Corals there grow one to two centimeters a year, laying down a calcium carbonate skeleton with a chemistry that depends on the water it formed in. In these skeletons, Cole and her colleagues looked at the strontium-to-calcium ratio and at different isotopes of oxygen because both track temperature. “As temperatures get warmer, we get less strontium in the coral,” Cole said. Sampling the extracted cores of fossilized corals millimeter by millimeter yields more than a dozen measurements per year—a monthly temperature log for the life of the coral colony.

Living colonies cover recent decades and can be checked against instrument data. The deeper history comes from dead coral heads, which can be dated using uranium-thorium methods to within a few years, even as far back as 500 years. “We were also able to find corals that are dead and washed up on the beach,” Cole said. “They may be 100 years old, 200 years old, or 900 years old, or even older.”

The result of sampling corals was a patchy but long temperature log—28 time-separated series from five islands, covering scattered years back to around 1100 CE. Set against that baseline, the last four decades look rather strange.

Fascinating stuff, we can almost map it to the Norman Conquest. Does anyone know if we have this data openly available anywhere?

DOI for Recent strengthening of eastern Pacific ENSO in the last millennium paleorecord is paywalled, unfortunately

This year is going to be fucked up.