Megadroughts loom large

Journal name:
Date published:
Published online

Climate warming looks set to plunge the American Southwest into decades-long drought by the end of the century.

Rick Wilking/Reuters

Such 'megadroughts' have hit the region (Lake Powell on the Colorado River, pictured) during the past millennium. To calculate how changes in temperature, rainfall and soil moisture will affect the likelihood of such events, Toby Ault of Cornell University in Ithaca, New York, and his colleagues ran simulations using climate models and two greenhouse-gas emission scenarios.

If emissions continue to rise unabated, the projected increase in regional mean temperature alone will boost the risk of a megadrought to 70–99% by 2100, depending on whether precipitation increases moderately, stays the same or decreases. If warming remains below 2 °C compared to temperatures seen in the second half of the twentieth century, that risk falls to less than 66%.

Sci. Adv. 2, e1600873 (2016)


  1. Report this comment #68809

    Janet Duignan said:

    I worry a lot about which climate models they chose to use – the real ones or the ones that have been 'adjusted' to fit the global warming by carbon emission fans.
    Please have a look at:
    At least they added a caveat about lower risk if warming remains below 2°C.

  2. Report this comment #68817

    Larry Gilman said:

    Ms. Duignan:

    1) I am not a climate scientist either, but it seems to me that revising models to successfully predict observed warming (i.e., build them so that, fed data only from earlier periods, they correctly predict later states) is not only legitimate, but practically the definition of how to make an effective ? ?real? ? model. Or perhaps I misunderstand your concern. In any case, the open-source original article reported on here identifies the model ensembles it uses, and anyone can follow up its references. If truly worried, one can do the homework.

    2) The Heritage Foundation is not a scientific organization. It does ideological work, not scientific work. But computational modeling of climate physics is a scientific enterprise. We should look to scientific sources for criticism of scientific enterprises — not to ideologically packaged claims about them.

Subscribe to comments

Additional data