2014 GRL article

Dynamical changes in the tropical Pacific warm pool and zonal SST gradient during the Pleistocene

Kelsey Dyez and Christina Ravelo

Published in Geophysical Research Letters, 2014

Spotlight summary of the article is here.


Carbon dioxide is a gas in the atmosphere produced by many natural and human processes. Naturally, it is produced as a byproduct of animals digesting their food and by decomposing organic material. Humans produce carbon dioxide gas by burning fossil fuels and deforesting landscapes. The amount of carbon dioxide in the atmosphere is currently going up as a result of human activity, so it is important to know how increasing carbon dioxide will affect Earth’s temperature and the average weather (climate) that humanity will experience. This knowledge is important for food crops, ecosystem management, building practices, weather prediction, drought prediction, forest fire prediction, etc. Such knowledge might induce policy changes to lower human contributions to carbon dioxide emissions.

One way to establish the relationship between temperature, average weather, and carbon dioxide is to look at times in the past when carbon dioxide was different from today and see how the climate was different. In this article, I look at the past 1.5 million years, a period of time known as the Pleistocene, which is long enough for the climate to have warmed and cooled ~15-20 times, enough to establish patterns for how climate warms and cools with changing amounts of carbon dioxide.

One great place to get constant records of climate changes is the chemistry of sediments from the bottom of the oceans. Compared to records from on land, temperature records from the ocean are not complicated by local effects of shade from trees or changes in altitude as mountains rise or wear away. Speaking of erosion, ocean sediments slowly build up continuously; on land dirt accumulates on the landscape sporadically, sometimes building up all at once in a flood, or even getting blown away, which makes climate records from on land easier to access, but harder to interpret.

Besides carbon dioxide, another idea for something that might change Earth’s temperature is natural changes in ocean circulation. Ocean circulation is chaotic and depends on many factors: wind over the ocean, the amount of salt in the ocean, the position of the continents, sea level, the difference in temperature between different parts of the ocean, etc. There is a straightforward way to test whether changes in temperature are due to carbon dioxide or changes in ocean circulation, and the argument goes as follows.

Over the course of a few months, carbon dioxide is mixed very well in the atmosphere by the wind. So it doesn’t matter where carbon dioxide is produced, it should have the same heat-trapping effect everywhere and raise the temperature of the Earth’s surface by the same amount everywhere. If we see places where the temperature changes by different amounts, it must be due to something other than carbon dioxide.

It has been known for a long time that when carbon dioxide goes up and down, the Arctic and the Antarctic warm and cool a lot more than other parts of the world and the oceans near the equator don’t change temperature nearly as much. It is thought that this “amplifying” effect at the poles is due to a feedback like the amount of ice covering the land surface. Just like feedback in a microphone, a small initial increase in carbon dioxide over the Arctic and Antarctic could start a chain reaction with the ice that in the end makes the temperature rise more than it would from just the carbon dioxide effect by itself.

In the wide open oceans near the equator, many of these types of feedback are not present: there isn’t any snow, permafrost, or ice cover, there can’t be forest fires, and there isn’t any vegetation on the surface to change. So for some time, it was thought that carbon dioxide might be the only thing that changed the temperature of the ocean near the equator. Here we show this cannot be the case.

Main points of arcticle:

In this article we report on four different locations, all in the open ocean near the equator in the Pacific ocean. Settling particles from the ocean surface were deposited on the ocean floor in layers and cores punched through these layers were collected in the 1990’s by a scientific drill ship and stored in giant refrigerators in Texas. Scientists had already analyzed the chemistry of three of these cores; in Santa Cruz I analyzed the chemistry of the fourth to get the past ocean surface temperature and I put all the records side-by-side to see how they were different. Differences between the temperature records cannot be due to carbon dioxide, and so must be due to changes coming from the ocean itself.

Here we show for the first time two ways the ocean was different in the past, which contributed to temperature changes beyond simply the effects of carbon dioxide. We show that in the west tropical Pacific, the warmest large body of surface water on the planet, at certain times in the past, colder water came up from the deeper in the ocean and cooled the surface. This is a new idea that hasn’t had a lot of support before. The other way the ocean changed in the past was that in the eastern Pacific, water temperature near the Galapagos cooled much more as the earth developed large icy glaciers in the Arctic and the Antarctic. This idea had been proposed before, but here it is drawn in sharper focus as we have another temperature record to use as a comparison. These ideas mean that we have to think about not just the temperature of water at the surface but also temperatures below the surface and how water moves within the ocean as the earth warms in the future.