Rainfall’s natural variation hides climate change signal

New research from The Australian National University (ANU) and ARC Centre of Excellence for Climate System Science suggests natural rainfall variation is so great that it could take a human lifetime for significant climate signals to appear in regional or global rainfall measures.

Even exceptional droughts like those over the Murray Darling Basin (2000-2009) and California (2011 to 2017) fit within the natural variations in the long-term precipitation records, according to the statistical method used by the researchers.

This has significant implications for policymakers in the water resources, irrigation and agricultural industries.

“Our findings suggest that for most parts of the world, we won’t be able to recognize long-term or permanent changes in annual rainfall driven by climate change until they have already occurred and persisted for some time,” said  Professor Michael Roderick from the ANU Research School of Earth Sciences.

“This means those who make decisions around the construction of desalination plants or introduce new policies to conserve water resources will effectively be making these decisions blind.

“Conversely, if they wait and don’t act until the precipitation changes are recognized they will be acting too late. It puts policymakers in an invidious position.”

To get their results the researchers first tested the statistical approach on the 244-year-long observational record of precipitation at the Radcliffe Observatory in Oxford, UK. They compared rainfall changes over 30-year-intervals. They found any changes over each interval were indistinguishable from random or natural variation.

They then applied the same process to California, which has a record going back to 1895, and the Murray Darling Basin from 1901-2007. In both cases, the long dry periods seem to fit within expected variations.

Finally, they applied the process to reliable global records that extended from 1940-2009. Only 14 percent of the global landmass showed, with 90 percent confidence, increases or decreases in precipitation outside natural variation.

Professor Graham Farquhar AO also from the ANU Research School of Biology said natural variation was so large in most regions that even if climate change was affecting rainfall, it was effectively hidden in the noise.

“We know that humans have already had a measurable influence on streamflows and groundwater levels through extraction and making significant changes to the landscape,” Professor Farquhar said.

“But the natural variability of precipitation found in this paper presents policymakers with a large known unknown that has to be factored into their estimates to effectively assess our long-term water resource needs.”

The research has been published in the journal Proceedings of the National Academy of Sciences.

Read more at Phys.org

Trackback from your site.

Comments (4)

  • Avatar

    Dr Pete Sudbury

    |

    Who would be a policy-maker? Every which way but lose!
    What I’m left wondering is if anyone is looking at increasing the statistical power of the analyses. Coming from a health background, it makes me think about meta-analysis, where combining together results from large numbers of trials allows much greater certainty over the validity of conclusions.
    Taking these three examples, any one of them at one time is indistinguishable from the extremes of natural variation, but if all three of them happen at the same time, and happen alongside other extreme events (like Capture Town being on the brink of running out of drinking water), then the statistical likelihood of it being due to chance is much lower.
    There are strong suggestions that is happening. When insurers like Zurich put extreme weather events and uncontrolled climate change at two of their top 5 global risks, they are probably not shooting in the dark, but looking at actual insurance costs in real time, and factoring in confounders, like continuing to build on flood plains, which are other causes of increased risk.
    Interesting stuff.
    Have s great day!

    Reply

  • Avatar

    jerry krause

    |

    Hi Readers,

    ““We know that humans have already had a measurable influence on streamflows and groundwater levels through extraction and making significant changes to the landscape,” Professor Farquhar said.”

    I wonder if Professor Farquhar has considered what has naturally happened in history. One example is that to which Pete has observed: “When insurers like Zurich put extreme weather events and uncontrolled climate change at two of their top 5 global risks, they are probably not shooting in the dark, but looking at actual insurance costs in real time, and factoring in confounders, like continuing to build on flood plains, which are other causes of increased risk.”

    The following is from (http://champoeg.org/1861-willamette-flood/)

    “Key Concepts:

    Oregon settlers built towns along the Willamette River for economic reasons, but without understanding the river’s flood history.
    The 1861 flood destroyed Champoeg and devastated much of the rest of the valley, causing hardship that lasted for years.
    Major floods occur about every 30 years, when deep snow in the Cascade Mountains melts suddenly.

    Toward the end of November, 1861, a huge weather system stretched from California to Washington, bringing 18 straight days of rain to the Willamette Valley. When the water crested on December 6, it was as much as 30 feet deep on the Champoeg bottomland. And when the water was gone, the dismayed residents found that the town was gone too, having simply floated away.”

    Much of Oregon east of the Cascade Mountains has very low annual precipitation because it is in the rain shadow of the Cascade Mounains. But rivers have been dammed to create reservoirs to store water for irrigation instead of permitting the natural flow, when it occurs to rapidly run out to the Pacific Ocean. Hence, land which grew mainly sagebrush, is a green garden. At other locations in eastern Oregon a significant aquifer has been tapped to make another region of green which did not naturally exist.

    I suspect, but do not know, that the volume of this aquifer might naturally oscillate as some of the snow melt of the Cascade Range, which is volcanic, goes underground instead of flowing down streams and rivers. Has anyone ever heard of Crater Lake at the top of one of these volcanic mountains?

    Have a good day, Jerry

    Reply

  • Avatar

    John Nicol

    |

    It is amazing that the fact that no noticeable change in a 244 year analysis is interpreted simply as being that “the noise level is too high.

    What are these people thinking? Surely this is at the pinnacle of “Group Think”. A far more intelligent interpretation must be that there is NO CLIMATE CHANGE AT OXFORD IN 244 YEARS!!!

    Reply

    • Avatar

      jerry krause

      |

      Hi John,

      Very well put: “The noise level is too high.” At (https://www.esrl.noaa.gov/gmd/grad/surfrad/metplot.html air temperatures are measured each minute and recorded each minute. So it can be seen, particularly during the nighttime, how the air temperatures can abruptly change, several times, for a short period of time by more than a degree or even two. So, it makes the calculation of the average temperature to a hundredth of a degree to be what I would describe as stupidity or a deception. And values of radiation measurements can significantly and abruptly change for moment after moment. And I have yet to read anyone acknowledging what anyone should know is causing these abrupt changes that are often observed.

      Have a good day, Jerry

      Reply

Leave a comment

Save my name, email, and website in this browser for the next time I comment.
Share via