Oxygen-Isotope Ratios

by Mark A. Tinsley
in partial fulfillment for the course ES 767, Quaternary Geology
November 2006

Oxygen-Isotope Dynamics
Isotopes (Generally)
Application of Oxygen-Isotope Dynamics to Paleoclimate
Oxygen Isotopes (Specifically)
The Practice of Oxygen-Isotope Studies


Without a doubt, an understanding of ancient climate is one of the most critical studies underway in Quaternary geology.   What were temperatures like during the last ice age?  How was vegetation distributed across North America during the late Wisconsin glaciation?  What caused camelids to migrate to North America during the Pleistocene Epoch?  These and countless other questions are the types of inquiries that drive Quaternary geology.  Furthermore, these are the types of questions that cannot be answered without an understanding of paleoclimate.  Unfortunately, no method exists to measure ancient climate directly (Aber, 2006).  As such, science has developed many proxy methods for conducting paleoclimatic studies.  This webpage is dedicated to one of these methods, namely, oxygen-isotope ratios.  In the paragraphs that follow, the subject of oxygen-isotope ratios will be systematically dissected and explained.  Subtopics include:  isotopes (generally), oxygen-isotopes (specifically), oxygen-isotope dynamics, application of oxygen-isotope dynamics to paleoclimate, and the practice of oxygen-isotope studies.

Isotopes (Generally)

Naturally, the inaugural question in any study such as this is the one regarding definition.  That is, "What is an isotope?"  One of the better extant definitions is that proffered by the INEEL Environment Surveillance, Education, and Research Program of the Idaho National Laboratory.  The INEEL defines isotopes as, "Two or more forms of an element having the same number of protons in the nucleus (or the same atomic number), but having different numbers of neutrons in the nucleus (or different atomic weights). Isotopes of single elements possess almost identical chemical properties. An example of isotopes are plutonium-238, plutonium-239, plutonium-240, and plutonium-241, each acts chemically like plutonium but have 144, 145, 146, and 147 neutrons, respectively" (Case et al., 2003).  In other words, an isotope is an element that is like other elements with the same atomic number except that it has a different number of neutrons.  Furthermore, since the number of neutrons is directly proportional to the atomic mass of an element, isotopes with a greater number of neutrons are weightier than those with fewer neutrons.  Using the INEEL illustration above, then, one can conclude that plutonium-240 is more massive than plutonium-238; that is, plutonium-240 weighs more than plutonium-238.  This principle will become important during the discussion on oxygen-isotope dynamics.

Isotopes can be either stable or unstable.  Stable isotopes are those isotopes that do not experience radioactive decay, whereas unstable isotopes are those that do experience radioactive decay.  Examples of unstable isotopes abound and include such infamous elements as Uranium-238, Plutonium-239, Carbon-14, and Phosphorus-32.  Over time these elements emit alpha, beta, or gamma particles to become wholly different elements and/or elements with a lower energy state; that is, they radioactively decay.  An alpha particle is nothing more than a helium (He) nucleus; therefore, alpha decay is when a radioactive isotope emits 2 protons and 2 neutrons (i.e. one He atom) from its nucleus (Radioactive Decay, 2000).  For example, uranium decays to thorium via alpha decay:

23892U  ---->  23490Th + 42H

Beta particles, on the other hand, are electrons that are emitted from the nucleus.  These nucleus-derived electrons are produced when a neutron decays to a proton in a process called weak interaction (Beta Reactivity, 2006).  One can think of this process as the splitting of a neutron into one proton and one electron.  The proton remains in the nucleus, thus converting it to a new element.  The electron, however, is emitted.  The following is an example of beta decay:                                                           

23490Th  ---->  23491Pa + 0-1e

Finally, gamma decay is the process whereby an element's nucleus goes from a higher energy state to a lower one via the emission of electromagnetic radiation (i.e. photons of energy).  Since the number of protons and neutrons does not change during gamma decay, the parent and daughter elements are the same (Gamma Decay, 2000).  The difference is that after gamma decay, the daughter element is less energetic.  Figure 1 is an illustration of gamma decay.

Figure 1:  Gamma Decay of the element Dysprosium (Dy).  Taken from Gamma Decay (2000).

