Paleoclimate Proxy Converter Calculator

Oxygen has two stable isotopes, ¹⁶O and ¹⁸O, and the ratio between them in ancient carbonates, ice, or cave deposits shifts in predictable ways with temperature — because lighter ¹⁶O evaporates more readily than heavier ¹⁸O, that fractionation is temperature-dependent and gets locked into whatever mineral forms at the time. This calculator applies three different, real published calibrations depending on which proxy you select: marine carbonates use the Shackleton (1974) quadratic, T = 16.9 − 4.38×Δ + 0.10×Δ², where Δ is the difference between the sample's δ18O and contemporary seawater's δ18O; ice cores use the simpler Jouzel spatial slope, dividing the offset from a −35‰ reference value by 0.67 to get a temperature anomaly added to your modern reference temperature; and speleothems use a calibration derived from Kim & O'Neil's 1997 equilibrium calcite-water oxygen isotope fractionation study, published in Geochimica et Cosmochimica Acta, which has slightly different coefficients from the marine version because cave calcite forms under different conditions than ocean carbonate. The calculator also reports a δD (deuterium) value, a sea-level proxy from benthic foraminifera isotope shift, and a deuterium-excess value that hints at moisture-source conditions. δD comes from whichever source you give it: enter a Measured δD from your own mass-spec run and that value is used directly, or leave the field at 0 and δD is estimated from the Global Meteoric Water Line (δD ≈ 8×δ18O + 10, Craig 1961).

That distinction is the whole story for d-excess, which is defined as δD − 8×δ18O: run it against an estimated δD and the algebra collapses to the water line's own 10‰ intercept no matter what you enter, so a d-excess that actually varies — and actually says something about moisture source — requires an independently measured δD. It is also only meaningful for precipitation-derived samples like ice cores, not marine carbonate. The biggest pitfall is picking the wrong proxy type for your sample: the three calibrations produce meaningfully different temperatures from the same δ18O input, and mixing up marine and ice-core interpretations will give you a physically nonsensical result.