More stories

  • in

    Pacific decadal variability over the last 2000 years and implications for climatic risk

    Power, S., Casey, T., Folland, C., Colman, A. & Mehta, V. Interdecadal modulation of the impact of ENSO on Australia. Clim. Dyn. 15, 319–324 (1999).
    Google Scholar 
    Mantua, N. & Hare, S. The Pacific decadal oscillation. J. Oceanogr. 58, 35–44 (2002).
    Google Scholar 
    Kiem, A. S., Franks, S. W. & Kuczera, G. Multi-decadal variability of flood risk. Geophys. Res. Lett. 30, 1035 (2003).
    Google Scholar 
    Kiem, A. S. & Franks, S. W. Multi-decadal variability of drought risk – Eastern Australia. Hydrol. Process. 18, 2039–2050 (2004).
    Google Scholar 
    Verdon, D. C., Kiem, A. S. & Franks, S. W. Multi-decadal variability of forest fire risk – Eastern Australia. Int. J. Wildland Fire 13, 165–171 (2004).
    Google Scholar 
    Parker, D. et al. Decadal to multidecadal variability and the climate change background. J. Geophys. Res. Atmos. https://doi.org/10.1029/2007JD008411 (2007).England, M. H. et al. Recent intensification of wind-driven circulation in the Pacific and the ongoing warming hiatus. Nat. Clim. Change https://doi.org/10.1038/NCLIMATE2106 (2014).Dai, A., Fyfe, J. C., Xie, S.-P. & Dai, X. Decadal modulation of global surface temperature by internal climate variability. Nat. Clim. Change 5, 555–559 (2015).
    Google Scholar 
    Henley, B. J. Pacific decadal variability: indices, patterns and tropical-extratropical interactions. Glob. Planet. Change 155, 42–55 (2017).
    Google Scholar 
    Magee, A. D. & Verdon-Kidd, D. C. Historical variability of Southwest Pacific tropical cyclone counts since 1855. Geophys. Res. Lett. https://doi.org/10.1029/2019GL082900 (2019).Gray, J. L., Verdon-Kidd, D. C., Callaghan, J. & English, N. B. On the recent hiatus of tropical cyclones landfalling in NSW, Australia. J. South. Hemisph. Earth Syst. Sci. 70, 180–192 (2020).
    Google Scholar 
    Meehl, G., Arblaster, J., Bitz, C., Chung, C. & Teng, H. Antarctic sea-ice expansion between 2000 and 2014 driven by tropical Pacific decadal climate variability. Nat. Geosci. 9, 590–595 (2016).CAS 

    Google Scholar 
    Clem, K. R. et al. Record warming at the South Pole during the past three decades. Nat. Clim. Change https://doi.org/10.1038/s41558-020-0815-z (2020).Turner, J. et al. Absence of 21st century warming on Antarctic Peninsula consistent with natural variability. Nature 535, 411–415 (2016).CAS 

    Google Scholar 
    Jones, J. M. et al. Assessing recent trends in high-latitude Southern Hemisphere surface climate. Nat. Clim. Change 6, 917–926 (2016).
    Google Scholar 
    Newman, M., Compo, G. P. & Alexander, M. A. ENSO-forced variability of the Pacific Decadal Oscillation. J. Clim. 16, 3853–3857 (2003).
    Google Scholar 
    Liu, Z. Y. Dynamics of interdecadal climate variability: a historical perspective. J. Clim. 25, 1963–1995 (2012).
    Google Scholar 
    Smith, D. M. et al. Role of volcanic and anthropogenic aerosols in the recent global surface warming slowdown. Nat. Clim. Change 6, 936–941 (2016).CAS 

    Google Scholar 
    Lou, J., Holbrook, N. J. & O’Kane, T. J. South Pacific decadal climate variability and potential predictability. J. Clim. 32, 6051–6069 (2019).
    Google Scholar 
    Mann, M. E., Steinman, B. A. & Miller, S. K. Absence of internal multidecadal and interdecadal oscillations in climate model simulations. Nat. Commun. 11, 49 (2020).CAS 

    Google Scholar 
    Mann, M. E., Steinman, B. A., Brouillette, D. J. & Miller, S. K. Multidecadal climate oscillations during the past millennium driven by volcanic forcing. Science 371, 1014–1019 (2021).CAS 

    Google Scholar 
    Folland, C. K., Renwick, J. A., Salinger, M. J. & Mullen, A. B. Relative influences of the Interdecadal Pacific Oscillation and ENSO on the South Pacific Convergence Zone. Geophys. Res. Lett. 29, 1643 (2002).
    Google Scholar 
    D’Arrigo, R. & Wilson, R. On the Asian expression of the PDO. Int. J. Climatol. 26, 1607–1617 (2006).
    Google Scholar 
    Henley, B. et al. A tripole index for the Interdecadal Pacific Oscillation. Clim. Dyn. 45, 3077–3090 (2015).
    Google Scholar 
    Bonfils, C. J. W. et al. Human influence on joint changes in temperature, rainfall and continental aridity. Nat. Clim. Change 10, 726–731 (2020).CAS 

    Google Scholar 
    Zhang, L., Kuczera, G., Kiem, A. S. & Willgoose, G. R. Using paleoclimate reconstructions to analyse hydrological epochs associated with Pacific decadal variability. Hydrol. Earth Syst. Sci. 22, 6399–6414 (2018).
    Google Scholar 
    Collins, M. et al. in Climate Change 2013: The Physical Science Basis (eds Stocker, T. F. et al.) 1029–1136 (Cambridge University Press, 2013).Armstrong, M. S., Kiem, A. S. & Vance, T. R. Comparing instrumental, palaeoclimate and projected rainfall data: Implications for water resources management and hydrological modelling. J. Hydrol. 31, 100728 (2020).
    Google Scholar 
    Buckley, B. M. et al. Interdecadal Pacific Oscillation reconstructed from trans-Pacific tree rings: 1350–2004 CE. Clim. Dyn. 53, 3181–3196 (2019).
    Google Scholar 
    Verdon, D. C. & Franks, S. W. Long-term behaviour of ENSO: Interactions with the PDO over the past 400 years inferred from palaeoclimate records. Geophys. Res. Lett. https://doi.org/10.1029/2005GL025052 (2006).Porter, S. E., Mosley-Thompson, E., Thompson, L. G. & Wilson, A. B. Reconstructing a Pacific Interdecadal Oscillation Index from a Pacific basin-wide collection of ice core records. J. Clim. 34, 3839–3852 (2021).
    Google Scholar 
    Vance, T. R., Roberts, J. L., Plummer, C. T., Kiem, A. S. & van Ommen, T. D. Interdecadal Pacific variability and eastern Australian megadroughts over the last millennium. Geophys. Res. Lett. 42, 129–137 (2015).
    Google Scholar 
    Roberts, J. et al. A 2000-year annual record of snow accumulation rates for Law Dome, East Antarctica. Clim. Past 11, 697–707 (2015).
    Google Scholar 
    van Ommen, T. D. & Morgan, V. Snowfall increase in coastal East Antarctica linked with southwest Western Australian drought. Nat. Geosci. 3, 267 (2010).
    Google Scholar 
    Vance, T. R., Ommen, T. D. V., Curran, M. A. J., Plummer, C. T. & Moy, A. D. A millennial proxy record of ENSO and Eastern Australian rainfall from the Law Dome Ice Core, East Antarctica. J. Clim. 26, 710–725 (2013).
    Google Scholar 
    Vance, T. R. et al. Optimal site selection for a high-resolution ice core record in East Antarctica. Clim. Past 12, 595–610 (2016).
    Google Scholar 
    Udy, D. G., Vance, T. R., Kiem, A. S., Holbrook, N. J. & Curran, M. A. J. Links between large-scale modes of climate variability and synoptic weather patterns in the southern Indian Ocean. J. Clim. 34, 883–899 (2021).
    Google Scholar 
    Crockart, C. K. et al. El Niño-Southern Oscillation signal in a new East Antarctic ice core, Mount Brown South. Clim. Past 17, 1795–1818 (2021).
    Google Scholar 
    Stevens, H. R. & Kiem, A. S. Developing hazard lines in response to coastal flooding and sea level change. Urban Pol. Res. 32, 341–360 (2014).
    Google Scholar 
    Magee, A. D., Verdon-Kidd, D. C., Diamond, H. J. & Kiem, A. S. Influence of ENSO, ENSO Modoki, and the IPO on tropical cyclogenesis: a spatial analysis of the southwest Pacific region. Int. J. Climatol. 37, 1118–1137 (2017).
    Google Scholar 
    McMahon, G. M. & Kiem, A. S. Large floods in South East Queensland, Australia: is it valid to assume they occur randomly? Austral. J. Water Resour. 22, 4–14 (2018).
    Google Scholar 
    Deb, P. et al. Causes of the Widespread 2019–2020 Australian Bushfire Season. Earth’s Future 8, e2020EF001671 (2020).
    Google Scholar 
    Magee, A. D. & Kiem, A. S. Using indicators of ENSO, IOD, and SAM to improve lead time and accuracy of tropical cyclone outlooks for Australia. J. Appl. Meteorol. Climatol. https://doi.org/10.1175/jamc-d-20-0131.1 (2020).van Dijk, A. I. J. M. et al. The millennium drought in southeast Australia (2001–2009): natural and human causes and implications for water resources, ecosystems, economy and society. Water Resour. Res. 49, 1–18 (2013).
    Google Scholar 
    Johnson, F. et al. Natural hazards in Australia: floods. Clim. Change 139, 21–35 (2016).
    Google Scholar 
    Holgate, C. M., van Dijk, A. I. J. M., Evans, J. P. & Pitman, A. J. Local and remote drivers of southeast Australian drought. Geophys. Res. Lett. https://doi.org/10.1029/2020GL090238 (2020).Chiew, F. S. H. Estimation of rainfall elasticity of streamflow in Australia. Hydrol. Sci. J. 51, 613–625 (2006).
    Google Scholar 
    Wooldridge, S. A., Franks, S. W. & Kalma, J. D. Hydrological implications of the Southern Oscillation: variability of the rainfall-runoff relationship. Hydrol. Sci. J. 46, 73–88 (2001).
    Google Scholar 
    Kiem, A. S. & Verdon-Kidd, D. C. Climatic drivers of Victorian streamflow: is ENSO the dominant influence? Austral. J. Water Resour. 13, 17–29 (2009).
    Google Scholar 
    Flack, A. L., Kiem, A. S., Vance, T. R., Tozer, C. R. & Roberts, J. L. Comparison of published palaeoclimate records suitable for reconstructing annual to sub-decadal hydroclimatic variability in eastern Australia: implications for water resource management and planning. Hydrol. Earth Syst. Sci. 24, 5699–5712 (2020).
    Google Scholar 
    Barr, C. et al. Holocene El Niño–Southern Oscillation variability reflected in subtropical Australian precipitation. Sci. Rep. 9, 1627 (2019).CAS 

