Skip to content

Paleoceanographic Proxies and Paleo-SST Methods

CSIR NET Earth Sciences Part C: foram oxygen isotopes and ice volume, Mg/Ca, alkenone, TEX86 and faunal SSTs, productivity, circulation and CO2 proxies.

  • 19 min read
  • Updated
On this page · 18 sections
  1. Prerequisites
  2. The Hook: A Thermometer That Measured Ice
  3. What a Proxy Is
  4. Stable Isotopes: Notation
  5. Oxygen Isotopes in Foraminifera
  6. Paleo-SST Methods
  7. Productivity and Nutrient Proxies
  8. Circulation and Ventilation Proxies
  9. Carbonate System and CO₂ Proxies
  10. Weathering and Provenance Proxies
  11. Dating Marine Records
  12. Memory Tricks
  13. Don’t Confuse
  14. Common Mistakes
  15. Practice
  16. Quick Revision Table
  17. Key Takeaways
  18. What’s Next

Prerequisites

The Hook: A Thermometer That Measured Ice

What a Proxy Is

A proxy is a measurable property of an archive that varies in a known way with an environmental variable we cannot measure directly. Using one involves three steps:

  1. Calibration: relate the proxy to the variable in modern samples (core tops, sediment traps, cultures).
  2. Assumptions: that the relationship held in the past and that nothing else changed the signal.
  3. Age model: put the measurements on a time scale.
ArchiveTypical resolutionSpan
Deep-sea sediment corescenturies to millennia (decades in fast-accumulating or laminated cores)whole Cenozoic and beyond
Coralsseasonal to annualcenturies per colony; fossil reefs for older windows
Speleothemsannual to decadalhundreds of kyr; U–Th dated
Ice coresannual to decadalabout 800 kyr (Antarctica)
Varves, tree ringsannualthousands of years

Marine microfossils are the workhorses: planktic and benthic foraminifera (calcite tests; isotopes and trace elements), coccolithophores (calcite plates and alkenones), diatoms and radiolarians (opal; assemblages, productivity), dinoflagellate cysts (organic walls; survive where carbonate dissolves), pteropods (aragonite) and ostracods.

Stable Isotopes: Notation

$$\boxed{\delta = \left(\frac{R_{sample}}{R_{standard}} - 1\right)\times 1000\ \text{‰}}$$

with \( R = {}^{18}\text{O}/{}^{16}\text{O} \) or \( {}^{13}\text{C}/{}^{12}\text{C} \). Carbonates are reported against VPDB, waters against VSMOW. Positive \( \delta \) means “heavier” (more of the rare isotope).

Fractionation during calcite precipitation is temperature-dependent: the colder the water, the more \(^{18}\)O the calcite takes up, by about 0.2–0.25‰ per °C.

Oxygen Isotopes in Foraminifera

Three things set foram δ¹⁸O

$$\delta^{18}\text{O}_{calcite} = f(\text{temperature}) + \delta^{18}\text{O}_{seawater}$$

and sea-water \( \delta^{18}\text{O} \) itself has two parts:

ControlDirectionSize
Temperaturecolder → heavier calciteabout 0.2–0.25‰ per °C
Global ice volumemore ice → heavier oceanabout 1.0‰ for the Last Glacial Maximum, when sea level was about 120–130 m lower
Local salinity / water balanceevaporation → heavier; rain and river water → lighterregional slope, often about 0.2–0.5‰ per salinity unit

Ice sheets are made of snow that has lost \(^{18}\)O at every step from the evaporating ocean (Rayleigh distillation); polar ice has \( \delta^{18}\text{O} \) of −30 to −50‰. Locking that water up leaves the ocean enriched.

The paleotemperature equation

Calibrations derived from inorganic and foram calcite (Epstein and Urey’s group in the early 1950s, refined later) have the quadratic form

$$\boxed{T\ (°\text{C}) = 16.9 - 4.38(\delta_c - \delta_w) + 0.10(\delta_c - \delta_w)^2}$$

in the version usually attributed to Shackleton (1974), where \( \delta_c \) is the calcite \( \delta^{18}\text{O} \) (VPDB) and \( \delta_w \) is sea-water \( \delta^{18}\text{O} \) expressed on the VPDB scale (about 0.27‰ below its VSMOW value). The slope near ordinary temperatures is about 4.2–4.4 °C per ‰.

