Limits of oxygen isotope palaeoaltimetry in Tibet

Measurements of stable water isotopes (oxygen and hydrogen) are commonly used to estimate palaeoelevation and quantify past changes in surface height across Tibet. Isotope palaeoaltimetry is often based on simple Rayleigh fractionation of a "parcel of air", but must make a considerable number of approximations and assumptions. In this paper, we elaborate on the practicability of oxygen water isotopes in palaeoaltimetry, and evaluate a recent challenge to the palaeoaltimetry community. First, we examine the isotopic composition of oxygen (δ18O) versus altitude relationship in a set of five topographic realisations of Tibet using an isotope-enabled palaeoclimate model for the mid-Eocene, a period where a variety of topographic 'uplift' models have been proposed, and compare it to modern relationships. Second, we investigate whether isotopic composition is a good predictor of more modest changes in topography, such as the introduction of a valley system or uplift of only part of the Tibetan region. The aim of the paper is not to perform a direct comparison to data, but to use the model to further refine knowledge of the strengths and limitations of using oxygen isotopes in palaeoaltimetry. We find that oxygen isotope palaeoaltimetry works surprisingly well, with the exception that it could not identify low elevation valley systems bounded by high elevations because the isotopic composition of the water in the air becomes depleted at the first high elevation that an air parcel passes over and does not recover when it descends into the valley. Hence, isotope-based elevations are biased towards mountain range peaks. Overall, the application of oxygen isotope palaeoaltimetry does have value, but would be further strengthened when employed together with isotope-enabled models. In conjunction with other techniques such as terrestrial thermal lapse rates and energy conservation approaches, over a wide spatial region, a more accurate and fully three-dimension view of complex palaeo-topography is increasingly possible, which will in turn improve the precision of these palaeoaltimeters.