Similar Holocene glaciation histories in tropical South America and Africa

Tropical glaciers have retreated alongside warming temperatures over the past century, yet the way in which these trends fit into a long-term geological context is largely unclear. Here, we present reconstructions of Holocene glacier extents relative to today from the Quelccaya ice cap (Peru) and the Rwenzori Mountains (Uganda) based on measurements of in situ14C and 10Be from recently exposed bedrock. Ice-extent histories are similar at both sites and suggest that ice was generally smaller than today during the first half of the Holocene and larger than today for most, if not all, of the past several millennia. The similar glaciation history in South America and Africa suggests that large-scale warming followed by cooling of the tropics during the late Holocene primarily drove ice extent, rather than regional changes in precipitation. Our results also imply that recent tropical ice retreat is anomalous in a multimillennial context.

Thinning of the Quelccaya Ice Cap over the last thirty years

Direct measurements of the decadal response of Tropical glaciers to environmental changes are difficult to acquire within their accumulation zones. In 2013, we used dual-frequency kinematic GPS to re-measure the surface elevations at 46 sites, from the margin to across the summit of the Quelccaya Ice Cap, first measured in 1983 using terrestrial surveying methods. In 2015, six additional sites on the western margin, first observed in 1978, were remeasured. Over the past 30 years, the ice cap summit has thinned by 4.41 ± 0.23 m (2σ), with a maximum ice loss at one site near the margin of 63.4 ± 0.34 m (2σ) over 37 years. Using geophysical methods that located the sub-glacial bedrock, we estimate the unit-volume of ice in 1983 along a profile from the 1983 margin to the summit and then the change in volume from 1983 to 2013 by differencing the surface elevations. Over the past 30 years, 21.2 ± 0.3 % (2σ) of the ice unit-volume has been lost suggesting an average annual mass balance rate of −0.5 ± 0.1 m w.e. a−1 (2σ). Increasing air temperature at high elevations of the Andes is likely a major driver of the observed changes. Specifically, within the ablation zone, thinning is likely caused by a 1–2 m w.e. a−1 increase in melting and sublimation above steady-state.Within the accumulation zone, analysis of annual, dry-season summit pits suggests that surface lowering may be caused by both a slight decrease in net snow accumulation and an increase in firnification rate, though this interpretation yet lacks statistical significance. The role of ice flux changes since 1983/4 remains unconstrained, awaiting updated measurements of ice surface velocities across the ice cap.