Complications

EffectWhat it does
Vital effectsspecies-specific offsets from equilibrium (photosymbionts, growth rate); use one species, or apply species offsets
Carbonate-ion effecthigher \( [\text{CO}_3^{2-}] \) lowers foram \( \delta^{18}\text{O} \) and \( \delta^{13}\text{C} \)
Habitat and seasonspecies live at different depths and seasons; a “SST” may really be thermocline or seasonal temperature
Dissolutionremoves thin, warm-water shells first; biases toward heavier values
Diagenesisrecrystallisation of old (Paleogene and older) planktic tests on the cold sea floor makes them too heavy, so early Cenozoic tropical SSTs were long underestimated
Bioturbationsmooths and shifts short events

Marine isotope stages

Glacial cycles in benthic \( \delta^{18}\text{O} \) are numbered as marine isotope stages (MIS), counting back from the present: odd = warm (interglacial or interstadial), even = cold.

StageAge (approx.)Note
MIS 1last about 14 kaHolocene (with the deglaciation)
MIS 2about 29–14 kaincludes the Last Glacial Maximum (about 21 ka)
MIS 5eabout 130–115 kaLast Interglacial (Eemian); warmer than today, sea level higher
MIS 11about 400 kalong interglacial

The LR04 stack (Lisiecki and Raymo, 2005) averages 57 benthic records into a global reference curve, used to date cores by correlation. Its main features: 41 kyr cycles before about 1.2 Ma, and dominant, saw-tooth 100 kyr cycles after the Mid-Pleistocene Transition (about 1.2–0.7 Ma).

The Cenozoic benthic δ¹⁸O curve

Compilations of benthic \( \delta^{18}\text{O} \) (Zachos and colleagues, 2001 and later) show the long cooling of the deep ocean since the early Eocene, with steps:

EventAge (approx.)Signal
Paleocene–Eocene Thermal Maximum (PETM)about 56 Masudden negative excursions in \( \delta^{18}\text{O} \) (warming) and \( \delta^{13}\text{C} \) (light carbon release); CCD shoaling
Early Eocene Climatic Optimumabout 53–50 Mawarmest deep water of the Cenozoic
Eocene–Oligocene transition (Oi-1)about 34 Ma+1‰ step in two pulses: first large Antarctic ice sheet and deep-water cooling; CCD deepened
Mid-Miocene Climatic Optimumabout 17–15 Mawarm interval
Middle Miocene climate transitionabout 14 MaAntarctic ice expanded permanently
Intensification of Northern Hemisphere glaciationabout 2.7 Maice-rafted debris in the North Atlantic

Paleo-SST Methods

Overview

MethodCarrierPrincipleStrengthsMain pitfalls
Foram δ¹⁸Ocalcite of planktic foramstemperature-dependent fractionationlong, widely availablemixes T, ice volume and salinity
Mg/Caplanktic foram calciteMg uptake rises about 9–10% per °Cindependent of \( \delta_w \); paired with δ¹⁸O on the same shellsdissolution lowers Mg/Ca; salinity and pH effects; sea-water Mg/Ca changes over Myr; cleaning protocol
Alkenone \( U^{K'}_{37} \)C\(_{37}\) ketones of haptophyte algae (Emiliania huxleyi, Gephyrocapsa)fewer double bonds in warmer waterrobust to dissolution; good for 0–28 °Csaturates near 28–29 °C; seasonality; lateral transport of fine organic particles
TEX\(_{86}\)membrane lipids (isoprenoid GDGTs) of marine Thaumarchaeotamore rings in warmer waterworks above 28 °C and back to the Cretaceoussubsurface signal; terrestrial input (check BIT index); calibration choice matters
Faunal transfer functions / modern analogue techniqueassemblages of forams, radiolarians, diatoms, coccoliths, dinocystsspecies distribution tracks temperaturebasis of CLIMAP (1976) ice-age mapno-analogue assemblages; dissolution; evolution limits it to the late Quaternary
Clumped isotopes (Δ\(_{47}\))carbonateabundance of \(^{13}\)C–\(^{18}\)O bonds depends only on temperatureindependent of \( \delta_w \)large samples, low precision; diagenetic resetting
Coral Sr/Ca and δ¹⁸Oaragonite of reef coralsSr/Ca falls with rising Tseasonal resolutionvital effects; limited to reef settings and coral lifetimes

