Climate Analysis from Vostok Ice Core Data

Objective

One of the primary results of paleoclimate research over the past decade has been evidence for human-influenced (anthropogenic) global warming. Results have been based on ice cores taken from undisturbed ice sheets, such as those in Antarctica . Ice cores provide uninterrupted information on important properties of paleoclimate, including local temperature, precipitation rate and humidity, as well as changes in atmospheric composition. In this lab we will work with the original data from the famous Vostok Ice Cores, which "carry the distinction of being the only ice cores that scientists are certain have remained undisturbed for the last interglacial and the penultimate glacial periods" (AGU, 1995).

Vostok is the Antarctic research base founded by the Soviet Union in 1957. Even though the US has bragging rights to a research base at the South Pole, the (now) Russian base at Vostok is much more important scientifically. First of all it sits over one of the thickest (>4km) portions of the East Antarctic Ice Sheet. Second, this ice is relatively undisturbed (elsewhere the ice has been subject to substantial flow and/or seasonal conditions that might destroy the ice's record). Third, under the 4km ice sheet at Vostok lies a lake the size of Lake Ontario that has been cut off from Earth surface conditions for perhaps 15 million years. In a recent, exciting scientific advance researchers drilled through the last bit of ice and have made contact with the lake water.

Antarctica with Lake Vostok circled

The subglacial Lake Vostok system

Part 1. Ice and Gas Ages in the Vostok Core

Open the excel file entitled totalvosdata.xls. Click on the Vostok tab of the Excel sheet. This dataset contains columns that give the depth (in meters) of the ice core, the "ice" and "gas" ages (in thousands of years ago), concentrations of carbon dioxide and methane found in the ice bubbles, the hydrogen isotopic ratios ("delta D"), and a column that provides information on dust.

The ice age (i.e., the age of the ice, not to be confused with our other use of the term Ice Age, which refers to a time in geologic history of pronounced glaciation) is obtained by counting layers of ice and, when layers are no longer clearly visible, modeling the flow of merged ice layers.

Annual ice layers

Continental ice sheets

The gas age is calculated assuming that the bubbles of gas can only be trapped effectively in layers of older ice (i.e., at a depth well below the surface, where pores in the ice close, sealing the air). This process is called sintering.

Sintering process - Raynaud, 1992

Plot both the ice age and the gas age as a function of depth.
(These instructions are for PC computers, instructions may be slightly different for a Mac.)
First select the entire depth column by clicking on the "A" at the top of it, then hold down the control key, and then select the ice age column by clicking on the "B" at the top. Once both are selected, click on the "Insert" tab, and insert a "Scatter with Straight Lines" plot. Now, right click on the graph and click "select data."  Add new series by clicking "Add." For plotting gas age, you will need to specify the numbers themselves, not the whole column. The formula for the X values should look like this: =Vostok!$A$3:$A$196 and Y values should be: =Vostok!$E$3:$E$196. Change series names in "Edit" to "Ice Age" and "Gas Age," respectively. Now click "OK", then click OK again to generate the graph. Give your chart appropriate labels for the title and axes and also add a legend- axes labels, titles, and legends can be found under the "Layout" tab. Be sure to label the appropriate units. If you don't know what they should be, then re-read this lab again from the beginning.

Question 1

The two age curves differ - why? How much younger, roughly, is a bubble of gas than the ice that surrounds it, at a depth of 1000 meters (hint: for this part you may need to look at the raw data for 1000 meters instead of your graph)? Be sure to include your complete graph in the assignment you turn in.

Part 2. The Temperature Record

The temperature of the environment when the ice or sediment in a core was deposited can be estimated using isotopes of hydrogen or oxygen. The Vostok record is most complete for the hydrogen isotopes, but the concepts of understanding how the isotope-temperature relationship works is easier with oxygen. In addition, oxygen can be used to determine temperatures from ocean sediment cores as well.

Oxygen has two stable isotopes of importance, ¹⁶O and ¹⁸O, which vary in mass. Differences in the amounts of these isotopes in a sample (air, water, ice, rocks, organisms) are measured by comparing the ratio of ¹⁸O/¹⁶O in a sample to that ratio in a standard, which for oxygen is average ocean water. This comparison is called d¹⁸O (pronounced delta-18-O). Variations in the d¹⁸O of the oxygen in the water molecule, H₂O, can be useful in understanding the hydrological cycle and the cycle of glaciations. Average ocean water has a value of 0‰ d¹⁸O (‰ is pronounced permil and represents one-thousandth. It is analogous to percent (%), which is one-hundredth).

Isotopic signature during glacial and interglacial periods

When ocean water evaporates, water with the lighter oxygen isotope (¹⁶O) evaporates more easily because it is lighter than a water molecule with the heavier oxygen isotope (¹⁸O). Therefore, water vapor in the atmosphere ends up with a smaller percentage of ¹⁸O in it than ocean water. Its d¹⁸O is a few permil negative, say around –3‰. When that vapor passes over land and condenses to form rain, the heavier isotopes that did make it into the clouds condense to a greater extent than the lighter isotopes, again due to mass. Precipitation that falls then has a more negative d¹⁸O than seawater, but more positive d¹⁸O than the clouds from which it fell. This effect is more pronounced in cold climates than in warmer ones, because temperature is what drives the evaporation/condensation processes. Because temperature drives this process, an equation can be derived to relate temperature to d¹⁸O (or Hydrogen). In fact, a decrease of one part per million d¹⁸O in ice reflects a 1.5°C drop in air temperature at the time it originally evaporated from the oceans. Therefore, times of glaciation are associated with more negative d¹⁸O in ice samples.