    Google Scholar 
    Tozer, C. R. et al. Reconstructing pre-instrumental streamflow in Eastern Australia using a water balance approach. J. Hydrol. 558, 632–646 (2018).
    Google Scholar 
    Etheridge, D. M., Steele, L. P., Lagenfelds, R. L. & Francey, R. J. Natural and anthropogenic changes in atmospheric CO2 over the last 1000 years from air in Antarctic ice and firn. J. Geophys. Res. 101, 4115–4128 (1996).CAS 

    Google Scholar 
    Massom, R. A. et al. Precipitation over the Interior East Antarctic Ice Sheet related to midlatitude blocking-high activity. J. Clim. 17, 1914–1928 (2004).
    Google Scholar 
    Morgan, V. et al. Site information and initial results from deep ice drilling on Law Dome. J. Glaciol. 43, 3–10 (1997).CAS 

    Google Scholar 
    Plummer, C. T. et al. An independently dated 200-yr volcanic record from Law Dome, East Antarctica, including a new perspective on the dating of the c. 1450s eruption of Kuwae, Vanuatu. Clim. Past 8, 1929–1940 (2012).
    Google Scholar 
    Curran, M. A. J., van Ommen, T. D., Morgan, V. I., Phillips, K. L. & Palmer, A. S. Ice core evidence for Antarctic sea ice decline since the 1950s. Science 302, 1203–1206 (2003).CAS 

    Google Scholar 
    Curran, M. A. J. & Palmer, A. S. Suppressed ion chromatography method for the routine determination of ultra low level anions and cations in ice cores. J. Chromatogr. A 919, 107–113 (2001).CAS 

    Google Scholar 
    van Ommen, T. D. & Morgan, V. Peroxide concentrations in the Dome Summit South ice core, Law Dome, Antarctica. J. Geophys. Res. 101, 15,147–15,152 (1996).
    Google Scholar 
    van Ommen, T. D. & Morgan, V. Calibrating the ice core paleothermometer using seasonality. J. Geophys. Res. 102, 9351–9357 (1997).
    Google Scholar 
    Kiem, A. S. et al. Learning from the past – using palaeoclimate data to better understand and manage drought in South East Queensland (SEQ), Australia. J. Hydrol. Reg. Stud. https://doi.org/10.1016/j.ejrh.2020.100686 (2020).Friedman, J. H. Multivariate adaptive regression spines. Ann. Stat. 19, 1–67 (1991).
    Google Scholar 
    Roberts, J. L. et al. Reconciling unevenly sampled palaeoclimate proxies: a Gaussian kernel correlation multiproxy reconstruction. J. Environ. Inform. https://doi.org/10.3808/jei.201900420 (2019).Mann, M. E. & Lees, J. Robust estimation of background noise and signal detection in climatic time series. Clim. Change 33, 409–455 (1996).
    Google Scholar  More

  • in

    Gran Turismo champion, reimagined urine — the week in infographics

    NEWS
    15 February 2022

    Gran Turismo champion, reimagined urine — the week in infographics

    Nature highlights three key graphics from the week in science and research.

    Twitter

    Facebook

    Email

    Artificial intelligence overtakes human gamersThis graphic shows one way in which an artificial intelligence (AI) is able to win against the best human players of the video game Gran Turismo. In a paper in Nature, a team of researchers introduce GT Sophy, which learns through a neural-network model. GT Sophy stands out for its performance against human drivers in a head-to-head competition. Far from using a lap-time advantage to outlast opponents, GT Sophy simply outraces them. Through the training process, GT Sophy learnt to take different lines through the corners in response to different conditions. Our graphic shows how, in one case, two human drivers attempted to block the preferred path of two GT Sophy cars, yet the AI succeeded in finding two trajectories that overcame this block and allowed its cars to overtake. You can read more about what it takes to win at racing (both real and simulated) in this News & Views article.

    The march of methaneLevels of methane, a potent greenhouse gas, have been growing for decades — but they began a rapid and mysterious uptick around 2007. Last year, methane concentrations in the atmosphere raced past 1,900 parts per billion, nearly triple pre-industrial levels, according to data released in January by the US National Oceanic and Atmospheric Administration. Where is it coming from? Potential explanations range from the expanding exploitation of oil and natural gas, and rising emissions from landfill, to growing livestock herds and increasing activity by microbes in wetlands. The spike has caused many researchers to worry that global warming is creating a feedback mechanism that will cause ever more methane to be released, making it even harder to rein in rising global temperatures.

    Source: NOAA

    Urine, reimaginedOur final graphic this week illustrates some of the many ways in which human urine could be recycled into useful products. Scientists say that urine diversion would have huge environmental and public-health benefits if deployed on a large scale. That’s in part because urine is rich in nutrients that could help to fertilize crops or feed into industrial processes; furthermore, not flushing urine down the drain could save vast amounts of water.But urine diversion and reuse would require “drastic reimagining of how we do human sanitation”, as a Feature reports. It would involve wide-scale use of special urine-diverting toilets, and even processing devices in your building’s basement.

    doi: https://doi.org/10.1038/d41586-022-00458-z

    Related Articles

    Neural networks overtake humans in Gran Turismo racing game

    Scientists raise alarm over ‘dangerously fast’ growth in atmospheric methane

    The urine revolution: how recycling pee could help to save the world

    Subjects

    Machine learning

    Water resources

    Climate change

    Latest on:

    Machine learning

    Neural networks overtake humans in Gran Turismo racing game
    News & Views 09 FEB 22

    Early prediction of preeclampsia in pregnancy with cell-free RNA
    Article 09 FEB 22

    DeepMind AI tackles one of chemistry’s most valuable techniques
    News 10 DEC 21

    Water resources

    The urine revolution: how recycling pee could help to save the world
    News Feature 09 FEB 22

    ‘Sky river’ brought Iran deadly floods — but also welcome water
    Research Highlight 08 DEC 21

    Brazil is in water crisis — it needs a drought plan
    Comment 08 DEC 21

    Climate change

    Scientists raise alarm over ‘dangerously fast’ growth in atmospheric methane
    News 08 FEB 22

    Climate pledges from top companies crumble under scrutiny
    News 07 FEB 22

    Emissions cuts take political and social innovation too
    Correspondence 01 FEB 22

    Jobs

    Flow cytometrist – Sorter Operator (m/f/div)

    Max Planck Institute for Biology of Ageing (MPIAGE)
    Köln, Germany

    Physicist, Engineer Electrical Engineering and/or photonics or similar (f/m/x) – Doctoral Fellowship – Optical technologies for space navigation

    German Aerospace Center (DLR)
    Oberpfaffenhofen, Germany

    Doctoral Researcher (f/m/d) 75 % for the project C2 Seismic imaging by full waveform inversion

    Karlsruhe Institute of Technology (KIT)
    Karlsruhe, Germany

    Research associate / PhD candidate (f/m/d) Embedded Electronic Systems

    Karlsruhe Institute of Technology (KIT)
    Karlsruhe, Germany More

  • in

    Rapid intensification of the emerging southwestern North American megadrought in 2020–2021

    Mankin, J. S. et al. NOAA Drought Task Force Report on the 2020–2021 Southwestern U.S. Drought (NOAA Drought Task Force, MAPP and NIDIS, 2021); https://www.drought.gov/sites/default/files/2021-09/NOAA-Drought-Task-Force-IV-Southwest-Drought-Report-9-23-21.pdfH.R.2030—Colorado River Drought Contingency Plan Authorization Act (US House of Representatives, 2019); https://www.congress.gov/bill/116th-congress/house-bill/2030Svoboda, M. et al. The Drought Monitor. Bull. Am. Meteorol. Soc. 83, 1181–1190 (2002).Article 

    Google Scholar 
    Cook, E. R. et al. Megadroughts in North America: placing IPCC projections of hydroclimatic change in a long-term palaeoclimate context. J. Quat. Sci. 25, 48–61 (2010).Article 

    Google Scholar 
    Williams, A. P. et al. Large contribution from anthropogenic warming to a developing North American megadrought. Science 368, 314–318 (2020).CAS 
    Article 

    Google Scholar 
    Cook, B. I. et al. North American megadroughts in the Common Era: reconstructions and simulations. WIREs Clim. Change 7, 411–432 (2016).Article 

    Google Scholar 
    Woodhouse, C. A. & Overpeck, J. T. 2000 years of drought variability in the central United States. Bull. Am. Meteorol. Soc. 79, 2693–2714 (1998).Article 

    Google Scholar 
    Breshears, D. D. et al. Regional vegetation die-off in response to global-change-type drought. Proc. Natl Acad. Sci. USA 102, 15144–15148 (2005).CAS 
    Article 

    Google Scholar 
    Williams, A. P. et al. Temperature as a potent driver of regional forest drought stress and tree mortality. Nat. Clim. Change 3, 292–297 (2013).Article 

    Google Scholar 
    Eyring, V. et al. Overview of the Coupled Model Intercomparison Project Phase 6 (CMIP6) experimental design and organization. Geosci. Model Dev. 9, 1937–1958 (2016).Barnett, T. P., Adam, J. C. & Lettenmaier, D. P. Potential impacts of a warming climate on water availability in snow-dominated regions. Nature 438, 303–309 (2005).CAS 
    Article 

    Google Scholar 
    Das, T., Pierce, D. W., Cayan, D. R., Vano, J. A. & Lettenmaier, D. P. The importance of warm season warming to western US streamflow changes. Geophys. Res. Lett. 38, L23403 (2011).Article 

    Google Scholar 
    Milly, P. C. D. & Dunne, K. A. Colorado River flow dwindles as warming-driven loss of reflective snow energizes evaporation. Science 367, 1252–1255 (2020).CAS 
    Article 

    Google Scholar 
    Pascolini-Campbell, M., Reager, J. T., Chandanpurkar, H. A. & Rodell, M. A 10 per cent increase in global land evapotranspiration from 2003 to 2019. Nature 593, 543–547 (2021).CAS 
    Article 

    Google Scholar 
    Udall, B. & Overpeck, J. The twenty‐first century Colorado River hot drought and implications for the future. Water Resour. Res. 53, 2404–2418 (2017).Article 

    Google Scholar 
    Cook, B. I. et al. Uncertainties, limits, and benefits of climate change mitigation for soil moisture drought in southwestern North America. Earth’s Future 9, e2021EF002014 (2021).Article 

    Google Scholar 
    Ault, T. R., Mankin, J. S., Cook, B. I. & Smerdon, J. E. Relative impacts of mitigation, temperature, and precipitation on 21st-century megadrought risk in the American Southwest. Sci. Adv. 2, e1600873 (2016).Article 

    Google Scholar 
    Woodhouse, C. A., Meko, D. M., MacDonald, G. M. & Stahle, D. W. A 1,200-year perspective of 21st century drought in southwestern North America. Proc. Natl Acad. Sci. USA 107, 21283–21288 (2010).CAS 
    Article 

    Google Scholar 
    Lepley, K. et al. A multi-century Sierra Nevada snowpack reconstruction modeled using upper-elevation coniferous tree rings (California, USA). Holocene 30, 1266–1278 (2020).Article 

    Google Scholar 
    Williams, A. P. et al. Tree rings and observations suggest no stable cycles in Sierra Nevada cool-season precipitation. Water Resour. Res. 57, e2020WR028599 (2021).Article 

    Google Scholar  More

  • in

    The urine revolution: how recycling pee could help to save the world

    NEWS FEATURE
    09 February 2022

    The urine revolution: how recycling pee could help to save the world

    Separating urine from the rest of sewage could mitigate some difficult environmental problems, but there are big obstacles to radically re-engineering one of the most basic aspects of life.