Mg/Ca

$$\boxed{\text{Mg/Ca} = B\,e^{AT}} \qquad \text{multi-species calibration (Anand and colleagues, 2003): } B = 0.38,\ A = 0.09$$

with Mg/Ca in mmol mol\(^{-1}\). An exponential means a constant percentage change per degree: \( e^{0.09} - 1 \approx 9.4\% \) per °C.

Alkenones

$$U^{K'}_{37} = \frac{[\text{C}_{37:2}]}{[\text{C}_{37:2}] + [\text{C}_{37:3}]}$$

where \( \text{C}_{37:2} \) and \( \text{C}_{37:3} \) are the di- and tri-unsaturated C\(_{37}\) methyl ketones. A widely used global core-top calibration (Müller and colleagues, 1998) is

$$\boxed{U^{K'}_{37} = 0.033\,T + 0.044}$$

TEX₈₆

$$\text{TEX}_{86} = \frac{\text{GDGT-2} + \text{GDGT-3} + \text{Cren}'}{\text{GDGT-1} + \text{GDGT-2} + \text{GDGT-3} + \text{Cren}'}$$

(numbers give the cyclopentane rings; Cren\(^\prime\) is the crenarchaeol isomer). One widely used calibration, TEX\(^{H}_{86}\) (Kim and colleagues, 2010), is \( T = 68.4\log_{10}(\text{TEX}_{86}) + 38.6 \). TEX\(_{86}\) = 0.70 gives about 28 °C. The BIT index (branched versus isoprenoid GDGTs) flags samples where soil-derived lipids could bias the result.

Combining proxies: sea-water δ¹⁸O and salinity

Measure Mg/Ca and \( \delta^{18}\text{O} \) on the same shells. Mg/Ca gives \( T \); put \( T \) into the paleotemperature equation to get \( \delta_w \); subtract the ice-volume contribution; convert the remainder to a local salinity change using a regional \( \delta_w \)–salinity slope. This is how monsoon-driven freshening of the Bay of Bengal and the Andaman Sea has been reconstructed (see the next page).

The uncertainties add up: about ±1 °C in Mg/Ca temperature alone gives about ±0.2‰ in \( \delta_w \), which is close to one salinity unit for a 0.25‰ slope.

Productivity and Nutrient Proxies

ProxyLogic
Organic carbon (TOC) and its MARmore export → more burial (but preservation also depends on oxygen and sedimentation rate)
Biogenic opal and CaCO\(_3\) MARexport of diatoms and coccoliths or forams
Excess (biogenic) Babarite forms in decaying organic aggregates; robust under oxic conditions
Upwelling indicator speciesplanktic foram Globigerina bulloides thrives in cool, nutrient-rich upwelled water
\( \delta^{15}\text{N} \) of organic matterhigh values when water-column denitrification removes light nitrate (Arabian Sea OMZ); also nutrient utilisation
Benthic foram assemblages and accumulation ratesflux of food to the sea floor; low-oxygen species
Redox-sensitive metals (U, Mo, Re) and laminationslow bottom-water oxygen; laminations mean no burrowing animals

Circulation and Ventilation Proxies

ProxyPrincipleTypical use
Benthic \( \delta^{13}\text{C} \) (Cibicidoides, which records bottom-water DIC)respiration adds light carbon as water ages; NADW has high \( \delta^{13}\text{C} \) (around +1‰), old Pacific deep water near 0‰glacial NADW shoaled; southern-source water filled the deep Atlantic
Cd/Ca of benthic foramsCd follows phosphatenutrient content of deep water; independent check on \( \delta^{13}\text{C} \)
Benthic–planktic \(^{14}\)C age differencetime since a water parcel left the surfaceventilation age of deep water
Nd isotopes (εNd) in authigenic coatings and fish teethset by continental sources around the basin, not altered by biology; NADW strongly negative (about −13 to −14), Pacific much less negativewater-mass provenance and mixing
\(^{231}\)Pa/\(^{230}\)Thboth made from U decay in sea water at a ratio of 0.093; Th is scavenged within decades, Pa lingers for a century or more and can be exported by fast deep flowratios well below 0.093 in North Atlantic sediment mean vigorous export; ratios near 0.093 mean weak overturning (Heinrich Stadial 1)
Sortable silt (mean size of 10–63 μm terrigenous silt)near-bottom current speedcontourite drifts, deep western boundary currents