Radioactive isotopes can be extremely useful in the dating of natural materials.  The only information needed is the half life of the element and its original concentration in the substance tested.  Carbon dating is a common radioactive isotope test performed on organic substances.  In Quaternary studies, however, scientists have found stable isotopes (i.e. those that do not decay) much more useful.  Two of the more common stable isotopes used in Quaternary geology are those of hydrogen and oxygen.  It is the latter that will be discussed herein.

Oxygen-Isotopes (Specifically)

Roughly fifteen isotopes of oxygen are known to exist.  These are Oxygen-12 through Oxygen-26.  Of these fifteen, however, only Oxygen-16, Oxygen-17, and Oxygen-18 are stable isotopes.  Furthermore, Oxygen-16 is the most abundant, constituting 99.762% of the Earth's oxygen (Firestone, 2000; Gibson, 2005).  Known as "light" oxygen, Oxygen-16 is composed of 8 protons and 8 neutrons (Herring, 2006).  Applying the definition of "isotopes" above, it is reasonable to conclude that Oxygen-17 and Oxygen-18 have 8 protons/9 neutrons and 8 protons/10 neutrons, respectively (see Figure 2 below).  Of these latter two oxygen-isotopes, though, only Oxygen-18 is found in sufficient quanities to be of use to science.  Known as "heavy" oxygen--because its mass is approximately 12.5% more than the mass of Oxygen-16--and having an abundance of around 0.200%, Oxygen-18 can be used along with Oxygen-16 to determine ancient climate.  This is because the ratio of Oxygen-18 to Oxygen-16 in water changes with the climate (Herring, 2006).  Incidentally, when speaking of oxygen isotopes throughout the remainder of this webpage, reference will be made only to the Oxygen-16 and Oxygen-18 varieties.


oxygen isotopes
Figure 2:  Oxygen-16 and Oxygen-18 isotopes.  White spheres are neutrons and blue spheres are protons. Taken from Herring (2006).

Oxygen-Isotope Dynamics

Water molecules are composed of two hydrogen atoms and one oxygen atom.  Because oxygen can be found principally in two stable varieties, it stands to reason that some water molecules will contain Oxygen-16 atoms and others will contain Oxygen-18 atoms.  As noted earlier, the isotope with a greater number of neutrons will be more massive than the one with a lesser number.  Therefore, it is reasonable to conclude that water molecules bearing Oxygen-18 will be slightly heavier than those containing Oxygen-16.  This is an important point in oxygen-isotope dynamics, and one that will be discussed in the following paragraphs.

The ratio of Oxygen-18 to Oxygen-16 in the Earth's atmosphere is dependent upon the dynamics of evaporation and condensation that occur at the atmospheric boundary layer and within the atmosphere itself.  This is because water molecules containing Oxygen-16 tend to evaporate more readily than those containing Oxygen-18, while molecules containing Oxygen-18 tend to condense more easily (Herring, 2006).  As such, evaporation of water from the ocean at lower latitudes tends to increase the atmospheric concentration of Oxygen-16 in those regions due to higher rates of H216O vaporization.  However, as Hadley cells and other global/regional atmospheric convections begin transporting the newly formed water vapor poleward, cooler temperatures and upper level divergences eventually cause condensation and precipitation to occur.  During the process of condensation, the first water molecules to precipitate are those containing Oxygen-18.  Therefore, initial precipitation during a rain or snow event is typically enriched in H218O and gradually becomes more and more "light."  This process is known as the "rainout effect," and it results in atmospheric Oxygen-18 decline (Gibson, 2005).  In fact, it is the cumulative effects of rainout that result in a lateral differentiation of 18O/16O concentration in the atmosphere.  That is, as one travels from the equator to the poles, the atmosphere generally becomes more and more depleted of Oxygen-18, until precipitation at the poles contains up to 5% less Oxygen-18 than does ocean water (Herring, 2006).  This decrease in Oxygen-18 levels with increasing latitude is known as the "latitude effect" (Kendall et al., 2004).  As a result, polar ice sheets tend to evidence a significant reduction in Oxygen-18.  Figure 3 illustrates this point graphically.

Figure 3:  Illustration of Oxygen-18 depletion with latitude.  Taken from Herring (2006).

