The Misuse Of The Scientific Method – A Nature Geoscienes Article “Changes In Hail And Flood Risk In High-Resolution Simulations Over Colorado’s Mountains” By Mahoney Et Al 2012

Recently, there were two articles in Nature Geosciences which concluded, based on multi-decadal climate model predictions, that there would not be hail in Colorado by the end of the current century. This study illustrates the misuse of the scientific method where a top-down multi-decadal climate prediction, which has shown no skill at predicting changes in hail statistics in hindcasts, is used to make forecasts decades from now.

Yet Nature Climate Change elected to publish such an untested paper. This research is another example of the epidemic of papers that purport to to be science but are actually just exercises with the models. Nature Climate Change, instead of presenting sound scientific research, apparently publishes anything that promotes their particular agenda in climate science.

The authors (and Nature Climate Change] ignore the fundamental limitations on this top-down, regional downscaled approach that we summarize in our article

Pielke Sr., R.A., and R.L. Wilby, 2011: Regional climate downscaling – what’s the point? EOS. January 31 2012 pages 52-53

The Mahoney et al Nature – Cliamte Change study is yet another example of the waste of money that I have discussed, for example, in my post

The Huge Waste Of Research Money In Providing Multi-Decadal Climate Projections For The New IPCC Report

The Mahoney etl al 2012 Nature Climate Change paper is

Kelly Mahoney, Michael A. Alexander, Gregory Thompson, Joseph J. Barsugli and James D. Scott, 2012: Changes in hail and flood risk in high-resolution simulations over Colorado’s mountains. Nature Climate Change. 10 January 2012 | DOI: 10.1038/NCLIMATE1344

with the abstract [highlight added]

“The effect of a warming climate on hailstorm frequency and intensity is largely unknown. Global climate models have too coarse resolution to simulate hailstorms explicitly; thus it is unclear if a warmer climate will change hailstorm frequency and intensity, and if so, whether such event will become more likely through intensified thunderstorms or less likely owing to overall warmer conditions. Here we investigate hail generation and maintenance for warm-season extreme precipitation events in Colorado, USA, for both present-day and projected future climates using high-resolution model simulations capable of resolving hailstorms. Most simulations indicate a near-elimination of hail at the surface in future simulations for this region, despite more intense future storms and significantly larger amounts of hail generated in-cloud. An increase in the height of the environmental melting level due to climate warming is found to be the primary reason for the disappearance of surface hail, as the warmer atmosphere increases the melting of frozen precipitation. A decrease in future surface hail at high-elevation locations may imply potential changes in both hail damage and flood risk.”

The authors conclude with the text

“For stakeholders and decision-makers trying to plan and prepare for potential changes, these results represent a step towards an improved understanding of the potential for climate change effects on extreme weather events. To optimize the utility of these findings, further research is underway to address a larger spectrum of future climate change scenarios and the sensitivity to model microphysics, and also to evaluate how findings in the in the Colorado Rocky Mountains compare to other geographic regions.”

The paper was highlighted in the journal  and discussed in the article

Walsh, Kevin, 2012: Climate science: A future Colorado without hail. Nature Climate ChangeVolume:2,Pages:78–79Year published:(2012)DOI:doi:10.1038/nclimate1396

which includes the text

“Writing in Nature Climate Change, Mahoney and co-workers present an investigation into changes in hail incidence and intensity in a hailstorm hotspot — the Colorado Rocky Mountains in the United States. They find that although it is likely that the intensity of hailstorms will increase in a warmer world, the climate may warm to a level where almost all hail formed over Colorado melts before it reaches the ground.”

“Mahoney and co-workers use a numerical atmospheric model that has a horizontal resolution fine enough to simulate thunderstorms directly, and represents precipitation processes with sufficient sophistication to produce hail. They ‘nest’ this model within a coarser resolution regional climate model, which is itself nested within a global climate model, to simulate the largest summer precipitation events in Colorado for both current (1971–2000) and possible future (2041–2070) climate conditions.”

The caveats to the study are relegated to the end of his article where Walsh writes

In all numerical modelling studies of this kind, confidence in such a prediction is improved if the model in question is able to skilfully simulate the current climate, and in this study there remains some room for improvement. The simulated freezing level in this study is higher than observed, and so all other things being equal, this would mean that their simulated hailstones would be more likely to melt in the current climate than occurs in reality. Add the additional warmth due to climate change and the stones would be even more likely to melt. Thus the prediction that summer hail will largely disappear from Colorado in the future may be overstated. To improve confidence in this result, it will need to be confirmed using other modelling systems and for other locations with similar climates.”

Walsh’s reporting the problem with their simulation of the current climate is refreshing. However, Walsh also writes that “[t]o improve confidence in this result, it will need to be confirmed using other modelling systems and for other locations with similar climates.”  He seems to ignore that you cannot test a model result with other models! Only real world observations can be used for such a test.

That the study has such a limitation in even predicting the current climate should have been a red flag to the Editor of Nature Climate Change. This article should not  have been published IF the goal of this journal is to be a credible professional resource instead of a tabloid magazine that publishes articles, regardless of scientific merit, just because they are sensational and promote their particular view of the climate science issue.

source of image

Comments Off on The Misuse Of The Scientific Method – A Nature Geoscienes Article “Changes In Hail And Flood Risk In High-Resolution Simulations Over Colorado’s Mountains” By Mahoney Et Al 2012

Filed under Climate Science Misconceptions, Climate Science Reporting, Research Papers

Comments are closed.