    Chelsea Wald

    0

    Chelsea Wald

    Chelsea Wald is a freelance reporter in The Hague, the Netherlands, and the author of Pipe Dreams: The Urgent Global Quest to Transform the Toilet.

    View author publications

    You can also search for this author in PubMed
     Google Scholar

    Twitter

    Facebook

    Email

    Download PDF

    Specialized toilet systems recover nitrogen and other nutrients from urine for use as fertilizers and other products.Credit: MAK/Georg Mayer/EOOS NEXT

    On Gotland, the largest island in Sweden, fresh water is scarce. At the same time, residents are battling dangerous amounts of pollution from agriculture and sewer systems that causes harmful algal blooms in the surrounding Baltic Sea. These can kill fish and make people ill.To help solve this set of environmental challenges, the island is pinning its hopes on a single, unlikely substance that connects them: human urine.Starting in 2021, a team of researchers began collaborating with a local company that rents out portable toilets. The goal is to collect more than 70,000 litres of urine over 3 years from waterless urinals and specialized toilets at several locations during the booming summer tourist season. The team is from the Swedish University of Agricultural Sciences (SLU) in Uppsala, which has spun off a company called Sanitation360. Using a process that the researchers developed, they are drying the urine into concrete-like chunks that they hammer into a powder and press into fertilizer pellets that fit into standard farming equipment. A local farmer uses the fertilizer to grow barley that will go to a brewery to make ale — which, after consumption, could enter the cycle all over again.The researchers aim to take urine reuse “beyond concept and into practice” on a large scale, says Prithvi Simha, a chemical-process engineer at the SLU and Sanitation360’s chief technology officer. The aim is to provide a model that regions around the world could follow. “The ambition is that everyone, everywhere, does this practice.”

    The Gotland experiment compared barley fertilized with urine (right) to plants grown without fertilizer (middle) and ones with mineral fertilizer (left).Credit: Jenna Senecal

    The Gotland project is part of a wave of similar efforts worldwide to separate urine from the rest of sewage and to recycle it into products such as fertilizer. That practice, known as urine diversion, is being studied by groups in the United States, Australia, Switzerland, Ethiopia and South Africa, among other places. The efforts reach far beyond the confines of university labs. Waterless urinals connect to basement treatment systems in offices in Oregon and the Netherlands. In Paris, there are plans to install urine-diverting toilets in a 1,000-resident eco-quarter being built in the 14th district of the city. The European Space Agency is to put 80 urine-diverting toilets into its Paris headquarters, which will begin operating later this year. According to proponents of urine diversion, it could see uses in sites from temporary military outposts to refugee encampments, rich urban centres and sprawling slums.Scientists say that urine diversion would have huge environmental and public-health benefits if deployed on a large scale around the world. That’s in part because urine is rich in nutrients that, instead of polluting water bodies, could go towards fertilizing crops or feed into industrial processes. According to Simha’s estimates, humans produce enough urine to replace about one-quarter of current nitrogen and phosphorus fertilizers worldwide; it also contains potassium and many micronutrients (see ‘What’s in urine’). On top of that, not flushing urine down the drain could save vast amounts of water and reduce some of the strain on ageing and overloaded sewer systems.

    Source: M. Qadir et al. Nat. Resour. Forum 44, 40–51 (2020)

    Thanks to advances in toilets and urine-treatment strategies, many components of urine diversion could soon be ready for widespread roll-out, according to experts in the field. But there are also big obstacles to radically re-engineering one of the most basic aspects of life. Researchers and companies need to solve a number of problems, from improving the design of urine-diverting toilets to making it easier to treat urine and turn it into valuable products. This could involve chemical-treatment systems connected to individual toilets or basement devices that serve entire buildings, with pick-up and maintenance services for the resulting concentrated or solidified product (see ‘From pee to products’). Then there are broader questions of social change and acceptance, related both to varying levels of cultural taboos around human waste and to deeply entrenched conventions about industrial sewage and food systems.Urine diversion and reuse is the type of “drastic reimagining of how we do human sanitation” that will become increasingly crucial as societies battle shortages in energy, water and raw materials for agriculture and industry, says biologist Lynn Broaddus, a sustainability consultant in Minneapolis, Minnesota, who is former president of the Water Environment Federation in Alexandria, Virginia, an association of water-quality professionals worldwide. “The fact of the matter is, it’s valuable stuff.”

    Mixed wasteUrine used to be a valuable commodity. In the past, some societies used it for fertilizing crops, tanning leather, washing clothes and producing gunpowder. Then, in the late nineteenth and early twentieth century, the modern model of centralized sewage management arose in England and spread worldwide, ultimately leading to what has been called urine blindness.In this model, flush toilets use water to quickly send urine, faeces and toilet paper into sewers, where it mixes with other liquids from households, industrial sources and sometimes storm run-off. At centralized treatment plants, an energy-intensive process uses microbes to clean the sewage.Depending on local regulations and a treatment plant’s condition, the wastewater discharged from the process can still contain a lot of nitrogen and other nutrients, as well as some other contaminants. And 57% of the world’s population isn’t connected to centralized sewer systems at all (see ‘Human sewage’).

    Source: C. Tuholske et al. PLoS ONE 16, e0258898 (2021).

    Scientists are working on ways to make centralized systems more sustainable and less polluting, but, beginning in Sweden in the 1990s, some researchers began pushing for more fundamental change. The end-of-pipe advances are “just, you know, another evolution of the same damn thing”, says Nancy Love, an environmental engineer at the University of Michigan in Ann Arbor. Urine diversion would be “transformative”, she says. In a study1 that modelled wastewater-management systems in three US states, she and her colleagues compared conventional wastewater systems with hypothetical ones that divert urine and use the recovered nutrients to replace synthetic fertilizers. They projected that communities with urine diversion could lower their overall greenhouse-gas emissions by up to 47%, energy consumption by up to 41%, freshwater use by about half, and nutrient pollution from the wastewater by up to 64%, depending on the technologies used.Still, the concept has remained niche, mostly limited to off-grid locales such as northern European eco-villages, rural outhouses and development projects in low-income settings.A lot of the lag is a result of the toilets themselves, says Tove Larsen, a chemical engineer at the Swiss Federal Institute of Aquatic Science and Technology (Eawag) in Dübendorf. First sold in the 1990s and 2000s, most urine-diverting toilets have a small basin at the front to capture the liquid — a set-up that requires careful aim. Other designs have incorporated foot-powered conveyor belts that let urine drain away while transporting the faeces to a composting vault, or sensors that operate valves to direct the urine to separate outlets.

    A prototype toilet that separates urine and dries it into a powder is being tested at the head office of VA SYD, the Swedish public water and wastewater utility, in Malmö.Credit: Lotte Kristoferitsch

    But in European pilot and demonstration projects, people failed to embrace their use, Larsen says, complaining that they were too unwieldy, smelly and unreliable. “We have really been stalled by this topic of toilets.”These concerns plagued the first large-scale use of urine-diversion toilets — a project in the 2000s in South Africa’s eThekwini municipality. After apartheid, the municipality’s boundaries suddenly expanded, causing authorities to take over responsibility for some poor rural areas where there was no toilet infrastructure and little water service, says Anthony Odili, who researches sanitation governance at the University of KwaZulu-Natal in Durban.After a cholera outbreak there in August 2000, the authorities quickly rolled out several types of sanitation that met financial and practical constraints, including about 80,000 urine-diversion dry toilets, most of which are still in use today. The urine drains below the toilet into the soil and the faeces falls into a vault, which, since 2016, the municipality has emptied every five years.
    The secret history of ancient toilets
    The project was successful at establishing safer sanitation in the region, Odili says. Social-science research, however, has revealed many problems with the programme. Although people felt that the toilets were better than nothing, Odili says, studies — including some he was involved in2 — later found that users generally disliked them. Many had been constructed with poor materials and were awkward to use. Although such toilets should prevent bad odours in theory, urine in the eThekwini ones often entered the vaults with the faeces, causing a terrible stink. People were “not able to breathe properly”, Odili says. What’s more, the urine remains largely unused.Ultimately, the decision to go with urine-diversion dry toilets, driven largely by public-health concerns, was top-down, and failed to take people’s preferences into account, Odili says. A 2017 study3 found that more than 95% of respondents in eThekwini aspired to the convenient, odourless flush toilets that wealthier white people use in the city — and that many have intentions to install them when their circumstances allow. In South Africa, toilets have long served as a symbol of racial disparity.A new design, however, could represent a breakthrough for urine diversion. Led by designer Harald Gründl and in collaboration with Larsen and others, in 2017, the Austrian design firm EOOS (which has since spun off the company EOOS Next) unveiled the Urine Trap. This removes the need for users to aim, and the urine-diverting function is almost invisible (see ‘A new kind of toilet’).

