Although quantitative isotopic data from speleothems has been used to evaluate isotope-enabled model simulations in the past, currently no consensus exists regarding the most appropriate methodology through which achieve this. Motivated by the fact that a number of modelling groups will be running isotope-enabled palaeoclimate simulations in the framework of the Coupled Model Intercomparison Project Phase 6, we have explored different approaches on how to use speleothem data for data-model comparisons in a paper that is now open for discussion at Climate of the Past Discussions (

Comas-Bru, L., Harrison, S. P., Werner, M., Rehfeld, K., Scroxton, N., Veiga-Pires, C., and SISAL working group members: Evaluating model outputs using integrated global speleothem records of climate change since the last glacial, Clim. Past Discuss., 2019, 1-53, 10.5194/cp-2019-25, 2019.

In this study, we show that the SISAL database reproduces the first-order spatial patterns of isotopic variability and is therefore useful to assess a model’s ability to simulate spatial isotopic trends. Our analyses provide a protocol for using speleothem isotopic data for model evaluation, including screening the observations, the optimum period for the modern observational baseline, and the selection of an appropriate time-window for creating means of the isotope data for palaeo time slices.

Figure 1 shows an example on how to circumvent the issue of speleothem records being discontinuous in nature, which complicates procuring large numbers of records if data-model comparisons are made using the traditional approach of comparing anomalies between a control period and a given palaeoclimate experiment. This example illustrates a technique through which the absolute isotopic values during any time period could be used for model evaluation.

Figure 1. (a) Mid-Holocene (MH) transect for the region NW to SE across North America, and (b) Last Glacial Maximum (LGM) transect for N-S from central Europe to southern Africa. Maps at the top of each panel show the simulated δ18Op (left), Mean Annual Temperature (MAT; centre) and Mean Annual Precipitation (MAP; right) from ECHAM5-wiso. The same scale is used for the δ18O, MAT and MAP maps. All transects show absolute δ18O values. In the δ18O maps, filled circles are SISAL δ18O averages for entities that cover both the MH and the modern reference period. Filled squares are SISAL entities that do not have a corresponding modern. Bottom plots of each panel show the simulated data extracted for each transect: black circles and whiskers are mean ±1 standard deviation of the data extracted along longitudinal sections in between the two great circle lines shown in solid grey lines in the top maps. The red line is the median of the extracted data. All data were extracted at steps of 1.12⁰ to coincide with the average model grid-size. Bottom plots in each panel also show SISAL δ18O (circles for low-elevation sites, < 1,000 masl; triangles for high-elevation sites, > 1,000 masl), pollen-based quantitative reconstructions of MAT (red squares; Bartlein et al., 2011) and MAP (blue squares; Bartlein et al., 2011). Pollen-based reconstructions have been converted to absolute values by adding the CRU-TS4.01 climatology (Harris et al., 2014).

Bartlein, P. J., Harrison, S. P., Brewer, S., Connor, S., Davis, B. A. S., Gajewski, K., Guiot, J., Harrison-Prentice, T. I., Henderson, A., Peyron, O., Prentice, I. C., Scholze, M., Seppä, H., Shuman, B., Sugita, S., Thompson, R. S., Viau, A. E., Williams, J., and Wu, H.: Pollen-based continental climate reconstructions at 6 and 21 ka: a global synthesis, Climate Dynamics, 37, 775-802,, 2011.

Harris, I., Jones, P. D., Osborn, T. J., and Lister, D. H.: Updated high-resolution grids of monthly climatic observations – the CRU TS3.10 Dataset, International Journal of Climatology, 34, 623- 642,, retrieved from on November 2018, 2014.