Thread by @MichaelEMann

I’m troubled by some very misleading statements and claims made at this public-facing @NOAA website on “Global Warming and Hurricanes”:
gfdl.noaa.gov/global-warming…
(thread)

@NOAA Among the problems are indefensible claims about the supposed non-significance of trends in hurricane activity. This includes tropical cyclone and hurricane counts and the proportion of storms that reach the highest (cat 4 and 5) categories.

@NOAA These claims are based on fitting linear trends to the associated time series.

More than a decade ago, Kerry Emanuel and I showed (meteo.psu.edu/holocene/publi…) that this approach is problematic.

@NOAA It treats externally- forced multidecadal variability (from aerosols, volcanoes, etc.) as if it is simply internal multidecadal (“AMO”) variability.

@NOAA This artificial inflation of the amplitude of the putative internal variability artificially increases the apparent “noise”, thereby reducing the signal-to-noise ratio and yielding false inferences regarding statistical significance.

@NOAA A correct approach employs a statistical model that explicitly accounts for the total forced component of variability and then isolates the residual internal variability component.

@NOAA Doing so yields a highly significant statistical relationship between rising temperatures and measures of tropical cyclone activity.

@NOAA Using sound statistical methods, hurricane expert @JBElsner and collaborators have demonstrated a very clear, statistically-significant relationship between sea surface temperature and the intensity of the strongest storms (nature.com/articles/natur…).

@NOAA @JBElsner They find a roughly 7% increase in maximum hurricane intensity for each 1C increase in sea surface temperature. Since destructive potential goes as the 3rd power of maximum wind speed, that corresponds to a roughly 23% increase in destructive potential.

@NOAA @JBElsner The tropical North Atlantic has warmed by about 1C.

@NOAA @JBElsner The idea that an internal “AMO” variability component can simply be defined by the residual variability around a linear trend in SST time series has been shown to be erroneous now in numerous publications.

@NOAA @JBElsner See:
agupubs.onlinelibrary.wiley.com/doi/abs/10.102…
agupubs.onlinelibrary.wiley.com/doi/full/10.10…
agupubs.onlinelibrary.wiley.com/doi/full/10.10…
science.sciencemag.org/content/347/62…
journals.ametsoc.org/doi/full/10.11…

@NOAA @JBElsner And yet, frustratingly, the NOAA webpage in question (gfdl.noaa.gov/global-warming…) is replete with results based on precisely such a procedure.

@NOAA @JBElsner Finally, the claim on that site that models show a decrease in the total number of tropical cyclones globally is wrong.

@NOAA @JBElsner Using a downscaling approach combined with state-of-the-art climate model simulations, leading researcher Kerry Emanuel finds an increase in both frequency and intensity for tropical cyclones globally, and in the North Atlantic specifically (pnas.org/content/110/30…).

@NOAA @JBElsner The modeling approach used by the GFDL group is the primary basis for the claim of a decrease in frequency. Many of us long suspected that was an artifact of too-low model resolution, which fails to resolve the eye wall dynamics of storms.

@NOAA @JBElsner Indeed, these researchers have recently conceded that the decrease in frequency evaporates when adequate model resolution is used:

https://twitter.com/MichaelEMann/status/1162076312992964609

@NOAA @JBElsner Using a downscaling approach combined with state-of-the-art climate model simulations, leading researcher Kerry Emanuel finds an increase in both frequency and intensity for tropical cyclones globally, and in the North Atlantic specifically (pnas.org/content/110/30…).

@NOAA @JBElsner The modeling approach used by the GFDL group is the primary basis for the claim of a decrease in frequency. Many of us long suspected that was an artifact of too-low model resolution, which fails to resolve the eye wall dynamics of storms.

@NOAA @JBElsner Indeed, these researchers have recently conceded that the decrease in frequency evaporates when adequate model resolution is used:

https://twitter.com/MichaelEMann/status/1162076312992964609

@NOAA @JBElsner The idea that an internal “AMO” variability component can simply be defined by the residual variability around a linear trend in SST time series has been shown to be erroneous now in numerous publications:

https://twitter.com/MichaelEMann/status/1170142983615930371

@NOAA @JBElsner

https://twitter.com/MichaelEMann/status/1170150920358940674

@NOAA @JBElsner

https://twitter.com/MichaelEMann/status/1170150920358940674

@NOAA Unsurprisingly, disinformers who have been exploiting this misinformation are upset I've drawn attention to it:
Ryan Maue (desmog.co.uk/2017/08/14/cli…)
Judith Curry (sourcewatch.org/index.php/Judi…)
Roger Pielke Jr (huffpost.com/entry/fivethir…)