    Source: EOOS

    It takes advantage of water’s tendency to cling to surfaces (known as the teapot effect because it’s like an inconveniently dribbling teapot) to direct urine down the front inner side of the toilet into a separate hole (see ‘How to recycle pee’). Developed with funding from the Bill & Melinda Gates Foundation in Seattle, Washington, which has supported a broad swathe of research into toilet innovation for low-income settings, the Urine Trap can be incorporated into everything from high-end ceramic pedestal models to plastic squat pans. LAUFEN, a manufacturer headquartered in Switzerland, is already producing one for the European market, called save!, although it is too costly for many consumers.The University of KwaZulu-Natal and the eThekwini municipality have also been testing versions of Urine Trap toilets that divert the urine and flush the solids. This time, the research is more focused on the user. Odili is optimistic that people will prefer the new urine-diversion toilets because they smell better and are easier to use, but he points out that men would have to sit down to urinate, which is a big cultural shift. But if the toilet is “also adopted and accepted in high-income areas — people from different racial groups here — it really will help in the roll-out”, he says. “We must always put on that racial lens,” he adds, to ensure that they’re not developing something that will be seen as ‘just for Black people’ or ‘just for poor people’.Uses for urineSeparating urine is just the first step in transforming sanitation. The next part is working out what to do with it. In rural areas, people could store it in vats to kill any pathogens and then apply it to fields. The World Health Organization provides guidelines for this practice.But urban settings are trickier — and that’s where most urine is produced. It’s not practical to add a separate set of sewer pipes throughout a city to move urine to a central location. And because urine is about 95% water, it is too expensive to store and transport. So researchers are focusing on drying, concentrating or otherwise extracting nutrients from urine at the toilet or building level, leaving the water behind.This isn’t easy, says Larsen. From an engineering perspective, “urine is a nasty solution”, she says. Aside from water, the largest portion is urea, a nitrogen-rich compound that bodies produce as a by-product of metabolizing proteins. Urea by itself is useful: a synthetic version is a common nitrogen fertilizer (see ‘Nitrogen demand’). But it’s also tricky: when combined with water, the urea transforms into ammonia gas, which helps to give urine its characteristic scent. If not contained, the ammonia stinks, pollutes the air and carries valuable nitrogen away. Catalysed by the widespread enzyme urease, this reaction, called urea hydrolysis, can take microseconds, making urease one of the most efficient enzymes known4.

    Source: FAO

    Some approaches allow the hydrolysis to go ahead. Researchers at Eawag have developed an advanced process for turning hydrolysed urine into a concentrated nutrient solution. First, in a tank, microorganisms transform the volatile ammonia into non-volatile ammonium nitrate, which is a common fertilizer. Then a distiller concentrates the liquid. A spin-off company called Vuna, also in Dübendorf, is working to commercialize both the system for use in buildings and the product, called Aurin, which has been approved in Switzerland for use on edible plants — a world first.Others try to stop the hydrolysis reaction by quickly raising or lowering the pH of the urine, which is usually neutral when it comes out of the body. On campus at the University of Michigan, a collaboration between Love and the non-profit Rich Earth Institute in Brattleboro, Vermont, is developing a system for buildings that squirts liquid citric acid down the pipes of a urine-diverting toilet and a waterless urinal. It then concentrates the urine through repeated freezing and thawing5.
    The new economy of excrement
    The SLU team doing the project on Gotland island, led by environmental engineer Björn Vinnerås, has worked out how to dry urine into a solid urea mixed with the other nutrients. The team is evaluating its latest prototype, a self-contained toilet including a built-in dryer, at the head office of the Swedish public water and wastewater utility VA SYD in Malmö.Other methods target individual nutrients from urine. These could more easily slot into existing supply chains for fertilizers and industrial chemicals, says chemical engineer William Tarpeh, a former postdoc of Love’s who is now at Stanford University in California.One well-established way of recovering phosphorus from hydrolysed urine is to add magnesium, which causes the precipitation of a fertilizer called struvite. And Tarpeh is experimenting with beads of adsorption materials that selectively pluck out nitrogen in the form of ammonia6 or phosphorus in the form of phosphate. His system uses another liquid, called a regenerant, to flow over the beads after they are spent. The regenerant carries off the nutrients and renews the beads for another round. It’s a low-tech, passive method, but the commercial regenerants are environmentally damaging. His team is now trying to make ones that are cheaper and more environmentally friendly (see ‘Future pollution’).

    Source: P. J. T. M. van Puijenbroek et al. J. Environ. Mgmt 231, 446–456 (2019)

    Other researchers are developing ways to produce electricity by putting urine into microbial fuel cells. In Cape Town, South Africa, another team has developed a method for making an unconventional construction brick by combining urine, sand and urease-producing bacteria in a mould; these calcify into any shape without the need for firing. And the European Space Agency is eyeing astronaut urine as a resource for building habitats on the Moon.“When I think about the big future of urine recovery and wastewater recovery, we want to be able to make as many products as possible,” Tarpeh says.As researchers pursue a slew of ideas to turn urine into commodities, they know it’s an uphill battle, particularly with entrenched industries. Fertilizer and food companies, farmers, toilet manufacturers and regulators are slow to make big changes to their practices. “There’s quite a lot of inertia,” says Simha.At the University of California, Berkeley, for example, a research and education installation of the LAUFEN save! toilet, including a drainpipe to a storage tank on the floor below, has unexpectedly taken nearly three years and cost more than US$50,000. That includes fees for architects, construction and complying with municipal codes, says environmental engineer Kevin Orner, now at West Virginia University in Morgantown — and it’s still not done. The lack of existing codes and regulations has caused troubles with facilities management, he says, which is why he is on a panel that is developing new codes.Some of the inertia might be due to concerns over customer resistance, but a 2021 survey of people in 16 countries7 indicated that willingness to consume urine-fertilized food approached 80% in places such as France, China and Uganda (see ‘Will people eat it?’).

    Source: Ref. 7

    Pam Elardo, who leads the Bureau of Wastewater Treatment as a deputy commissioner in the New York City Department of Environmental Protection, says she supports innovations such as urine diversion, because further reducing pollution and recovering resources are key goals for her utility. The most practical and cost-effective approach to urine diversion for a city such as New York, she foresees, would be off-grid systems for renovated or new buildings, supported by maintenance and collection operations. If innovators can work that out, she says, “they should go for it”.Given the advances, Larsen predicts that mass production and automation of urine-diversion technologies could be around the corner. And that would improve the business cases for this transformation in dealing with waste. Urine diversion “is the right technology”, she says. “It’s the only technology which can solve the problem of nutrients from households in a reasonable time. But people have to dare.”

    Nature 602, 202-206 (2022)
    doi: https://doi.org/10.1038/d41586-022-00338-6

    References1.Hilton, S. P., Keoleian, G. A., Daigger, G. T., Zhou, B. & Love, N. G. Environ. Sci. Technol. 55, 593–603 (2021).PubMed 
    Article 

    Google Scholar 
    2.Sutherland, C. et al. Perceptions on Emptying of Urine Diverting Dehydration Toilets. Phase 2: Post eThekwini Municipality UDDT Emptying Programme (Univ. KwaZulu-Natal, 2018).
    Google Scholar 
    3.Mkhize, N., Taylor, M., Udert, K. M., Gounden, T. G. & Buckley, C. A. J. Water Sanit. Hyg. Dev. 7, 111–120 (2017).PubMed 
    Article 

    Google Scholar 
    4.Mazzei, L., Cianci, M., Benini, S. & Ciurli, S. Angew. Chem. Int. Edn Engl. 58, 7415–7419 (2019).Article 

    Google Scholar 
    5.Noe-Hays, A., Homeyer, R. J., Davis, A. P. & Love, N. G. ACS EST Engg. https://doi.org/10.1021/acsestengg.1c00271 (2021).Article 

    Google Scholar 
    6.Clark, B. & Tarpeh, W. A. Chem. Eur. J. 26, 10099–10112 (2020).PubMed 
    Article 

    Google Scholar 
    7.Simha, P. et al. Sci. Total Environ. 765, 144438 (2021).PubMed 
    Article 

    Google Scholar 
    Download references

    Related Articles

    Toilets – what will it take to fix them?

    The secret history of ancient toilets

    The new economy of excrement

    Subjects

    Chemistry

    Developing world

    Water resources

    Latest on:

    Chemistry

    Nuclear-fusion reactor smashes energy record
    News 09 FEB 22

    Automated iterative Csp3-C bond formation
    Article 08 FEB 22

    Structure of a B12-dependent radical SAM enzyme in carbapenem biosynthesis
    Article 02 FEB 22

    Developing world

    The wise counsel that can drive public-health research in Kenya
    Career Feature 09 FEB 22

    Africa is bringing vaccine manufacturing home
    Editorial 09 FEB 22

    Female scientists can advance by saying: ‘Yes, I’ll do it’
    Career Feature 08 FEB 22

    Water resources

    ‘Sky river’ brought Iran deadly floods — but also welcome water
    Research Highlight 08 DEC 21

    Brazil is in water crisis — it needs a drought plan
    Comment 08 DEC 21

    Global potential for harvesting drinking water from air using solar energy
    Article 27 OCT 21

    Jobs

    postdoctoral fellow

    Ghent University (UGent)
    Ghent, Belgium

    Open Rank, Term Tenure Track

    The University of Texas MD Anderson Cancer Center
    Houston, TX, United States

    R464 Project Research Scientist, Cancer Metabolism

    Francis Crick Institute
    London, United Kingdom

    Postdoctoral Training Fellow

    Francis Crick Institute
    London, United Kingdom More

  • in

    RNA test detects deadly pregnancy disorder early

    NATURE PODCAST
    09 February 2022

    RNA test detects deadly pregnancy disorder early

    RNA in blood reveals signs of pre-eclampsia before symptoms occur, and the issue of urine in our sewage and what can be done about it.

    Nick Petrić Howe

    &

    Shamini Bundell

    Nick Petrić Howe

    View author publications

    You can also search for this author in PubMed
     Google Scholar

    Shamini Bundell

    View author publications

    You can also search for this author in PubMed
     Google Scholar

    Twitter

    Facebook

    Email

    Subscribe
    Subscribe

    iTunes
    Google Podcast
    acast
    RSS

    Listen to the latest from the world of science, brought to you by Nick Petrić Howe and Shamini Bundell.

    Your browser does not support the audio element.

