpreting coral cross-sections, due to uncertainty about the process of coral growth. Some corals appeared to grow fastest in the warmer wet season and others in the cooler dry season.


Coral growth layers

X-ray examination reveals that massive coral has layers of different density, due to seasonal variations, like the annual rings of tree trunks. Counting of the density layers in large colonies of coral provides annual dating of the layers for several hundreds of years. Massive coral cores of the Porites type on Australia’s Great Barrier Reef (GBR) have been dated back to 1479 AD (1). New skeleton is not only added to the outside of the existing material, but also to a depth of several millimetres. So the process involves two aspects - the building of a scaffolding of skeleton, and subsequent infilling of surface tissue. On the GBR the average growth rate is around 12 mm/year, varying between 5 and 30 mm/year.

The average annual calcification rate for ten very large coral colonies on the GBR was relatively high during 1769-98 and 1940-1960, presumably indicating warm conditions, but low between 1819-30, indicating colder conditions. These 10 coral cores were calibrated by matching with water temperature measurements since 1905 (3). There appear to have been 12 major swings of around 0.6 K below the present, during 1745-1905, i.e. roughly once each 14 years. The corals also indicate a warming of over 1 K during 1979-1997.

However the growth rate only gives an indication of temperatures when the temperature is a limiting factor (18-22°C) and when other factors do not limit the growth. There are two more accurate methods to infer sea surface temperature variations from the annual growth layers of massive coral.


Oxygen isotope method

The ratio of isotopes of oxygen (18O/16O) in layers of the calcium carbonate provides an accurate measure of the SST at the time of forming each layer, and the ratio of carbon isotopes (14C/12C) is used to confirm the age of the older specimens. Anomalously large concentrations of trace elements in the skeleton, such as barium, cadmium and manganese, reflect more-than-normal deep-ocean upwelling and/or river discharges. Certain inshore corals contain fluorescent bands which glow in UV light and provide evidence on the local rainfall runoff.


Strontium/calcium method

The coral absorbs less strontium when the temperature rises, and temperatures can be determined to within 0.5 K from the strontium/calcium ratio. This technique allows monthly 'mean' temperature anomaly estimates, from slices of coral corresponding to only a month's growth (2). The gradualness of ocean-temperature changes makes them less susceptible to the random disturbances which affect air and land temperatures, so underlying temperature trends are discerned more readily.



  1. Lough, J.M. and D.J. Barnes 1997. Coral records of past climates and environments. Aust. Meteor. & Ocean. Soc. Bull. 10, 84-90.
  2. Shorter, D. 1997. ANU taps coral secrets. Report in the Canberra Times of 25/3/1997 of work by M. McCulloch and C. Alibert at the Australian National University.
  3. Lough, J.M. & D.J. Barnes 1997. Several centuries of variation in skeletal extension, density and calcification in massive Porites colonies from the Great barrier Reef. J. Exptl. Marine Biol. & Ecol., 211, 29-67.