From a contemporary perspective, the world ocean is generally in a state of isotopoic equilibrium.  That is to say, though there is certainly lateral variation in oxygen-isotope ratios (implied in Figure 3 above), the average18O/16O concentration does not change overall.  This is because evaporation is balanced by condensation/precipitation worldwide.  In other words, enrichment of Oxygen-18 in equatorial regions due to evaporation is balanced by enrichment of Oxygen-16 due to both affluent glacial meltwater (see Figure 3) and cumulative global precipitation.   As will be shown later, however, this is not necessarily the case over long periods of geologic time.

The oxygen-isotope concentration (d18O) of a water sample is determined with respect to a datum known as standard mean ocean water (SMOW).  SMOW is a sample of well-mixed ocean water that contains a precisely known 18O/16O concentration and is currently stored at the International Atomic Energy Agency in Vienna, Austria (Groning & Frohlich, 2006).   Oxygen-isotope concentrations are typically expressed in parts per thousand (0/00) using the following mathematical equation:

 d18O = [(18O/16O) sample - SMOW / SMOW] X 1000

where d18O = 00/00 for SMOW (Aber, 2006; Kennett, 1982).  Presently, the world ocean has a d18O = 0.00/00, and glacial ice sheets exhibit d18O values ranging from -200/00 to -300/00 (Aber, 2006--see Figure 4).  The next section, however, will detail conditions in which these numbers experience marked flux.                                   

Figure 4:  Modern-day oxygen-isotope concentrations for the hydrosphere, atmosphere, and cryosphere.  Values for "d" are given in parts per thousand (0/00).  Information derived from Aber (2006) and Herring (2006).

Application of Oxygen-Isotope Dynamics to Paleoclimate

The three principles that drive application of oxygen-isotope dynamics to paleoclimate are the following:   

1.  As the 18O/16O ratio for landed ice decreases, the18O/16O ratio for ocean water increases.
2.  As the
18O/16O ratio for landed ice increases, the 18O/16O ratio for ocean water decreases.
3.  The
18O/16O ratio for both landed ice and ocean water is atmospheric temperature dependent.

Naturally, discussion of these principles is in order. This is best done by way of illustration.

All other conditions remaining constant, suppose that the Earth's atmosphere began to cool.   In this scenario, landed ice volume would obviously increase due to the equatorward extension of freezing conditions.  Considering latitude effect, this equatorward movement of cooler temperatures would cause rain out to begin at lower latitudes, thus decreasing the 18O/16O ratio of the polar ice sheets (Herring, 2006; Kennett, 1982).  In short, cooler temperatures worldwide would result in more water being trapped on land, and the water thus trapped would have a lower 18O/16O ratio due to the combined effects of temperature distribution, rain out, and latitude.  Furthermore, increased sequestration of water on land would cause the 18O/16O ratio for ocean water to increase.  That is, since more Oxygen-16-rich water would be locked in polar ice sheets, there would be less of it in the world ocean and, therefore, the ratio of 18O/16O would necessarily increase (Hambrey & Alean, 2004; Aber, 2006).  On the level of fundamental scientific assumption, what is being described herein is nothing more than the conservation of matter.  Thus, for a cooling scenario, the interplay of the above three principles is quite clear.

Conversely, were the Earth's atmosphere to warm, landed ice volume would decrease due to the poleward retreat of freezing conditions.  This poleward movement of cooler temperatures would result in rain out at higher latitudes and a corresponding increase in 18O/16O ratios for the polar ice sheets (Herring, 2006; Kennett, 1982).  Finally, decreased volumes of landed ice would return Oxygen-16-rich water to the world ocean, thus decreasing the overall 18O/16O ratio of seawater (Hambrey & Alean, 2004).  Again, the relationship of the above principles is easily discernable.

From these illustrations, then, two sub-principles of principle #3 (above) become evident:

3a.  As atmospheric temperature decreases, the 18O/16O ratio for landed ice decreases and the 18O/16O ratio for ocean water increases.
3b.  As atmospheric temperature increases, the
18O/16O ratio for landed ice increases and the 18O/16O ratio for ocean water decreases.

Incidentally, it is these latter two sub-principles that are directly applied to paleoclimates.  That is, when it is found in the terrestrial ice record that 18O/16O ratios are lower than presently observed or in the marine record that 18O/16O ratios are higher than presently observed, then it is assumed that paleotemperatures--and thus paleoclimates--were cooler.  On the other hand, if 18O/16O ratios in the terrestrial record are found to be higher and 18O/16O ratios in the marine record are determined to be lower than at present, then warmer paleoclimates are posited.