    Download MP3

    In this episode:00:46 Predicting pre-eclampsiaCell-free RNA circulates in the blood and can give clues as to what is going on in the body. This can be used to detect disease before symptoms occur. Now researchers have analysed cell-free RNA in pregnant people and have found it can give early warning signs of a serious, and sometimes fatal, disorder of pregnancy — pre-eclampsia.Research Article: Moufarrej et al.07:19 Research HighlightsUpgrading machine vision by modelling it on human eyes, and stacked skeletons which could show attempts at repair after European tomb raiders.Research Highlight: Retina-like sensors give machines better visionResearch Highlight: ‘Spines on posts’ hint at ancient devotion to the dead09:55 The problems of peeSewage and the way it is managed can cause serious problems, for example contaminants in waste can lead to harmful algal blooms. One of the major causes of this is urine, and so some researchers have been promoting a deceptively simple solution — separate out the urine.News Feature: The urine revolution: how recycling pee could help to save the worldBook Review: Toilets – what will it take to fix them?16:40 Briefing ChatWe discuss some highlights from the Nature Briefing. This time, how China has planned to make this year’s Winter Olympics carbon neutral, and how a new radio telescope in Namibia will help us understand black holes.Nature News: China’s Winter Olympics are carbon-neutral — how?Nature News: Major African radio telescope will help to image black holesSubscribe to Nature Briefing, an unmissable daily round-up of science news, opinion and analysis free in your inbox every weekday.Never miss an episode: Subscribe to the Nature Podcast on Apple Podcasts, Google Podcasts, Spotify or your favourite podcast app. Head here for the Nature Podcast RSS feed.

    doi: https://doi.org/10.1038/d41586-022-00390-2

    Related Articles

    The urine revolution: how recycling pee could help to save the world

    Major African radio telescope will help to image black holes

    Subjects

    Biotechnology

    Water resources

    Astronomical instrumentation

    Latest on:

    Biotechnology

    Reboot biomedical R&D in the global public interest
    Comment 09 FEB 22

    Early prediction of preeclampsia in pregnancy with cell-free RNA
    Article 09 FEB 22

    Genome-edited powdery mildew resistance in wheat without growth penalties
    Article 09 FEB 22

    Water resources

    The urine revolution: how recycling pee could help to save the world
    News Feature 09 FEB 22

    ‘Sky river’ brought Iran deadly floods — but also welcome water
    Research Highlight 08 DEC 21

    Brazil is in water crisis — it needs a drought plan
    Comment 08 DEC 21

    Astronomical instrumentation

    Major African radio telescope will help to image black holes
    News 04 FEB 22

    Omicron’s origins, where’s Webb — the week in infographics
    News 01 FEB 22

    NASA’s Webb telescope is an initial success — despite formidable odds
    Editorial 25 JAN 22

    Jobs

    Research Associate (Postdoc) in AI-accelerated Signal Processing for Satellite Communications

    Interdisciplinary Centre for Security, Reliability and Trust (SnT), University of Luxembourg
    Luxembourg, Luxembourg

    Research Associate / PhD Student / Postdoc „Combinatorial Image Analysis“ (m/f/x)

    Technische Universität Dresden (TU Dresden)
    01069 Dresden, Germany

    Assistant Professor (Tenure Track)

    The University of British Columbia (UBC)
    Vancouver, Canada

    M.Sc. Student Chemical Engineering, Process Engineering, Mechanical Engineering or similar – Techno-economic analysis of a solar ammonia and fertilizer production

    German Aerospace Center (DLR)
    Köln, Germany More

  • in

    Accounting for interactions between Sustainable Development Goals is essential for water pollution control in China

    Identifying SDGs related to nutrient pollution in Chinese water systemsThe SDGs (and their targets) that are relevant to nutrient pollution in Chinese water systems are identified based on the existing literature (Supplementary Table 1) and expert judgments. The targets of the 17 SDGs are officially listed as one-sentence statements that guide SDG implementation. Based on these one-sentence statements, we identify keywords for each SDG target (Supplementary Table 5). In this way, the potential link between a target and nutrient pollution is investigated by performing a keyword search in the existing literature. The keywords for each target are compared to different keywords related to nutrient pollution, such as “nutrient pollution”, “nutrient management”, or “water quality”, to span the array of academic literature that potentially exists on the subject. Additional keywords such as “China” or “Chinese waters” are added to the query, so the literature review is made specific to the national or regional context. In the case where no specific information is found, information is then extrapolated from global studies (Supplementary Table 1). We define three levels of the relevance of targets to nutrient pollution: “high”, “moderate” and “low”. The level referred to as “high” encompasses the targets that address the direct sources of nutrient losses to the Chinese water systems. The targets identified as being of “moderate” relevance comprise factors that address the impacts of nutrient pollution on aquatic ecosystems and human health or influence the resilience of ecosystems to nutrient pollution. The targets of “low” relevance cover technological, social, administrative or economic interventions indirectly related to nutrient pollution in water systems. The targets identified as “high”, “moderate” and “low” are listed in Fig. 1 and Supplementary Table 1.Assessing synergies and tradeoffs between SDGs 6 and 14 and other SDGsThe target-level interactions between SDGs relevant to nutrient pollution in inland freshwater and coastal waters in China are assessed. More specifically, an innovative aspect of this research lies in its evaluation of the potential positive (synergetic) and negative (tradeoff) interactions existing between SDG targets in the context of nutrient pollution in Chinese water systems. These interactions are assessed based on the existing literature (Supplementary Table 1) and our expert judgments on water pollution in China using the seven-point-scale framework of Griggs et al.25, who classified the interactions at 7 levels: (−3) canceling, (−2) counteracting, (−1) constraining, (0) consistent, (+1) enabling, (+2) reinforcing, and (+3) indivisible. The negative levels refer to tradeoffs, while the positive levels refer to synergies. Zero refers to a neutral relationship between targets. The definitions of the 7 levels of interactions are summarized in Supplementary Table 2.The identified interactions have a direction, either unidirectional (one direction) or bidirectional (two directions). A unidirectional interaction means that target A affects target B, but target B does not affect target A, whereas a bidirectional interaction means that target A affects target B, and target B also affects target A. An example of a unidirectional interaction is the tradeoff between targets 2.3 and 6.4: water use in intensive agriculture to meet target 2.3, aiming to double agricultural productivity, may counteract reducing water scarcity to meet target 6.4. An example of bidirectional interaction is the synergy between targets 11.6 and 6.3: reducing water pollution in cities by improving wastewater management to meet target 11.6 is indivisible from improving water quality by halving the proportion of untreated wastewater to meet target 6.3, and vice versa. The identified interactions and their directions are illustrated in Figs. 2 and 3 and explained in Supplementary Tables 3, 4. We realize that such an assessment of the interactions can differ among experts and therefore require continuous iterations and improvements. The interactions that we identified, however, provide a primary and good basis for such continuous effort that contributes to understanding how SDGs are interrelated in the context of water pollution in China.ScenariosWe explore future (1 baseline + 5 alternative) scenarios to achieve the SDGs for improved river and coastal water quality in China using the MARINA 2.0 model. Our five alternative scenarios are developed to reduce water pollution while benefitting agriculture, sewage, food consumption, and climate mitigation by accounting for the interactions between the SDGs. We account for synergies and tradeoffs in developing these scenarios through the following steps. First, we make an inventory of the measures that are effective in reducing nutrient pollution in Chinese water systems based on existing scenario analyses52,53,54,55. Next, based on the identified SDG interactions, we identify the measures that contribute to achieving SDGs 6 and 14 as well as SDGs 2, 11, 12, and 13 simultaneously. In other words, we try to include in our scenarios only the measures that promote synergies and avoid tradeoffs between SDGs 6 and 14 and SDGs 2, 11, 12, and 13. For example, agricultural practices and technologies to improve nutrient use efficiencies are adopted in the alternative scenarios, which reduces nutrient losses to waters from agriculture for SDGs 6 and 14 while maintaining food production for SDG 2 (synergies between SDGs). Measures to control water pollution, such as reducing fertilizer use, which may result in yield losses, are thus not considered, as they can lead to challenges in achieving SDG 2 (tradeoffs between SDGs). In other words, the five alternative scenarios are developed based on measures of action promoting the synergies and mitigating the tradeoffs between key SDGs (i.e., water, agriculture, sewage, food consumption, and climate mitigation) (Supplementary Table 6). The interactions (synergies and tradeoffs) addressed by each specific assumption in the alternative scenarios are presented in Supplementary Table 7 and Supplementary Figs. 3–7.The baseline SSP5-RCP8.5 scenario assumes relatively low population growth, fast economic growth, high fossil fuel consumption, and high international trade, increasing productivity in agriculture and environmental policies for local issues16,56,57. As a result, in 2050, sewage systems will be slightly improved compared to those today. Not all wastewater will be connected to sewage systems, especially in rural areas, where only 10% of wastewater will be collected (Supplementary Table 6). Nutrient removal during treatment will remain low or moderate at ~12–47% for N and 44–75% for P in rural and urban areas (Supplementary Table 6). Crops will be produced with fewer resources (e.g., nutrients, land, and water) because of increased productivity. Animal production will be intensive and industrialized to meet the increasing preference for meat-rich diets. Improved manure management is implemented to reduce emissions of NH3 and N2O during manure storage and housing. A total of 15–41% of crop residues and 70% of animal manure will be recycled in agriculture (Supplementary Table 6). The remainder will be lost to the environment. The import of food for consumption will be 17% higher in 2050 than in 2012 (Supplementary Table 6). The greenhouse gas (GHG) emissions of China, as well as those of other countries, will be high due to high fossil fuel consumption.The SE (improved sewage treatment) scenario builds on the SSP5-RCP8.5 and assumes further improved sewage systems by 2050 based on the targets of SDG 11 “Sustainable Cities and Communities”. According to current Chinese policies, wastewater connected to sewage systems will reach 70–95% in urban areas and 60% in one-third of China’s counties, including rural and urban areas, by 2050 (Supplementary Table 6). We, therefore, assume in this scenario that by 2050, all wastewater will be connected to centralized (in urban areas) or decentralized (in rural areas) sewage systems, following Strokal et al.52 Nutrient removal during treatment is assumed to reach 80% for N and 90% for P by adopting the best treatment technologies22,52 (Supplementary Table 6). These scenario assumptions promote 9 synergies and mitigate 3 tradeoffs between SDGs for clean water (SDGs 6 and 14) and SDG 11 (Supplementary Fig. 3).The AG (improved nutrient use efficiencies in agriculture) scenario builds on the SSP5-RCP8.5 and assumes further improved nutrient use efficiencies in agriculture by 2050 based on the targets of SDG 2 “Zero Hunger”. In this scenario, crops will be fertilized according to their needs for nutrients based on a balanced fertilization approach53,54. As a result, the use of synthetic fertilizers will be largely reduced compared to the baseline, without yield loss. Recycling up to 80% of straw residues on cropland will largely reduce air pollution due to straw burning (Supplementary Table 6). Animal production will be more efficient by using improved animal feeding and genetically modified animals that use nutrients more efficiently58. In the AG scenario, N and P excretions are thus 12% lower than in the baseline SSP5-RCP8.5 (Supplementary Table 6). Improved manure management is incorporated to reduce NH3 and N2O emissions during manure storage59,60,61. In the AG scenario, the direct discharge of manure will be restricted by policies; thus, all manure is assumed to be treated and recycled on cropland. These scenario assumptions promote 8 synergies and mitigate 10 tradeoffs between SDGs for clean water (SDG 6 and 14) and SDG 2 (Supplementary Fig. 4).The AG + SE scenario combines the storylines of the SE and AG scenarios that are developed based on SDGs 2 and 11. The AG + SE scenario assumes improved sewage systems and nutrient use efficiencies in agriculture. This scenario will promote 17 synergies and mitigate 13 tradeoffs between SDGs for clean water (SDG 6 and 14) and SDGs 2 and 11 (Supplementary Fig. 5).The AG + SE + SFC (sustainable food consumption in addition to AG + SE) scenario builds on AG + SE scenario and assumes additionally healthier diets and less food waste by 2050 based on the targets of SDG 12 “Responsible Consumption and Production”. In this scenario, society will follow Chinese dietary guidelines (CDGs)62, which recommend consuming less meat and more milk, eggs, vegetables, and fruits. Food waste will be reduced by 20% through responsible consumption, improved food processing, and storage facilities55. The reduction in meat consumption and food waste will result in a 20% reduction in the requirements for crop and animal production. China may remain a large importer of soybean due to limited land resources and increasing food demand63. For soybeans, we assume that approximately 80% of the soybean consumption in 2050 will be imported from abroad, following the assumption in Ma et al.55 In addition to the above assumptions, this scenario assumes the further improved management of animal manure. In the AG + SE scenario, many river basins do not have enough arable land to recycle all the manure produced in the basin. Therefore, the AG + SE + SFC scenario assumes that the excessive manure will be either treated (as effectively as wastewater) or exported to other regions in China to be recycled. Finally, atmospheric N deposition is assumed to be reduced by 50% relative to that in the SSP5-RCP8.5 by reducing NH3 and nitrogen oxide (NOx) emissions in the agricultural and nonagricultural sectors (e.g., controlling NH3 and NOx emissions from industries). These scenario assumptions promote 42 synergies and mitigate 13 tradeoffs between SDGs for clean water (SDGs 6 and 14) and SDGs 2, 11, and 12 (Supplementary Fig. 6).The AG + SE + SFC + CLI (climate mitigation in addition to AG + SE + SFC) scenario builds on the AG + SE + SFC scenario and additionally assumes a global effort in climate mitigation by 2050 based on the targets of SDG 13 “Climate Action”. In an earlier study using the MARINA 2.0 model16, the baseline SSP5-RCP8.5 scenario assumes high GHG concentrations under higher fossil fuel consumption, which will lead to considerable climate change and thus affect hydrology (e.g., river discharge). The AG + SE + SFC + CLI scenario assumes that GHG emissions will be reduced to the level of the RCP2.6 scenario by 2050, which implies efforts by countries worldwide to reduce GHG emissions to achieve Paris Agreement temperature targets64. The lower GHG emissions in the future may result in fewer increases in precipitation and river discharge than in the baseline, thus lessening the decrease in the in-river retention of nutrients. The river export of nutrients may thus be reduced by climate mitigation in this scenario compared to the baseline. These scenario assumptions promote 56 synergies and mitigate 13 tradeoffs between SDGs for clean water (SDGs 6 and 14) and SDGs 2, 11, 12, and 13 (Supplementary Fig. 7).MARINA 2.0 modelWe use the MARINA 2.0 model16 to explore future nutrient pollution in the rivers and coastal waters of China. This model is developed to quantify the river export of TDN and TDP in four forms by rivers at the subbasin scale from different sources16. The four nutrient forms are dissolved inorganic N (DIN), dissolved organic N (DON), dissolved inorganic P (DIP), and dissolved organic P (DOP). TDN is the sum of DIN and DON, and TDP is the sum of DIP and DOP.The MARINA 2.0 model quantifies the river export of TDN and TDP as a function of N and P inputs to surface waters (rivers) from diffuse and point sources and retentions of N and P in rivers based on Eq. 1, respectively16,29:$${M}_{F.y.j}=(RSdi{f}_{F.y.j}+RSpn{t}_{F.y.j})cdot F{E}_{riv.F.outlet.j}cdot F{E}_{riv.F.mouth.j}$$
    (1)
    where MF.y.j is the river export of N and P in form F (DIN, DON, DIP, DOP) by source y from subbasin j (kg year-1). RSdifF.y.j is the N and P inputs in form F to rivers (surface waters) from diffuse sources y in subbasin j (kg year−1). RSpntF.y.j is the N and P inputs in form F to rivers from point sources y in subbasin j (kg year−1). FEriv.F.outlet.j is the fraction of N and P in form F exported to the outlet of subbasin j (0–1). FEriv.F.mouth.j is the fraction of N and P in form F exported from the outlet of subbasin j to the river mouth (0–1). The detailed equations to quantify RSdifF.y.j, RSpntF.y.j, FEriv.F.outlet.j and FEriv.F.mouth.j are available in the SI of Wang et al.16.We model nutrient pollution in the rivers and coastal waters of six large rivers in China (Supplementary Fig. 1). These rivers include the Liao, Hai, and Yellow Rivers draining into the Bohai Gulf; the Huai River draining into the Yellow Sea; the Yangtze River draining into the East China Sea; and the Pearl River draining into the South China Sea. We select these rivers because they contribute largely to nutrient pollution in the coastal waters of China. According to Wang et al.16, these six rivers contributed ~90% to the river export of TDN and TDP to the Chinese seas in 2012. The drainage basins of the Yellow, Yangtze, and Pearl Rivers are divided into upstream, middle-stream and downstream subbasins, respectively, following Wang et al.16 The names of the subbasins are available in Supplementary Fig. 2.Indicators for SDGs 6 and 14Two indicators are calculated from the MARINA 2.0 model results to assess whether SDGs 6 and 14 are met. We use water quality standards for N and P concentrations as the indicator for SDG 6 and the Indicator for Coastal Eutrophication Potential (ICEP) for SDG14. Below, we describe how these indicators are chosen based on the UN-defined indicators and how they are calculated.The goal of SDG 6 is to “ensure the availability and sustainable management of water and sanitation for all”65. One important indicator for assessing SDG 6 is the “6.3.2 proportion of bodies of water with good ambient water quality”, according to the global indicator list from the UN66. In this study, we take an indicator for “good ambient water quality” from the Chinese “Environmental Quality Standard for Surface Water”23. This standard was adopted by “China’s National Plan on Implementation of the 2030 Agenda for Sustainable Development” to achieve SDG 618. China developed this plan to translate each target of the SDGs into “action plans”, considering opportunities and challenges that it faces in implementing the 2030 Agenda. According to the Chinese “Environmental Quality Standard for Surface Water”, “third grade” (grade III) refers to good ambient water quality23. For “grade-III” water in rivers, the concentration of NH3 may not exceed 1.0 mg-N/L, and that of total P (TP) may not exceed 0.2 mg-P/L. The MARINA 2.0 model quantifies DIN (including NH3, NO3−, and NO2) and TDP but not NH3 and TP. Therefore, we calculate N and P concentrations at the outlets of subbasins using modeled DIN and TDP loads and river discharges at the outlets. We compare the calculated concentrations of DIN and TDP with the water quality standards for “grade-III” water and discuss whether our scenarios contribute to the achievement of SDG 6.The goal of SDG 14 is to “conserve and sustainably use the oceans, seas and marine resources for sustainable development”65. The UN’s global indicator list suggests “14.1.1 Index of Coastal Eutrophication” as an indicator for this SDG66. Therefore, we take ICEP as an indicator for assessing the potential of coastal eutrophication for SDG 14, as it indicates the potential for the new production of harmful algae in coastal waters. This indicator is calculated by comparing the N, P, and silica (Si) loads and the Redfield molar ratios (C:N:P:Si ratios: 106:16:1:20) (see Garnier et al.43 for the detailed approach to quantifying the ICEP). Positive ICEP values indicate relatively high potentials for harmful algal blooms when rivers deliver excess N or P over Si to the sea. Negative ICEP values indicate relatively low potentials for harmful algal blooms. We calculate the ICEP values for the six Chinese rivers using the modeled river export of TDN and TDP from the MARINA 2.0 model. Based on the results, we discuss whether our scenarios contribute to the achievement of SDG 14. More

