Using a combination of experimental data taken from ice cores from the frigid glacier of Greenland and computer models of global circulation, scientists have partially solved a mystery about a lagged melting of Greenland 20,000 years ago, when the glaciers began retreating. The results by Buizert et al in Science provide important insight into the flaws in particular measurements we use as corollaries for measuring CO2 and temperature in the past and how will help to improve climate models in the future.
The CO2-Temperature Connection
Scientists have strong evidence that carbon dioxide (CO2) levels in the atmosphere and temperature are closely linked. As more CO2 enters the atmosphere, more infrared radiation emitted from Earth is absorbed by the atmosphere and re-emitted back to the surface, thus warming Earth. This does not mean CO2 and temperature should rise at the same rate! There’s no reason to expect this, given the complexity of feedback systems in the climate system, but a doubling of CO2 in the atmosphere is predicted to lead to a 2-4 degree Celsius increase in mean surface temperatures.
The CO2 atmospheric increase is happening very quickly now, quicker than ever before in the known history of the planet, but we can also look at slower patterns of CO2 in the atmosphere throughout Earth’s history, oscillations that are part of the natural variability or cycles of the climate. These patterns are crucial to understand the Earth’s feedback systems. Beginning about 23,000 years ago, ice sheets across the Northern Hemisphere began to melt and, over thousands of years, sea levels rose over 100 meters as all that ice turned into ocean water. Scientists measured CO2 and temperature (more on how they did that below!) in Arctic and Greenland regions – the Arctic data showed levels indicating melting around 20,000 years ago, however Greenland shows a lagged melting starting almost five thousand years later! This discrepancy has mystified scientists until now. To understand the results, we need to take a slight detour to understand what the scientists measure to determine CO2 and temperature levels, and how Buizert et al may have found some flaws…
Drilling in the Far North
Scientists can determine CO2 levels during these periods far in the past by taking ice core samples from the regions of interest. As snow falls, it traps air bubble in layers of ice that continue to get pushed deeper and deeper, farther from the surface. Scientists can extract samples, measure the CO2 levels, and determine how old they are (there is a tremendously detailed Wikipedia page on this for more information). This requires advanced equipment and extremely hardy researchers – an impressive scientific feat!:
Now, scientists also want to measure temperature from those same periods so they can correlate CO2 with temperature to understand how they have behave together or affect one another. This is a little tricker – proxies are required that can be used to infer temperature based on related chemistry principles. Usually, oxygen isotopes are also taken from the ice cores as a temperature proxy. Isotopes contain different numbers of neutrons – more neutrons mean that given isotope is heavier. Oxygen exists in two main isotopic forms – O18 and O16, where the number indicates the sum of protons and neutrons. Oxygen always has 8 protons, so O18 has 10 neutrons compared to 8 in O16.
Now, being lighter, there is a slight preference for O16 to evaporate compared to the heavier O18, since it is easier for the lighter isotope to overcome the gravitational force. As the reasoning goes, increased temperature leads to more evaporation, which leaves behind more O18 in the water. Thus, researchers use the O18/O16 isotopic ratio in ice cores as a proxy to measure temperature and correlate this with CO2.
This method had seemed to work in most cases, but Greenland isotopic ratios showed no temperature increase in the same time period when Arctic and Southern Hemisphere measurements did, and when CO2 concentrations were rising across all regions. Buizert et al decided to explore the issue a bit more.
The answer to the mystery is twofold – one determined from computer modeling, the other by using a different isotope proxy (nitrogen) to measure temperature.
The Seesaw Effect
Buizert et al used general circulation models (GCM), the gold standard for climate modeling, to examine how melting freshwater distributes across the globe. Remember that we’re looking at a time 20,000 years ago when all the glaciers are melting, adding enough freshwater to significantly disturb the status quo. The authors found that this huge flux of freshwater that prevents the Northerm Hemisphere from warming. Normally, there is s a large north-south circulation loop in which warm water moves from the Equator to the Northern Hemisphere, rising and warming the frigid northern waters. The glacial loss led to freshwater that essentially blocked this salt water, keeping it in the Southern Hemisphere and creating a temperature differential. Thus Greenland stayed cold, or at least colder than the Southern Hemisphere and other regions of the Arctic.
A New Temperature Proxy
But Buizert et al didn’t stop there – they also took a different type of measurement from Greenland ice cores – nitrogen isotopes. As more and more snow falls, the snow below the top layer becomes denser to the greater and great weight forcing down on it. This creates a change in the ratio of isotopic N15 ( to N14, I believe) that ends up being temperature dependent. Thus, the authors could also use this as a temperature proxy, and, when they analyzed the data in this way, found a small temperature increase in the Greenland data at the same time as other regions of the world.
This finding is important in multiple ways. One – this combined with the circulation model gives a more complete story. Incoming freshwater blocks the saltwater circulation that normally warms the North, leading Greenland regions to be colder. But there was a temperature increase correlating with the CO2 increases around the same time, just not the same magnitude as other parts of the world due to the circulation effects. So there were multiple mechanisms at play!
Second, these findings indicate that the oxygen isotope measurement may have its flaws – more research will now need to be done to understand when it works and what causes it to go awry. But these findings are very helpful in better understanding a new climatic mechanism – freshwater flux from glacial melting – and providing insight into the reliability of experimental measurements.
Buizert C, Gkinis V, Severinghaus JP, He F, Lecavalier BS, Kindler P, Leuenberger M, Carlson AE, Vinther B, Masson-Delmotte V, White JW, Liu Z, Otto-Bliesner B, & Brook EJ (2014). Greenland temperature response to climate forcing during the last deglaciation. Science (New York, N.Y.), 345 (6201), 1177-80 PMID: 25190795