Carbonate System and CO₂ Proxies

ProxyRecords
\( \delta^{11}\text{B} \) of foram calcitesea-water pH (borate fraction depends on pH, and borate is isotopically lighter); with a second carbonate parameter gives pCO\(_2\)
B/Cacarbonate ion
Alkenone carbon-isotope fractionation (\( \varepsilon_p \))dissolved CO\(_2\) (with growth-rate assumptions)
Foram shell weight; fragmentation; benthic B/Cadeep-water \( [\text{CO}_3^{2-}] \), dissolution
Ice-core air bubblesatmospheric CO\(_2\) directly: about 180 ppm in glacials and 280 ppm in interglacials over the last 800 kyr

Weathering and Provenance Proxies

  • Sea-water \(^{87}\)Sr/\(^{86}\)Sr, recorded in marine carbonates, rose from about 0.7078 in the late Eocene to about 0.7092 today. Rivers draining old continental crust (high ratio) compete with hydrothermal and basalt weathering inputs (low ratio). The rise has been linked to Himalayan uplift and weathering (Raymo and Ruddiman’s hypothesis of the early 1990s), which would also draw down CO\(_2\).
  • Os and Li isotopes of sea water: further weathering tracers.
  • Detrital \(^{87}\)Sr/\(^{86}\)Sr and εNd: distinguish Himalayan from Indian peninsular (Deccan) sources in Bay of Bengal and Arabian Sea sediment.
  • Clay mineralogy and major-element ratios (K/Al, chemical index of alteration): intensity of chemical weathering on land.

Dating Marine Records

  • Radiocarbon: conventional age \( t = -8033\ln F \), with \( F \) the fraction modern (8033 yr is the Libby mean life). Subtract a marine reservoir age (about 400 yr on average for surface water, with a regional correction \( \Delta R \)), then calibrate to calendar years.
  • Correlation to LR04, magnetostratigraphy, biostratigraphy, tephra layers, and orbital tuning for older records. More in Oceanic Sediments, Diagenesis and Marine Mineral Resources.

Memory Tricks

Don’t Confuse

ThisvsThat
\( \delta^{18}\text{O}_{calcite} \) (T + \( \delta_w \))\( \delta^{18}\text{O}_{seawater} \) (ice volume + local salinity)
VPDB (carbonate standard)VSMOW (water standard)
Planktic forams (surface or thermocline)Benthic forams (bottom water)
Mg/Ca (exponential in T)\( U^{K'}_{37} \) (linear in T)
Alkenones (haptophyte algae, photic zone)TEX\(_{86}\) (archaea, often subsurface)
\( \delta^{13}\text{C} \) (nutrient and ageing tracer, biology alters it)εNd (quasi-conservative provenance tracer)
Lysocline-driven dissolution (lowers Mg/Ca, raises \( \delta^{18}\text{O} \))True cooling
Conventional \(^{14}\)C ageCalibrated calendar age

Common Mistakes

Practice

All questions below are original practice questions written for this page. They are not from any CSIR NET paper.

Level 1: Foundation

  1. What is the standard for reporting the \( \delta^{18}\text{O} \) of carbonates?
  2. Is MIS 6 a glacial or an interglacial stage?
  3. Which organisms produce the alkenones used in \( U^{K'}_{37} \)?
Answers
  1. VPDB (Vienna Pee Dee Belemnite).
  2. Glacial (even numbers are cold).
  3. Haptophyte algae such as the coccolithophores Emiliania huxleyi and Gephyrocapsa oceanica.

Level 2: Part B Style

Practice Q1. Using Mg/Ca = 0.38 e\(^{0.09T}\), a planktic foram Mg/Ca of 3.5 mmol mol\(^{-1}\) gives about:

(a) 21.3 °C (b) 24.7 °C (c) 26.7 °C (d) 38.9 °C

Answer: (b) 24.7 °C

\( T = \ln(3.5/0.38)/0.09 = \ln 9.21/0.09 = 2.220/0.09 \approx 24.7 \) °C. (d) treats the relation as linear (3.5/0.09).