  • in

    Storing frozen water to adapt to climate change

    1.Nüsser, M. et al. Reg. Environ. Change 19, 1327–1337 (2019).Article 

    Google Scholar 
    2.Shafiq, M. U. et al. J. Climatol. Weather Forecasting 4, 165 (2016).
    Google Scholar 
    3.Nüsser, M., Dame, J., Kraus, B., Baghel, R. & Schmidt, S. Reg. Environ. Change 19, 1327–1337 (2019).Article 

    Google Scholar 
    4.Leh Ladakh Population Census 2011 – 2021: Leh Ladakh Municipal Committee City Population Census 2011-2021 Jammu and Kashmir (Census Organization of India, 2021); https://www.census2011.co.in/data/town/800047-leh-ladakh-jammu-and-kashmir.html5.Oerlemans, J. et al. Cryosphere 15, 3007–3012 (2021).Article 

    Google Scholar 
    6.Schmidt, S. & Nüsser, M. Geosciences 7, 27 (2017).Article 

    Google Scholar 
    7.Brombierstäudl, D., Schmidt, S. & Nüsser, M. Sci. Total Environ. 780, 146604 (2021).Article 

    Google Scholar 
    8.Maheshwary, S. et al. In Proc. IASS Annual Symposium 2019 – Structural Membranes 2019: Form and Force (eds Lázaro, C. et al.) (International Association for Shell and Spatial Structures 2019).9.Tveiten, I. N. Glacier Growing – A Local Response to Water Scarcity Baltistan and Gilgit, Pakistan. MSc thesis, Norwegian Univ. of Life Science (2007).10.Faraz, S. The glacier ‘marriages’ in Pakistan’s high Himalayas. The Third Pole (3 June 2020); https://www.thethirdpole.net/en/climate/the-glacier-marriages-in-pakistans-high-himalayas/ More