The Practice of Oxygen-Isotope Studies

There are essentially two geologic records from which scientists have gathered and continue to gather oxygen-isotope data.  These are ice cores from glacial ice sheets (Figure 5) and sediments from the marine environment.  The former include cores gathered during the Greenland Ice Core Project (GRIP) and Greenland Ice Sheet Project 2 (GISP2), as well as the famous Vostok core from Antarctica.  The Vostok core is, of course, the most famous ice core due to its impressive length--some 3.7 km covering over 420,000 years of Earth's history (Hambrey & Alean, 2004)!  Analysis of glacial ice easily yields oxygen-isotope ratios and has been of inestimable value in the study of Quaternary geology.

Figure 5:  Montage of photos taken during experimental ice core drilling in Greenland, 2005.  Taken from Riebeek (2006b).

However, it is the study of marine sediments that has truly revolutionized Quaternary geology.  This is principally due to the longer spans of geologic time available in the marine record.  Through the efforts of scientists such as C. Emiliani and N. Shackleton, the analysis of oxygen-isotope ratios in marine sediments has allowed geologists and climatologists to estimate paleotemperatures for nearly the entire Quaternary period (Gibbard, 2006; Kennett, 1982).  In fact, it was Shackleton who helped devise the equation [1967] that is still used today for estimating paleotemperatures from fossil foraminifera in marine sedimentary deposits:

T(oC) = 16.9 – 4.38(d18Oc – d18Ow) + 0.10(d18Oc – d18Ow)2

where d18Oc = oxygen-isotope ratio for calcite in foraminifera and d18Ow = oxygen-isotope ratio for ancient seawater (Kennett, 1982; Costa et al., 2006).  Foraminifera are used for oxygen-isotope analysis because the shells of these organisms are made from calcium carbonate (CaCO3)--an oxygen-isotope bearing compound.  These fossils act like veritable event recorders in nature.  As planktonic and benthic foraminifera build their shells, they incorporate oxygen in ratios proportional to the 18O/16O ratio of the contemporary seawater.  As such, when they die and become buried in marine sediment, they record indirectly in their carcasses the 18O/16O ratio of the seawater at the time of deposition.  The modifiers "proportional" and "indirectly" are employed here because even though the18O/16O ratio of ancient seawater can be determined from foraminifera shells, it is not done so directly.  Herring offers the best explanation for this when he writes,

"As the shells form, they tend to incorporate more heavy oxygen than light oxygen, regardless of the oxygen ratio in the water.  The biological and chemical processes that cause the shells to incorporate greater proportions of heavy oxygen become even more pronounced as the temperature drops, so that shells formed in cold waters have an even larger proportion of heavy oxygen than shells formed in warmer waters, where the difference is less notable.  This temperature-based skew effect means that the oxygen isotope make-up of shells would not precisely match the make-up of the ocean water in which they grew.  Scientists must correct for this skew if they are to learn about the ratio of oxygen isotopes in the ocean waters where the shells formed" (Herring, 2006).

Hence the reason for Shackleton's complicated formula above!  It must be noted, however, that foraminifera are not the only organisms that can be used for marine oxygen-isotope studies.  Other calcium carbonate bearing organisms, plants, and corals may be utilized, as well as those containing silicon dioxide (
SiO2) (Herring, 2006).   It is foraminifera of both the benthic and planktonic variety that have been analyzed most extensively, though, due to their shear volumes in the marine record.


As stated in the introduction, no method exists to measure paleoclimates directly.  However, through the use of oxygen-isotope ratios, science has devised a means by which to estimate ancient climate by proxy.  Though by no means infallible, the method of paleoclimate determination using oxygen-isotope ratios has nonetheless revolutionized Quaternary geology.  Today, scientists are quite confident in their estimations of paleoclimate (see Figure 6).  Countless research projects and scholarly works have been completed that have as their foundational methodology the use of oxygen-isotope ratios.  What is more, the list continues to grow.  Where the use of oxygen-isotopes will take science, no one can accurately predict; however, one thing is for certain:  oxygen-isotope studies will be part and parcel of Quaternary geology for the forseeable future.

Figure 6:  Example of a paleoclimate graph derived from oxygen-isotope ratios in marine sediments.
Adapted from Riebeek (2006a).