Practice Q2. An alkenone \( U^{K'}_{37} \) value of 0.80 with \( U^{K'}_{37} = 0.033T + 0.044 \) corresponds to:

(a) 22.9 °C (b) 24.2 °C (c) 25.9 °C (d) 26.4 °C

Answer: (a) 22.9 °C

\( (0.80 - 0.044)/0.033 = 0.756/0.033 \approx 22.9 \) °C. (b) ignores the intercept (0.80/0.033).

Practice Q3. Benthic foram \( \delta^{18}\text{O} \) increases by 1.7‰ from an interglacial to a glacial. Taking the ice-volume effect as 1.0‰ and 0.25‰ per °C, the deep-water cooling is about:

(a) 1.0 °C (b) 2.8 °C (c) 4.0 °C (d) 6.8 °C

Answer: (b) 2.8 °C

Temperature part \( = 1.7 - 1.0 = 0.7 \)‰; \( 0.7/0.25 = 2.8 \) °C. (d) ignores ice volume (1.7/0.25).

Practice Q4. Which proxy is least affected by changes in sea-water \( \delta^{18}\text{O} \)?

(a) planktic foram \( \delta^{18}\text{O} \) (b) benthic foram \( \delta^{18}\text{O} \) (c) clumped isotopes (Δ\(_{47}\)) (d) coral \( \delta^{18}\text{O} \)

Answer: (c)

Δ\(_{47}\) depends on the ordering of heavy isotopes within the carbonate itself and is independent of the water’s isotopic composition. All the others depend on \( \delta_w \).

Level 3: Part C Style

Practice Q5. Paired measurements on Globigerinoides ruber from a Bay of Bengal core:

Mg/Ca (mmol mol\(^{-1}\))\( \delta^{18}\text{O}_c \) (‰ VPDB)
Core top4.2−2.4
LGM3.2−0.6

Use Mg/Ca = 0.38 e\(^{0.09T}\), \( T = 16.9 - 4.38(\delta_c - \delta_w) + 0.10(\delta_c - \delta_w)^2 \) (\( \delta_w \) on the VPDB scale), an LGM ice-volume effect of +1.0‰ and a local slope of 0.25‰ per salinity unit. Which statements are correct?

  1. The LGM was about 3 °C cooler than the core top.
  2. Local \( \delta_w \) (after removing the ice-volume effect) was about 0.16‰ higher at the LGM.
  3. The LGM surface water was about 0.6–0.7 salinity units saltier, consistent with a weaker monsoon.
  4. Without the ice-volume correction, the inferred salinity change would be about four times smaller.

(a) 1 and 2 only (b) 1, 2 and 3 only (c) 2, 3 and 4 only (d) 1, 2, 3 and 4

Answer: (b) 1, 2 and 3 only
  • Temperatures: core top \( \ln(4.2/0.38)/0.09 = 26.7 \) °C; LGM \( \ln(3.2/0.38)/0.09 = 23.7 \) °C. Cooling 3.0 °C. Statement 1 true.
  • Solve the quadratic for \( \delta_c - \delta_w \): at 26.7 °C it is −2.13‰; at 23.7 °C it is −1.50‰.
  • \( \delta_w \) (VPDB): core top \( -2.4 + 2.13 = -0.27 \)‰; LGM \( -0.6 + 1.50 = +0.90 \)‰. Change = +1.16‰.
  • Remove ice volume: \( 1.16 - 1.0 = +0.16 \)‰. Statement 2 true.
  • Salinity: \( 0.16/0.25 \approx +0.65 \). Saltier, as expected with less monsoon runoff and rain. Statement 3 true.
  • Statement 4 false: without the correction the change would be \( 1.16/0.25 \approx 4.6 \) units, about seven times larger, not smaller.

Practice Q6. A North Atlantic core shows, across a Heinrich layer: benthic \( \delta^{13}\text{C} \) falling from +1.0‰ to +0.2‰, sediment \(^{231}\)Pa/\(^{230}\)Th rising from 0.055 to 0.090, and a layer rich in coarse ice-rafted grains. Which interpretations are correct?