  • in

    Mt. Everest’s highest glacier is a sentinel for accelerating ice loss

    MeteorologyWe reconstruct the meteorology at the South Col using observations from the automatic weather station (AWS) there (at 7945 m a.s.l)5 to downscale ERA5 reanalysis via a parsimonious blend of bias correction and machine learning. Initial screening indicates strong correlations between hourly ERA5 pressure level data bilinearly interpolated to South Col and air temperatures (r = 0.98), wind speed (r = 0.94), and relative humidity (r = 0.80) observed by the AWS. We therefore apply a simple empirical quantile mapping correction29 to remove systematic bias for these variables.Incident shortwave (SW) and longwave (LW) radiation are not available on ERA5 pressure levels, so we reconstruct them by downscaling the transmissivity (τ) and emissivity (α) of the atmosphere, defined:$$tau = {{{mathrm{SW/}}}}{Psi}$$
    (1)
    $${{{mathrm{and}}}};alpha = {{{mathrm{LW/}}}}sigma T_{{{mathrm{a}}}}^4$$
    (2)
    Where Ψ is the theoretical top of atmosphere solar radiation, σ is the Stefan Boltzmann constant (5.67 × 10−8 W m−2 K−4), and Ta is the 2 m air temperature (Kelvin). Observed values for τ and α are evaluated using AWS measurements of incident radiation and air temperature (using calculations of solar geometry to compute Ψ). We then train a Random Forest model (with 100 trees and a minimum leaf size of three) using Python’s Scikit Learn (version 0.20.1), modeling τ and α as a function of the ERA5 predictors in Methods Table 1. SW and LW could then be computed from:$${{{mathrm{SW}}}} = tau {Psi}$$
    (3)
    $${{{mathrm{LW}}}} = varepsilon T^4$$
    (4)
    Where T is the estimate from the bias-corrected ERA5 data.Table 1 Predictor variables used in the machine learning downscaling.Full size tableWe calibrate the bias correction and RF models using between 5012 (wind speed) and 12,810 (air temperature) overlapping hours of AWS observations and ERA5 data (May 2019 to December 2020). We evaluate the performance using a fivefold cross-validation, with results indicating very strong agreement between the observed and downscaled meteorology: hourly Pearson correlations range from 0.83 (relative humidity) to 0.98 (air temperature), translating respectively to root mean square errors between ~31 and 8% of the observed means (Supplementary Fig. 7). We also detect no sign of a seasonal dependence in the performance of bias correction and RF models (Supplementary Fig. 8). The resulting downscaled ERA5 data provide a complete annual series of hourly values for 1950–2019.We estimate precipitation at the South Col also using ERA5. First, we linearly interpolate the reanalysis data to the location of the Phortse AWS5 and then compute the ratio of the total observed precipitation (Po) and ERA5 precipitation (PE) during the overlapping period (April 2019-November 2020). We then multiply all reanalysis precipitation by this scalar to produce a corrected precipitation series (PE’) for Phortse 1950–2019:$$P_{{{mathrm{E}}}}^prime = frac{{P_{{{mathrm{o}}}}}}{{P_{{{mathrm{E}}}}}}P_{{{mathrm{E}}}}$$
    (5)
    To extrapolate to the South Col, we assume that precipitation decays exponentially with increasing elevation30. However, we recalibrate the regression using the Phortse and Basecamp AWSs because these new sites have weighing precipitation gauges protected by double alter shields5, and hence are less prone to under-catch error (Supplementary Fig. 9). Note that, as described below, the precipitation estimate is adjusted to an “effective” flux before being used to simulate glacier mass balance changes.Mass balanceWe use the precipitation, along with the other downscaled meteorological variables, to force the COSIPY model20 at hourly resolution for 1950–2019. First, we compute the effective precipitation (which implicitly includes the net effects of avalanching and wind transport, as well as correcting for any systematic bias in the downscaling/extrapolation method described above) required for the glacier to be in equilibrium for the period 1950–1959, iterating until the surface mass balance is zero. This is achieved when the precipitation is decreased by 65%. We then run two simulations with COSIPY. The first (referred to as the “snow” simulation) assumes a starting snowpack that is arbitrarily deep (20 m) to ensure that it remains present throughout the entire (70-year) simulation. We set the initial surface density of the snowpack to 350 kg m−3, the bottom density to 800 kg m−3, and linearly interpolate between. The second simulation (hereafter the “ice” simulation) uses the same effective precipitation but assumes no initial snowpack. However, snow is free to accumulate in the model in response to meteorological forcing. The algorithms and parameter values used in our application of COSIPY are outlined in Methods Table 2.Table 2 Parameterizations and parameter values used in the COSIPY model runs.Full size tableMeltThe surface melt rate depends on the surface energy balance (SEB):$$Q_{{{mathrm{h}}}} + Q_{{{mathrm{l}}}} + Q_{{{{mathrm{lw}}}}} + Q_{{{{mathrm{sw}}}}} + Q_{{{mathrm{g}}}} + Q_{{{mathrm{r}}}} – Q_{{{mathrm{m}}}} = 0$$
    (6)
    where Q denotes energy flux (W m−2) and the subscripts h, l, lw, sw, g, and r refer to the sensible, latent, net longwave radiative, net shortwave radiative, ground, and precipitation heat fluxes, respectively. The fluxes are defined as positive when directed towards the surface. The energy consumed in melting (Qm) is also defined as positive, meaning the melt rate (M; mm w.e. s−1 or kg s−1) can be calculated:$$M = frac{1}{{L_{{{mathrm{f}}}}}}mathop {sum}Hleft( {Q_{{{mathrm{m}}}},T_{{{mathrm{s}}}}} right)Q_i$$
    (7)
    in which H(Q,Ts) is a Heaviside function that returns a value of one unless both the sum of the first six terms in Eq. (6) (Qm = ∑Qi, with i indexing terms Qh to Qr) is positive and the surface temperature is also at the melting point; otherwise, it returns zero. The melt total over a period of ∆t seconds can then be expressed:$$M = Pfrac{{{Delta}t}}{{L_{{{mathrm{f}}}}}}{sum} {overline {Q_i} }$$
    (8)
    where P is the fraction of ∆t during which melting conditions occurred, and the overbar for energy component Qi indicates the mean value calculated during melting conditions. In terms of energy components Eq. (8) is the major driver of the amplification in melt totals between the snow and ice COSIPY simulations, increasing by a factor of 4.4; the energy sinks (sum of the remaining terms) amplify by a factor of 3.6 (Methods Fig. 3). The resulting amplification in mean melt rate (left( {acute{A} = frac{{left( {{sum} {overline {Q_i} } } right)_{{{{mathrm{ice}}}}}}}{{left( {{sum} {overline {Q_i} } } right)_{{{{mathrm{snow}}}}}}}} right)), though, is by almost a factor of 500. To understand this result, note that the proportional increase in melt rate can be written:$$acute{A} = frac{{koverline {Q_{{{{mathrm{sw}}}}}} – joverline {Q_{{{{mathrm{sinks}}}}}} }}{{overline {Q_{{{{mathrm{sw}}}}}} – overline {Q_{{{{mathrm{sinks}}}}}} }}$$
    (9)
    where (overline {Q_{{{{mathrm{sw}}}}}}) and (overline {Q_{{{{mathrm{sinks}}}}}}) are the mean energy gains and losses, respectively, during melt conditions in the snow simulation, and k and j are the proportional increases in these terms when transitioning to an ice surface (4.3 and 3.6, respectively). Critically, Eq. (9) reveals that (acute{A}) is inversely proportional to the baseline melt rate in the snow simulation (left( {overline {Q_{{{{mathrm{sw}}}}}} – overline {Q_{{{{mathrm{sinks}}}}}} } right)). The very low melt rate in the snow scenario (3.3 mm w.e. a−1), therefore acts to amplify the numerator of in Eq. (9).Sublimation and melt sensitivity to climate forcingTotal sublimation (S, mm w.e. or kg) can be written:$$S = rho ;U;C_{{{mathrm{e}}}}(e_{{{mathrm{s}}}} – e_{{{mathrm{a}}}}{Upsilon})frac{varepsilon }{{P_{{{mathrm{a}}}}}}{Delta}t$$
    (10)
    where ρ is the air density (kg m−3), U is the wind speed (m s−1), Ce is the turbulent exchange coefficient for moisture (dimensionless), ε is the ratio of gas constants for water vapor and air (0.622), Pa is air pressure (Pa), and Υ is relative humidity (fraction). The saturation vapor pressure for the surface (es) and the near-surface atmosphere (ea) are functions of the surface (Ts) and air temperature (Ta), respectively. If we assume that Ts = Ta (which is a reasonable simplification at the South Col where air temperature does not rise above 0 °C), and use the Clausius Clapeyron equation:$$e_{{{mathrm{s}}}} = e_0e^{frac{L}{{R_{{{mathrm{v}}}}}}left( { – frac{1}{{T_{{{mathrm{a}}}}}} + frac{1}{{273.15}}} right)}$$
    (11)
    in which e0 is the saturation vapor pressure at the melting point, L is the latent heat of sublimation (2.83 × 106 J kg−1), and Rv is the gas constant for moist air (461 J K−1); Eq. (10) becomes:$$S = rho ;U;C_{{{mathrm{e}}}}(1 – {Upsilon})e_0e^{frac{L}{{R_{{{mathrm{v}}}}}}left( { – frac{1}{{T_{{{mathrm{a}}}}}} + frac{1}{{273.15}}} right)}frac{varepsilon }{{P_{{{mathrm{a}}}}}}{Delta}t$$
    (12)
