Twenty-five ice cores trace West Antarctic snowfall swings to tropical Rossby waves

Surface mass balance, the net gain or loss of ice at an ice sheet's surface, comes mainly from snowfall and from wind moving snow around ^[9]. A single core records one site's share of both. The Utah team set 25 cores ^[6] against historical climate datasets covering 1900 to 1999 ^[9]. Co-author Summer Rupper, a University of Utah geography professor, said the combination is what is new. "What's unique about this study is leveraging all these cores together to look more carefully at the drivers of the variability, not just the trends. We tend to focus on the trends, but variability matters a lot for ice sheet stability," she said ^[7].

The cores are shallow by design. Rupper's lab used a portable drill that reaches the upper 50 metres of the sheet, snow that fell in the last century or two ^[10]. She recovered some of the 5-centimetre-wide cores herself on field trips in 2010 and 2011 ^[10]. Fifty metres is about 5 percent of the sheet's average thickness of 1,040 metres ^[1]. For recent weather, the alternative would be instruments, and they do not last there. "You can't keep weather stations maintained. They just get buried by snow. They get blown over," Rupper said ^[11]. She also named the limit: "Ice cores can't record the same things that a weather station would, but we have proxies." ^[12]

The team reports that conditions in the tropics, especially in the Pacific and Indian oceans, affect how much snow falls in West Antarctica ^[2]. The evidence is observational: core layers set against 20th-century weather data ^[2]. A physical pathway is what makes a causal reading reasonable. According to first author Ella Hunter, a graduate student in atmospheric sciences, El Nino and the Madden-Julian Oscillation send Rossby waves south toward Antarctica ^[3]. Those waves shift winds and storm tracks, changing how much moisture reaches the ice ^[3]. "Tropical convection or heating in the tropics from sea surface temperatures can create storms that propagate these Rossby waves," Hunter said. "And they are really effective at transporting heat and moisture down to the Antarctic ice sheet." ^[14]

The second result is that West Antarctica does not act as one system. The team identified an eastern and a western region that respond differently to climate influences ^[13]. The news account of the study does not describe how the two regions differ, or give the share of year-to-year snowfall variation that the tropical patterns explain.

The ice sheet covers 2 million square kilometres, and if it all melted, global sea level would rise by more than 4.6 metres ^[8]. I think the evidence supports carrying tropical variability into estimates of the sheet's mass, on two conditions. The tropical share has to be large next to the long-term trend. The relationship also has to hold outside the window it was measured in, a century that ends in 1999 ^[9]. A 20th-century record cannot show whether the link behaves the same way in a warmer climate.