Glaciers in the high mountains of Asia are a crucial water resource, but are at risk from global warming. Modelling suggests that the glaciers will shed mass in direct proportion to the warming to which they are exposed. See Letter p.257
The 2015 Paris climate agreement aspires to limit the average global temperature rise to 1.5 °C above pre-industrial levels by the end of this century (see go.nature.com/2efxikg). Intensive work is under way to determine what return we would get for the formidable effort required to achieve this goal. On page 257, Kraaijenbrink et al.1 contribute to this work by modelling glaciers in the high mountains of Asia (HMA) and showing how these glaciers respond to different levels of warming. The authors demonstrate that even if the 1.5 °C target is met, about one-third of the present-day mass of HMA glaciers will be lost by the end of this century. The work could have major consequences for communities in the region that depend on glacial meltwater for domestic use, hydropower and irrigation.
In 2013, the Intergovernmental Panel on Climate Change reported climate projections2 based on four scenarios called Representative Concentration Pathways (RCPs) that differ in terms of the amounts of greenhouse gases in the atmosphere. For example, RCP2.6 assumes that greenhouse-gas emissions peak during 2010–2020 and subsequently decline, whereas RCP8.5 assumes that emissions continue to rise steeply throughout the twenty-first century. Kraaijenbrink and colleagues considered 110 climate simulations obtained from the Coupled Model Intercomparison Project3, each based on one of the four RCPs. Of the 110 simulations, the authors identified 6 (a subset of the RCP2.6 simulations) that yielded an average global temperature rise of 1.4–1.6 °C by the end of this century (2071–2100), compared with pre-industrial times (1851–.80). When projecting climate, such 30-year averages are used to present estimates. Scenarios of warming are constructed with respect to pre-industrial times, but because of lack of information it is necessary to begin glaciological simulations at the present day. In these six simulations, HMA glaciers suffered more warming (2.1 °C) than the global average (1.5 °C) and there was substantial variation between subregions — from warming of 1.9 °C in the Eastern Himalaya to warming of 2.3 °C in the Hindu Kush mountain range.
Considerable parts of HMA glaciers are covered by a layer of debris, caused by the erosion of material surrounding the glacier. Depending on its thickness, this debris can either suppress or accelerate glacial melt. Meltwater ponds on glacial surfaces can also enhance mass loss (Fig. 1). Kraaijenbrink et al. therefore developed a glacier model that incorporates the effects of both debris and meltwater ponds. They then used satellite imagery to measure the extent of these features on more than 33,000 HMA glaciers — those with an area larger than 0.4 square kilometres — accounting for almost all of the ice in the region.
Next, the authors calculated the present-day mass of the glaciers using a model4 called GlabTop2. They estimated that there are almost 5 × 1012 tonnes of ice currently stored in HMA glaciers — substantially less than the amount suggested by some previous estimates5. Finally, the authors combined their glacier model with projected temperature and precipitation changes from the climate simulations to determine the glaciers' long-term evolution.
Kraaijenbrink et al. show that about two-thirds of the ice currently stored in HMA glaciers will be lost by 2071–2100 if no efforts are made to prevent climate change. This corresponds to RCP8.5, and is the track on which we are currently travelling, despite signs of more-invigorated policy-making relating to climate. By contrast, the authors find that only about one-third of the ice will be lost if the 1.5 °C target is achieved. Substantial differences in precipitation changes across the 110 climate simulations have no marked impact on their projections.
One of the authors' most striking findings is that the mass loss is directly proportional to the warming to which the glaciers are exposed (Fig. 2). In particular, their results suggest that about 7% of HMA ice would be retained for every 1 °C of averted warming. If there was no warming between 1851–1880 and 2071–2100, about 22% of the present-day ice mass would be lost over this time span. In fact, in one of their own simulations (which was not part of the 110 simulations), in which there is no warming after the present day, the authors find a mass loss of 14% by 2071–2100 — implying that about 8% of the present ice mass was lost between 1851–1880 and today. Mass loss in the absence of warming reflects a slow adjustment to an earlier climate because it can take decades for glaciers to reach the size sustained by a given climate. More speculatively, the work indicates that a warming of about 11 °C would be required to remove all of the ice. Such a value is much higher than any of those in the 110 climate simulations, suggesting that HMA glaciers are not going to disappear altogether by 2100, let alone by 2035 (as has previously been suggested6).
Kraaijenbrink and colleagues' glacier model has some innovative features that might raise eyebrows among glaciologists, but it is difficult to find fault with it as a pioneering effort. Nevertheless, more work is needed to validate the model because most of the required measurements have not yet been made. The authors are operating in 'reconnaissance mode' and a long list of questions remains for the future — from whether all the climate simulations deserve equal standing, to how best to validate the glacier model, to the practical implications of subregional variations in glaciers' response to warming. The authors have shown that achieving the 1.5 °C target will conserve a substantial fraction of Asia's water resources and that, if we fail in this regard, we will pay in direct proportion to the extent of the failure.