    which can be differentiated with respect to U, Ta, and Υ to explore the sensitivity of sublimation to changes in these meteorological parameters. In turn, the contribution of temporal trends (left( {frac{{{mathrm{d}}x}}{{{mathrm{d}}t}}} right)) in these variables to the trend sublimation can be evaluated via the chain rule:$$frac{{{mathrm{d}}S}}{{{mathrm{d}}t}} = frac{{partial S}}{{partial U}}frac{{{mathrm{d}}U}}{{{mathrm{d}}t}} + frac{{partial S}}{{partial Y}}frac{{{mathrm{d}}Y}}{{{mathrm{d}}t}} + frac{{partial S}}{{partial T_{{{mathrm{a}}}}}}frac{{{mathrm{d}}T_{{{mathrm{a}}}}}}{{{mathrm{d}}t}}$$
    (13)
    With:$$frac{{partial S}}{{partial U}} = rho ;C_{{{mathrm{e}}}};e_0left( {1 – {Upsilon}} right)e^{frac{L}{{R_{{{mathrm{v}}}}}}left( { – frac{1}{{T_{{{mathrm{a}}}}}} + frac{1}{{273.15}}} right)}frac{varepsilon }{{P_{{{mathrm{a}}}}}}{Delta}t$$
    (14)
    $$frac{{partial S}}{{partial T_{{{mathrm{a}}}}}} = L;C_{{{mathrm{e}}}};e_0;rho ;Uleft( {1 – {Upsilon}} right)frac{{e^{frac{L}{{R_{{{mathrm{v}}}}}}left( { – frac{1}{{T_{{{mathrm{a}}}}}} + frac{1}{{273.15}}} right)}}}{{R_{{{mathrm{v}}}}P_{{{mathrm{a}}}}T_{{{mathrm{a}}}}^2}}varepsilon {Delta}t$$
    (15)
    $$frac{{partial S}}{{partial {Upsilon}}} = – rho ;C_{{{mathrm{e}}}};e_0;Ue^{frac{L}{{R_{{{mathrm{v}}}}}}left( { – frac{1}{{T_{{{mathrm{a}}}}}} + frac{1}{{273.15}}} right)}frac{varepsilon }{{P_{{{mathrm{a}}}}}}{Delta}t$$
    (16)
    To evaluate Eq. (13) we compute the derivatives (Eqs. (14)–(16)) using the mean meteorology at the South Col during the ERA5 reconstruction (1950–2019), and ∆t to the number of seconds in 1 year (3.2 × 107 s). We prescribe the turbulent exchange coefficient (Ce) using the output from the COSIPY snow simulation, dividing the simulated sublimation by (rho ;U(e_{{{mathrm{s}}}} – e_{{{mathrm{a}}}}Upsilon )frac{varepsilon }{{P_{{{mathrm{a}}}}}}Delta t) (see Eq. (12)).Inserting these values into Eqs. (14) (16) yields:$$frac{{partial S}}{{dU}} = 6.0;{{{mathrm{mm}}}};{{{mathrm{w}}}}{{{mathrm{.e}}}}{{{mathrm{.}}}};{{{mathrm{a}}}}^{ – 1}{{{mathrm{m}}}}^{ – 1};{{{mathrm{s}}}}^1$$$$frac{{partial S}}{{d{Upsilon}}} = – 1.8;{{{mathrm{mm}}}};{{{mathrm{w}}}}{{{mathrm{.e}}}}{{{mathrm{.}}}};{{{mathrm{a}}}}^{ – 1}% ^{ – 1}$$$$frac{{partial S}}{{dT_{{{mathrm{a}}}}}} = 6.7,{{{mathrm{mm}}}},{{{mathrm{w}}}}{{{mathrm{.e}}}}{{{mathrm{.}}}},{{{mathrm{a}}}}^{ – 1},^circ {{{mathrm{C}}}}^{ – 1}$$We then estimate the time derivatives for Eq. (13) using the Theil-Sen slope estimation, yielding:$$frac{{{{d}}{U}}}{{{{d}}t}} = – 0.08;{mathrm{m}};{mathrm{s}}^{-1} ;{{{mathrm{decade}}}}^{ – 1};{{{mathrm{and}}}};frac{{partial S}}{{partial {U}}}frac{{{{d}}{U}}}{{{{d}}t}}; = ;0.05;{{{mathrm{mm}}}};{{{mathrm{w}}}}{{{mathrm{.e}}}}.;{{{mathrm{a}}}}^{ – 2}$$$$frac{{{{d}}{Upsilon}}}{{{{d}}t}} = – 0.6% ;{{{mathrm{decade}}}}^{ – 1};{{{mathrm{and}}}};frac{{partial S}}{{partial {Upsilon}}}frac{{{{d}}{Upsilon}}}{{{{d}}t}}; = ;0.11;{{{mathrm{mm}}}};{{{mathrm{w}}}}{{{mathrm{.e}}}}.;{{{mathrm{a}}}}^{ – 2}$$$$frac{{{{d}}T_a}}{{{{d}}t}} = 0.16;^circ {{{mathrm{C}}}};{{{mathrm{decade}}}}^{ – 1}{{{mathrm{and}}}};frac{{partial S}}{{partial T_{{{mathrm{a}}}}}}frac{{{{d}}T_{{{mathrm{a}}}}}}{{{{d}}t}}; = ;0.11;{{{mathrm{mm}}}};{{{mathrm{w}}}}{{{mathrm{.e}}}}{{{mathrm{.}}}};{{{mathrm{a}}}}^{ – 2}$$Summing these terms indicates a theoretical trend (left( {frac{{{mathrm{d}}S}}{{{mathrm{d}}t}}} right)) of 0.27 mm w.e. a−2, in reasonably close agreement with the 0.22 mm w.e. a−2 derived from the COSIPY snow simulation and reported in the main text. This theoretical analysis also indicates that 82% of the trend can be attributed to increasing air temperature (41%) and declining relative humidity (41%), with strengthening winds explaining the remaining 18%.An alternative (empirical) method to estimate the sensitivity of sublimation in the COSIPY snow simulation is to use multiple linear regression:$$S = alpha + beta _{{{mathrm{U}}}}U + beta _{{{mathrm{{Upsilon}}}}}{Upsilon} + beta _{T_{{{mathrm{a}}}}}T_{{{mathrm{a}}}}$$
    (17)
    Where the slope coefficients (βx) are linear approximations of the derivatives, relating the changes in the annual mean of the meteorological variables to the total annual sublimation. Performing the regression (Supplementary Fig. 11) lends support to the interpretation from the theoretical analysis above, attributing 49, 26, and 25% of the sublimation increase to the trends in air temperature relative humidity, and wind speed, respectively.In the main text, we highlight that sublimation and melt rates may differ in their response to climate forcing. To support this assertion, we repeat the sensitivity assessment above, evaluating the derivatives of the melt rate with respect to air temperature, wind speed, and relative humidity.We simplify the analysis by assuming that the proportion of time that the surface is melting (P) is constant (see Eq. (8)). Although physically incomplete, we note that there is no temporal trend in P for the COSIPY snow simulation (p  > 0.05 according to Seil-Then slope estimation). With this simplification, the sensitivity of the melt rate to changes in meteorological component x can then be written:$$frac{{{mathrm{d}}M}}{{{mathrm{d}}x}} = Pfrac{{{Delta}t}}{{L_{{{mathrm{f}}}}}}{sum} {frac{{partial underline {Q_i} }}{{partial x}}}$$
    (18)
    Wind speed, air temperature, and relative humidity appear in the expressions for the sensible, latent, and longwave heat fluxes (Qh, Ql, and Qlw, respectively):$$Q_h = rho ;U;c_{{{mathrm{p}}}};C_{{{mathrm{h}}}}(T_{{{mathrm{a}}}} – 273.15)$$
    (19)
    $$Q_l = rho ;U;C_{{{mathrm{e}}}};L_{{{mathrm{v}}}}({Upsilon}e_{{{mathrm{a}}}} – 611.3)frac{varepsilon }{{P_{{{mathrm{a}}}}}}$$
    (20)
    $$Q_{lw} = alpha sigma T_{{{mathrm{a}}}}^4 – 312.5$$
    (21)
    In which we have assumed melting conditions (Ts = 273.15, e0 = 611.3 Pa; Lv is the latent heat of vaporization [2.5 × 105 J kg−1], and the longwave thermal radiation emitted by the snow surface is 312.5 W m−2); cp is the specific heat content of the air (1004.7 J kg−1 K−1). It has been concluded31 that the incident longwave flux Qlw↓ in the Himalaya may be estimated from Υ and Ta:$$Q_{{{{mathrm{lw}}}}} downarrow = c_1 + c_2{Upsilon} + c_3sigma T_{{{mathrm{a}}}}^4 – 312$$
    (22)
    where the cx terms are empirically determined coefficients, whose value depends on cloudiness. Optimizing this expression for the South Col AWS, we found c1 = −17 (−168) W m−2, c2 = 0.73 (2.12) W m−2 %−1 and c3 = 0.57 (0.84) (dimensionless) for clear (cloudy) conditions.The derivative of these fluxes with respect to air temperature is then:$$frac{{partial Q_{{{mathrm{h}}}}}}{{partial T_{{{mathrm{a}}}}}} = rho ;U;c_{{{mathrm{p}}}};C_{{{mathrm{h}}}}$$
    (23)
    $$frac{{partial Q_{{{mathrm{l}}}}}}{{partial T_{{{mathrm{a}}}}}} = rho ;U;C_{{{mathrm{e}}}};L_{{{mathrm{v}}}};L;{Upsilon};varepsilon ;e_0frac{{e^{frac{L}{{R_{{{mathrm{v}}}}}}left( { – frac{1}{{T_{{{mathrm{a}}}}}} + frac{1}{{273.15}}} right)}}}{{R_{{{mathrm{v}}}}P_{{{mathrm{a}}}}T_{{{mathrm{a}}}}^2}}$$
    (24)
    $$frac{{partial Q_{{{{mathrm{lw}}}}}}}{{partial T_{{{mathrm{a}}}}}} = 4sigma c_3T_{{{mathrm{a}}}}^3$$
    (25)
    Note that all terms in Eqs. (23)–(25) are positive, outlining the physical basis for why melt rates should increase with rising air temperature32.The derivative of these fluxes with respect to relative humidity is:$$frac{{partial Q_{{{mathrm{l}}}}}}{{partial {Upsilon}}} = rho ;U;C_{{{mathrm{e}}}};L_{{{mathrm{v}}}};e_0e^{frac{L}{{R_{{{mathrm{v}}}}}}left( { – frac{1}{{T_{{{mathrm{a}}}}}} + frac{1}{{273.15}}} right)}frac{varepsilon }{{P_{{{mathrm{a}}}}}}$$
    (26)
    $$frac{{partial Q_{{{{mathrm{lw}}}}}}}{{partial {Upsilon}}} = c_2$$
    (27)
    Because all terms are positive in Eqs. (26) and (27), increases in relative humidity also drive increases in the melt rate.The derivatives of the sensible and latent heat fluxes with respect to wind speed are then:$$frac{{partial Q_{{{mathrm{h}}}}}}{{partial U}} = rho ;c_{{{mathrm{p}}}};C_{{{mathrm{h}}}}(T_{{{mathrm{a}}}} – 273.15)$$
    (28)
    $$frac{{partial Q_{{{mathrm{l}}}}}}{{partial U}} = rho ;C_{{{mathrm{e}}}};L_{{{mathrm{v}}}}frac{varepsilon }{{P_{{{mathrm{a}}}}}}(Ye_0e^{frac{L}{{R_{{{mathrm{v}}}}}}left( { – frac{1}{{T_{{{mathrm{a}}}}}} + frac{1}{{273.15}}} right)} – 611.3)$$
    (29)
    Because Ta is always less than 273.15 K during melt events at the South Col in the COSIPY snow simulation (Supplementary Fig. 12), (e_0;e^{frac{L}{{R_{{{mathrm{v}}}}}}left( { – frac{1}{{T_{{{mathrm{a}}}}}} + frac{1}{{273.15}}} right)}) must be less than 611.3 Pa. Hence, the last terms in Eqs. (28) and (29) are negative, and so increases in wind speed act to reduce the melt rate.In summary, then, theory indicates that rising air temperatures should accelerate both sublimation and melt rates (Eqs. (15) and (24)). However, increases in wind speed and relative humidity will have opposite effects. Due to the persistence of freezing air temperatures during surface melt events, faster winds act to enhance sublimation (Eq. (14)) but reduce melting (Eqs. (28) and (29)), whereas increasing relative humidity amplifies melting (Eqs. (26) and (27)) but dampens sublimation (Eq. (16)). More