Late-Holocene climate evolution at the WAIS Divide site, West Antarctica: bubble number-density estimates J.M. FEGYVERESI, 1 R.B. ALLEY, 1 M.K. SPENCER, 2 J.J. FITZPATRICK, 3 E.J. STEIG, 4 J.W.C. WHITE, 5 J.R. McCONNELL, 6 K.C. TAYLOR 6 1 Department of Geosciences, The Pennsylvania State University, and The Earth and Environmental Systems Institute, University Park, Pennsylvania 16802, USA E-mail: jmf439@psu.edu 2 Department of Geology and Physics, Lake Superior State University, Sault Ste. Marie, Michigan 49783, USA 3 US Geological Survey, Geology and Environmental Change Science Center, PO Box 25046, MS 980, Denver, Colorado 80225, USA 4 Department of Earth and Space Sciences and Quaternary Research Center, University of Washington, Seattle, Washington 98195, USA 5 Institute of Arctic and Alpine Research, UCB 450, University of Colorado at Boulder, Boulder, Colorado 80309-0450, USA 6 Desert Research Institute, 2215 Raggio Parkway, Reno, Nevada 89512-1095, USA ABSTRACT. A surface cooling of 1.78 8C occurred over the two millennia prior to 1700 CE at the West Antarctic ice sheet (WAIS) Divide site, based on trends in observed bubble number-density of samples from the WDC06A ice core, and on an independently constructed accumulation-rate history using annual-layer dating corrected for density variations and thinning from ice flow. Density increase and grain growth in polar firn are both controlled by temperature and accumulation rate, and the integrated effects are recorded in the number-density of bubbles as the firn changes to ice. Number- density is conserved in bubbly ice following pore close-off, allowing reconstruction of either paleotemperature or paleo-accumulation rate if the other is known. A quantitative late-Holocene paleoclimate reconstruction is presented for West Antarctica using data obtained from the WAIS Divide WDC06A ice core and a steady-state bubble number-density model. The resultant temperature history agrees closely with independent reconstructions based on stable-isotopic ratios of ice. The 1.78C cooling trend observed is consistent with a decrease in Antarctic summer duration from changing orbital obliquity, although it remains possible that elevation change at the site contributed part of the signal. Accumulation rate and temperature dropped together, broadly consistent with control by saturation vapor pressure. INTRODUCTION Knowledge of climate history is of value in understanding and attributing climate change (e.g. Solomon and others, 2007). Ice cores contribute in important ways, supplement- ing and extending instrumental data (e.g. Schneider and Steig, 2008; Steig and others, 2009). Data from ice cores are especially valuable because they provide multiple, inde- pendent constraints on the history of key climatic variables such as temperature (e.g. Cuffey and Paterson, 2010). Ice-core techniques for estimating past temperatures include the interpretation of stable-isotope ratios of ice (d 18 O or dD; e.g. Epstein and others, 1970), occurrence of melt layers (e.g. Das and Alley, 2008), thermal fractionation of gases under temperature gradients across firn, and depth or age of firn (e.g. Severinghaus and others, 1998), together with the inversion of temperatures measured in boreholes (e.g. Cuffey and others, 1994). All of these have a strong basis in physics but are influenced by processes other than simply temperature change. Spencer and others (2006) showed that bubble number- density in polar ice can be quantitatively modeled as a function of temperature and accumulation rate. Here we apply their model to new bubble number-density data obtained from an ice core at the West Antarctic ice sheet (WAIS) Divide site to produce a new paleoclimate reconstruction for the two millennia prior to 1700 CE. The WAIS Divide drilling site (Fig. 1) is located at 79828.058 0 S, 112805.189 0 W, 160 km from the previous Byrd ice-core drilling site and 24 km from the current West Antarctic ice-flow divide (on the Ross Sea side). This drilling location was chosen because it provides an excellent high- time-resolution analogue to the central Greenland ice cores, in terms of ice accumulation rate (22 cm a –1 ), thickness (3465 m), average annual surface temperature (–31.18C), and gas-age–ice-age difference (225 years) (Conway and others, unpublished). BUBBLE NUMBER-DENSITY PALEOCLIMATOLOGY The technique of Spencer and others (2006) allows reconstruction of either paleotemperatures or paleo-accu- mulation rates by fitting a semi-empirical steady-state model to measured number-densities of bubbles in (ice-core) glacier ice (see also Lipenkov and others, 1998; Alley and Fitzpatrick, 1999). Temperature and accumulation rate are the primary drivers of polar firn densification, and the integrated effects of these two drivers on density and grain growth are recorded in the number-density of bubbles formed during the sintering process as the firn is transformed into ice at the pore close-off depth (Spencer, 1999; Spencer and others, 2006). That number-density is then conserved in the bubbly ice following pore close-off. This allows use of a Journal of Glaciology, Vol. 57, No. 204, 2011 629