Aber, James S.  (2006).  ES 331/767 Lecture 11:  Paleoclimate Reconstruction.  Retrieved October 7, 2006, from the Emporia State University Web site:  http://academic.emporia.edu/aberjame/ice/lec11/lec11.htm.

Beta Reactivity.  (2006).  Retrieved November 11, 2006, from the Georgia State University Web site:  http://hyperphysics.phy-astr.gsu.edu/hbase/nuclear/beta.html.

Case, M., Halford, D., & Martin, C.  (2003).  Appendix E:  Glossary.  Retrieved November 11, 2006, from the INEEL Environmental Surveillance, Education, and Research Program Web site:  http://www.stoller-eser.com/Annuals/2003/AppendixE.htm.

Costa, K. B., Toledo, F. A. L., Pivel, M. A. G., Moura, C. A. V., Chemale, F.  (2006).  Evaluation of Two Genera of Benthic Foraminifera for Down-Core Paleotemperature Studies in the Western South Atlantic.  Brazilian Journal of Oceanography, 54(1), 75-84.

Firestone, Richard B.  Isotopes of Oxygen.  Retrieved November 11, 2006, from the Lawrence Berkeley National Laboratory Web site:  http://ie.lbl.gov/education/parent/O_iso.htm.

Gamma Decay.  (2000).  Retrieved November 11, 2006, from the Lawrence Berkeley National Laboratory Web site:  http://www.ibl.gov/abc/wallchart/chapters/03/3.html.

Garrison, Tom.  (2006).  Essentials of Oceanography (4th ed.).  Belmont, CA:  Thomson Brooks/Cole.

Gibbard, Philip.  (2006).  Professor Sir Nicholas Shackleton.  Boreas, 35, 385-390.

Gibson, John.  (2005).  Oxygen.  Retrieved November 11, 2006, from the Sustainability of Semi-Arid Hydrology and Riparian Areas Web site:  http://www.sahra.arizona.edu/programs/isotopes/oxygen.html.

GISP2--Greenland Ice Sheet Project 2.  (2006).  Retrieved November 11, 2006, from the University of New Hampshire Web site:  http://www.gisp2.sr.unh.edu.

Groning, M., & Frohlich, K.  (2006).  Part II:  Example of Reference Materials Certified for Stable Isotope Composition.  Retrieved November 11, 2006, from the International Atomic Energy Agency Web site:  http://www.iaea.org/programmes/aqcs/pdf/reference_2.pdf.

Hambrey, M., & Alean, J.  (2004).  Glaciers (2nd ed.).  Cambridge, England:  University Press.

Herring, David.  (2006).  Paleoclimatology:  the Oxygen Balance.  Retrieved November 11, 2006, from the National Aeronautics and Space Administration Web site:  http://earthobservatory.nasa.gov/Study/Paleoclimatology_OxygenBalance/oxygen_balance.html.

Kendall, C., Caldwell, E., & Snyder, D.  (2004).  Resources on Isotopes, Periodic Table:  Oxygen.  Retrieved November 11, 2006, from the United States Geologic Survey Web site:  http://wwwrcamnl.wr.usgs.gov/isoig/period/o_iig.html.

Kennett, James.  (1982).  Marine Geology.  Englewood Cliffs, NJ:  Prentice-Hall.

Radioactive Decay.  (2000).  Retrieved November 11, 2006, from the Boston University Web site:  http://www.physics.bu.edu/py106/notes/RadioactiveDecay.html.

Riebeek, Holli. (2006a). Paleoclimatology:  Explaining the Evidence.  Retrieved November 27, 2006, from the National Aeronautics and Space Administration Web site:  http://earthobservatory.nasa.gov/Study/Paleoclimatology_Evidence/paleoclimatology_evidence.html.

________.  (2006b).  Paleoclimatology:  The Ice Core Record.  Retrieved November 11, 2006, from the National Aeronautics and Space Administration Web site:  http://earthobservatory.nasa.gov/Study/Paleoclimatology_IceCores/.

Stauffer, Bernhard, et al.  (2006).  Greenland Ice Core Project:  An ESF Research Programme--Final Report.  Retrieved November 11, 2006, from the National Oceanic and Atmospheric Administration Web site:  http://www.ncdc.noaa.gov/paleo/icecore/greenland/summit/document.