Rivers originating in the high mountains of Asia are among the most meltwater-dependent river systems on Earth, yet large human populations depend on their resources downstream 1 . Across High Asia's river basins, there is large variation in the contribution of glacier and snow melt to total runo 2 , which is poorly quantified. The lack of understanding of the hydrological regimes of High Asia's rivers is one of the main sources of uncertainty in assessing the regional hydrological impacts of climate change 3 . Here we use a large-scale, high-resolution cryospheric-hydrological model to quantify the upstream hydrological regimes of the Indus, Ganges, Brahmaputra, Salween and Mekong rivers. Subsequently, we analyse the impacts of climate change on future water availability in these basins using the latest climate model ensemble. Despite large di erences in runo composition and regimes between basins and between tributaries within basins, we project an increase in runo at least until 2050 caused primarily by an increase in precipitation in the upper Ganges, Brahmaputra, Salween and Mekong basins and from accelerated melt in the upper Indus Basin. These findings have immediate consequences for climate change policies where a transition towards coping with intra-annual shifts in water availability is desirable.

This is a PDF file of a peer-reviewed paper that has been accepted for publication. Although unedited, the content has been subjected to preliminary formatting. Nature is providing this early version of the typeset paper as a service to our authors and readers. The text and figures will undergo copyediting and a proof review before the paper is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers apply.

Glaciers in the high mountains of Asia (HMA) make a substantial contribution to the water supply of millions of people 1,2 , and they are retreating and losing mass as a result of anthropogenic climate change 3 at similar rates to those seen elsewhere 4,5 . In the Paris Agreement of 2015, 195 nations agreed on the aspiration to limit the level of global temperature rise to 1.5 degrees Celsius ( °C) above preindustrial levels. However, it is not known what an increase of 1.5 °C would mean for the glaciers in HMA. Here we show that a global temperature rise of 1.5 °C will lead to a warming of 2.1 ± 0.1 °C in HMA, and that 64 ± 7 per cent of the present-day ice mass stored in the HMA glaciers will remain by the end of the century. The 1.5 °C goal is extremely ambitious and is projected by only a small number of climate models of the conservative IPCC's Representative Concentration Pathway (RCP)2.6 ensemble. Projections for RCP4.5, RCP6.0 and RCP8.5 reveal that much of the glacier ice is likely to disappear, with projected mass losses of 49 ± 7 per cent, 51 ± 6 per cent and 64 ± 5 per cent, respectively, by the end of the century; these projections have potentially serious consequences for regional water management and mountain communities.Temperatures are rising faster in high-altitude regions, including HMA, than in low-lying plains 6 . Possible explanations for this elevation-dependent warming in mountains include the effects of snow albedo and surface-based feedback, water vapour changes and latent heat release, radiative flux changes, surface heat loss and temperature change, and aerosols. A global ensemble of 110 general circulation model (GCM) runs spanning the full range of radiative forcing defined in the Coupled Model Intercomparison Project Phase 5 (CMIP5) 7 (RCP2.6 (n = 25), RCP4.5 (n = 35), RCP6.0 (n = 18) and RCP8.5 (n = 32); Supplementary Table 5) shows an evident relation between radiative forcing and projected temperature increase from pre-industrial conditions (1851-1880) to the end of this century (2071-2100, EOC) (Fig. 1). It also shows that the glacierized areas of HMA are consistently warming at much higher rates than the global average and that the difference between global and HMA temperature rises is increasing with radiative forcing (Fig. 1). Compared to the global warming of land masses only, the enhanced warming in HMA is less pronounced, but still evident. From the GCM ensemble, we have selected models that result in a 1.5 °C temperature rise globally relative to pre-industrial conditions (n = 6, see Methods). All of the selected models originate from the conservative RCP2.6 model ensemble. The 1.5 °C global increase implies a warming of 2.1 ± 0.1 °C for the glacierized areas in HMA (Fig. 2). Although there is considerable regional variation, with the Hindu Kush warming the most (2.3 °C) and the Eastern Himalaya the least (1.9 °C), all regions warm by more than 1.5 °C. These spatial patterns persist for the RCP2.6, RCP4.5, RCP6.0 and RCP8.5 scenarios, for which considerably higher warming...

Abstract. Mountain ranges in Asia are important water suppliers, especially if downstream climates are arid, water demands are high and glaciers are abundant. In such basins, the hydrological cycle depends heavily on high-altitude precipitation. Yet direct observations of high-altitude precipitation are lacking and satellite derived products are of insufficient resolution and quality to capture spatial variation and magnitude of mountain precipitation. Here we use glacier mass balances to inversely infer the high-altitude precipitation in the upper Indus basin and show that the amount of precipitation required to sustain the observed mass balances of large glacier systems is far beyond what is observed at valley stations or estimated by gridded precipitation products. An independent validation with observed river flow confirms that the water balance can indeed only be closed when the highaltitude precipitation on average is more than twice as high and in extreme cases up to a factor of 10 higher than previously thought. We conclude that these findings alter the present understanding of high-altitude hydrology and will have an important bearing on climate change impact studies, planning and design of hydropower plants and irrigation reservoirs as well as the regional geopolitical situation in general.

The Indus basin heavily depends on its upstream mountainous part for the downstream supply of water while downstream demands are high. Since downstream demands will likely continue to increase, accurate hydrological projections for the future supply are important. We use an ensemble of statistically downscaled CMIP5 General Circulation Model outputs for RCP4.5 and RCP8.5 to force a cryospheric-hydrological model and generate transient hydrological projections for the entire 21st century for the upper Indus basin. Three methodological advances are introduced: (i) A new precipitation dataset that corrects for the underestimation of high-altitude precipitation is used. (ii) The model is calibrated using data on river runoff, snow cover and geodetic glacier mass balance. (iii) An advanced statistical downscaling technique is used that accounts for changes in precipitation extremes. The analysis of the results focuses on changes in sources of runoff, seasonality and hydrological extremes. We conclude that the future of the upper Indus basin’s water availability is highly uncertain in the long run, mainly due to the large spread in the future precipitation projections. Despite large uncertainties in the future climate and long-term water availability, basin-wide patterns and trends of seasonal shifts in water availability are consistent across climate change scenarios. Most prominent is the attenuation of the annual hydrograph and shift from summer peak flow towards the other seasons for most ensemble members. In addition there are distinct spatial patterns in the response that relate to monsoon influence and the importance of meltwater. Analysis of future hydrological extremes reveals that increases in intensity and frequency of extreme discharges are very likely for most of the upper Indus basin and most ensemble members.