Now we turn to hydrogen, because its isotopes are used in this activity (and, it was the first data, before oxygen, to be produced from the ice cores). The hydrogen isotopes of interest to paleoclimate studies are ¹H and ²H. Since ²H is such an important isotope to nuclear chemistry, it has its own name: Deuterium. The fractionation of heavy and light hydrogen isotopes behaves similarly to oxygen isotopes. Both oxygen and hydrogen isotopic fractionation are related to temperature (and thus if you have one you can relate it to the other), and in this activity, we use the hydrogen isotopes.

Structure of Deuterium

We want to convert the deuterium isotopic ratios to temperature changes (delta temperatures) that describe variations in the temperature of the ocean from which the ice was originally evaporated. This has been determined by past studies, and you can reproduce this equation in Excel by doing the following. Type "delta temp" in the first row of a new column (column H), and "(deg C)" in the second row. Now use Excel to calculate the delta temperatures, filling in your new column. Note that the "delta temperature" is relative to the current temperature today.

Use the following conversion factor: Type "= (C3 + 440) / 6.2" into the first cell, then put the cursor on the lower right corner of the cell until it turns into a gray cross. Click and hold the left mouse button and drag down to the end of the column. The cross cursor applies the relationship of the first cell to other cells.

Highlight the column (right-click Format Cells) and give the numbers two decimal places.

Save your work. Think about why the delta temperatures change with deuterium isotope ratio in the way they do. If you don't understand, then you should be sure to read through the suggested readings listed at the end of this lab and look at the web notes for the lecture on Climate Change and the Ice Ages on Canvas, Lecture Schedule.

Question 2

Plot the delta temperature curve as a function of ice age (ice age is on the X axis; use a "Scatter with Straight Lines" graph) and copy it into your report. How many degrees of temperature has climate varied in the past, as indicated in these data? Over what time scale does this occur?Be sure to include your graph in the assignment you turn in.

Question 3

According to your graph, approximately when did the last period of full glacial conditions begin and end (no transition zones)? When did the current interglacial time period begin and end? Be sure to include your graph in the assignment you turn in.

Part 3. The Atmospheric Composition

Another piece of information that comes from ice cores is the amount of dust that was in the atmosphere. During the last ice age, sea level was 100-120 meters lower and much more land was exposed . In addition, it was much colder, and that had an effect on the distribution of vegetation. (We will not explore the dust record in this activity.) In this section, you will put together the different pieces of information that come from the core. Use the following graphs of CO₂ and CH₄ concentration, as well as delta temperature as a function of age, to answer the questions for this section.

Carbon dioxide and methane concentrations plotted with delta temperature

Question 4

Note the time of the major warming events. Then look at how CO₂ and CH₄ change during the same time. From the data provided in this lab, can you tell which changes first, temperature or greenhouse gas (CO₂, CH₄) composition? (Note the direction of the x-axis.) Explain the pattern you see in the graph.

Part 4. What can we learn from this analysis of past climate about climate on earth today?

Now we will reproduce the steps necessary to convert the isotope data to the air temperature of the past. First, add another column to your Excel file to estimate the temperature at Vostok. Label the column "Vostok Temp", with units deg C. To calculate the values for this column, subtract 55.5 degrees from the numbers in the delta temperature column. The resulting value is an estimate of the Vostok air temperature itself.

Create scatter plots (without connecting lines) of CO₂ (x-axis) vs. temperature (y-axis) and CH₄ (x-axis) vs. temperature (y-axis) using the Vostok data. Use the Insert a linear "trend line" (which is found under the layout tab) and report the equation and r² value by checking those boxes in the option menu for trend lines of each scatter plot in your lab report. The r-squared (r²) is a statistical measure of how well actual data fit a linear regression model (note: the r² is NOT the radius squared and is NOT the R² inverse-square law); an r-squared of 1.0 (100%) indicates a perfect fit of the data to a straight line. So if you had an r² of 1.0 for the CO₂ vs. temperature plot, if you knew the CO₂ concentration you could predict the corresponding temperature 100% of the time (in the real world, very few relationships are 100% perfect).

Question 5

Copy your scatter plots with trendlines into your report. Why is the r² value for CO₂ or CH₄ less than one? In other words, explain why the relationships are not perfect fits (the fit of the trendline to the data is not perfect, there are points that do not lie right on the trendline)?

Question 6

Predict the temperature at Vostok today. Use today's CO₂ concentration (~400 ppmv; see https://www.co2.earth/) to solve the linear regression equation from the past relationship between CO₂ and temperature (Q6). How does this calculated temperature differ from the surface temperature today at Vostok? Explain why these may be different. (You can look up this temperature with a weather website (e.g., https://goo.gl/K4m8iL); if necessary, convert temperature to Celsius.)

Submit a document containing answer to questions 1 through 6 and graphs for questions 1, 2 and 5.

Sources

Monnin et al. "Atmospheric CO2 concentrations over
the last glacial termination" Science v.291, 112-114, 5 January 2001.

Dansgaard, W., H.B. Clausen, N. Gundestrup, C. U. Hammer, S. J. Johnsen, P. M. Kristinsdottir, and N. Reeh, A New Greenland Deep Ice Core, Science, Vol. 218, 1992, p.1273-1277.

"Deciphering Mysteries of Past Climate From Antarctic Ice Cores" Earth in Space (American Geophysical Union), Vol. 8, No. 3, November 1995, p. 9.

Crane, Robert G., James F. Kasting, and Lee R. Kump. The Earth System. New Jersey: Prentice Hall, 1999.

http://www.aad.gov.au/asset/images/525_ul-core.jpg

A great source of paleoclimatological data is the NOAA web site.