  1. Nutrient-rich, southern-sourced or poorly ventilated water replaced NADW at the site.
  2. Pa export from the basin decreased, consistent with a weaker AMOC.
  3. The IRD layer records iceberg discharge.
  4. The \( \delta^{13}\text{C} \) drop by itself proves that surface productivity rose.

(a) 1, 2 and 3 only (b) 1 and 4 only (c) 2 and 3 only (d) 1, 2, 3 and 4

Answer: (a) 1, 2 and 3 only

Low benthic \( \delta^{13}\text{C} \) means bottom water with more respired carbon: reduced NADW influence (1). Pa/Th approaching the production ratio 0.093 means Pa was no longer exported, i.e. weak overturning (2). Heinrich layers are IRD from iceberg armadas (3). Statement 4 is false: a benthic \( \delta^{13}\text{C} \) drop has several possible causes (water-mass change, ageing, a whole-ocean carbon shift); it does not by itself prove a productivity rise.

Practice Q7. Consider these statements about paleo-SST methods:

  1. Partial dissolution of planktic foram tests on the sea floor tends to bias Mg/Ca temperatures low.
  2. \( U^{K'}_{37} \) loses sensitivity above about 28 °C.
  3. Recrystallised Eocene planktic forams tend to give SSTs that are too warm.
  4. A high BIT index warns that terrestrial GDGTs may bias TEX\(_{86}\).

(a) 1, 2 and 4 only (b) 2 and 3 only (c) 1 and 3 only (d) 1, 2, 3 and 4

Answer: (a) 1, 2 and 4 only

1: Mg-rich calcite dissolves preferentially, lowering Mg/Ca and apparent temperature. 2: the index approaches 1. 4: BIT measures soil-derived branched GDGTs. 3 is false: recrystallisation on the cold sea floor adds heavy \( \delta^{18}\text{O} \), so it makes tropical SSTs look too cold, the “cool tropics paradox” of early studies.

Quick Revision Table

IdeaFact or formula
δ notation\( \delta = (R_{sa}/R_{std} - 1)\times 1000 \); VPDB carbonate, VSMOW water
Paleotemperature\( T = 16.9 - 4.38(\delta_c - \delta_w) + 0.10(\delta_c - \delta_w)^2 \); about 0.23‰ per °C
Ice volumeabout +1.0‰ at the LGM; sea level about 120–130 m lower
HistoryEmiliani 1955; Shackleton 1967 (ice volume); Hays, Imbrie, Shackleton 1976 (orbital cycles)
MISodd warm, even cold; MIS 2 LGM, MIS 5e Last Interglacial; LR04 stack (57 records)
Cenozoic stepsPETM about 56 Ma; Oi-1 about 34 Ma; MMCT about 14 Ma; NHG about 2.7 Ma; MPT about 1.2–0.7 Ma
Mg/Ca\( 0.38e^{0.09T} \); about 9% per °C; dissolution lowers it
Alkenones\( U^{K'}_{37} = \text{C}_{37:2}/(\text{C}_{37:2} + \text{C}_{37:3}) = 0.033T + 0.044 \); saturates near 28 °C
TEX\(_{86}\)archaeal GDGTs; TEX\(^{H}_{86}\): \( T = 68.4\log\text{TEX}_{86} + 38.6 \); check BIT
Faunal methodstransfer functions, MAT; CLIMAP 1976
Clumped isotopesΔ\(_{47}\); independent of \( \delta_w \)
ProductivityTOC, opal, excess Ba, G. bulloides, \( \delta^{15}\text{N} \), redox metals
Circulationbenthic \( \delta^{13}\text{C} \), Cd/Ca, B–P \(^{14}\)C, εNd, Pa/Th (production 0.093), sortable silt
CO\(_2\)\( \delta^{11}\text{B} \) → pH; ice cores 180–280 ppm
Weatheringsea-water \(^{87}\)Sr/\(^{86}\)Sr about 0.7078 → 0.7092 since late Eocene; clays; detrital εNd
\(^{14}\)C\( t = -8033\ln F \); reservoir about 400 yr

Key Takeaways

What’s Next