Sunday, April 20, 2014

Is Asian Pollution Intensifying Pacific Storms? Separating the Hype from Reality.

The media over the past week have given a lot of play to a paper in the journal Proceedings of the National Academy of Sciences (PNAS)  that claims that Asian pollution is enhancing storms over the Pacific Ocean, with effects extending globally.

We are talking headlines in hundreds of major media sources around the world.  As illustrated below, the headlines have been pretty scary.  But as I will show below, the PNAS paper really proves no such thing and this situation is another example of unhealthy and counterproductive hype and exaggeration in the media.






The paper, Assessing the Effects of Anthropogenic Aerosols on Pacific Storm Track Using a Multitscale Global Climate Model (found here) is by Yuan Wang of Texas A&M and collaborators, with the last author being Mario Molina, a member of the National Academy of Sciences (keep this in mind, it will be important later).

The paper claims particles ejected into the atmosphere over Asia (aerosols) are causing "intensification of the Pacific storm track" and "invigorated midlatitude cyclones."   In a press release by Texas A&M, co-author Renyi Zhang states:

“There appears to be little doubt that these particles from Asia affect storms sweeping across the Pacific and subsequently the weather patterns in North America and the rest of the world,” Zhang says of the findings.
“The climate model is quite clear on this point. The  aerosols formed by human activities from fast-growing Asian economies do impact storm formation and global air circulation downstream.  They tend to make storms deeper and stronger and more intense, and these storms also have more precipitation in them. "
And Tech Times states:
Yuan Wang believes the polar vortex that brought frigid temperatures to much of the eastern United States may have been the result of Pacific storms driven by air particles released over China and other nations .    

In a press release Wang stated:

"This cold winter in the U.S. probably had something to do with stronger cyclones over the Pacific,"
So Asians keeping themselves warm is causing cold waves over the eastern U.S. according to an author of this paper.  As we shall see below, such claims are entirely unwarranted.


A little analysis quickly reveals that there is far less  than meets the eye to this paper.  And that there is very little proof that Asian pollution is doing anything to Pacific storms.
Let me begin by noting that the Wang et al. study is limited to modeling.  There is no observational analysis showing that observed storms are intensifying.  No observations showing direct effects of Asian pollution on storms.

This study is limited to looking at two global model simulations:  one assuming the aerosol (particle) conditions of 1850 (preindustrial, PI) and the other the pollution characteristics of 2000 (present day, PD).  So they did not show using observations that storms are getting stronger or that there was a connection with increased aerosols.

What they did do was to run a global climate model twice with an without modern aerosols in the atmosphere.   This study really doesn't isolate Asian aerosols changes as suggested by the text and the resolution is so coarse (around 200- 250 km between the grid points) that they really can't simulate Pacific storms properly in any case.    Their innovation is to run a different approach for simulating convection (e.g., thunderstorms), using explicit simulations, with cloud and aerosol physics modules, rather than what is called parameterized physics (where processes are inferred from larger scale variables).  They never show that their approach does a better job in duplicating observed clouds and storms than more traditional approaches.

Now the fun part.   They describe the results of huge changes in aerosols (again between 1850 and 2000).  But they never show how the winds are changing in the storms.  Or how the central pressures are changing.  Rather they show very indirect parameters.  Kind of strange that they took this approach.   

They show that precipitation over the NW Pacific increases by 2.4%!   And that high clouds increase by 2.6% .   So a huge change in particles over Asia only increases precipitation and clouds by around 2.5% and ON THIS BASIS they are suggesting noticeable strengthening of Pacific storms.  I could make a snide remark on how ridiculous this is, but I won't.   And they found that north-south heat transport over the NW Pacific increased by 5% for this 150-year aerosol change.  Not much.  And for most of their domain (basically China to the dateline) the differences between 1850 and 2000 were NOT STATISTICALLY SIGNIFICANT.


What they should have noted was how AMAZINGLY INSENSITIVE the Pacific storm track is to increased particles in the atmosphere.  

Next, they compare their simulations to a similar experiment using a more traditional (CAM5) climate model. Never do they show that their model is better by the way.  Complexity does not necessarily produce better results.   Anyway, the CAM5 model gives the opposite results, with the storm precipitation and high clouds decreasing.   Quite honestly, this makes just as much sense as their results with more a complex model.   With reflective aerosols pushing out from China and India around 30-40N, that might cause cooling in the latitudinal band, weakening the north-south temperatures gradient, and thus attenuating storms (midlatitude storms depend on north-south gradient for their energy).

And there is something really strange about this study.  They ran the model globally but never show the influence over the eastern Pacific and North America.   Why not?  Why make all kinds of inferences and suggestions about the downstream effects, but don't show the directly relevant information that you have.  

The question you all should be asking at this point is:  how in the world did this paper get published in a supposedly peer reviewed journal?  If I had been a reviewer, I would have immediately rejected it for the reasons noted above and many more.

And there is a dirty little secret here: papers published in PNAS are often...how do I say itpolitely?... "lightly reviewed" , particularly when a member of the National Academy is an author.   Until recently, a member of the Academy could simply wave any paper through.  After some complaints about poor manuscripts they developed an alternative means where the Academy member can serve as editor and evaluate the reviews him/herself.  You notice a member of the Academy was involved with the paper and it is clear that the paper did not undergo the "standard", more rigorous, review process.  And only two reviewers, which is less than usual for most major papers.


What I find particularly disturbing is that  this paper is just one in a series of similarly speculative, generally weak, manuscripts that hype some environmental threat from global warming or some other environmental threat. The media, does not seriously review the work, and goes bonkers over it, with the latest threat becoming headline material all over the world.   There are so many examples of this. For example:


  • That Global Warming is causing the jet to "wobble" producing more cold waves in the eastern U.S.
  • This week this was another paper of this ilk claiming that Global Warming is causing a "dipole pattern" over North America with ridging in the west (think drought) and troughing in the east (think polar vortex).  Future blog material!


Global warming is a serious issue for mankind and rational discussion of its impacts is being impeded by the huge media hype for weak,speculative papers, most of which will be proven to be in error.  And advocacy groups concerned with global warming are not helping their cause by incessantly crying wolf.  The world will eventually stop listening.  Many folks already have.


9 comments:

RickG said...

Serious matters are never well served by weak science. It's a shame that this happens. The news writers don't have resources, time or inclination to dig in to the details to verify claims. Especially when the claims will make sensational headlines. And the article is "peer reviewed," right? So it must be valid, right? Aaaaaggghhhh!

Thanks, Cliff, for you efforts to sort things out. Would that you had a broader audience.

Fixed Carbon said...

Serious science about climate change is not served by a bias against preliminary results when and only when they contradict your hunches. Bias against false positives carries the risk of false negatives. If one asserts over and over again, that "global warming is coming, but not yet", one needs to show the science about "when" and what you will accept as evidence that the time has come. Cliff what is your model of "when," and how will we know that the time has come?

Unknown said...

Why did they stop at 2000, when the pollution has gotten steadily worse the past decade or so? Seems they could gone up to 2010 at least.

-Douglas

Cliff Mass said...

Fixed Carbon,
An interesting question. First, the media needs to get beyond the press releases and abstracts, and get opinions from other experts before they go into headline mode. If they had done so with this paper, they would have realized how little meat there was on the bone. Secondly, it is good to give the scientific process time to work (reactions, comments, alternative papers) by the scientific community. Typical time scale: 1-2 years considering the review process. A good example of this is the snow pack controversy here in the NW. There were claims in a few papers of HUGE declines. Big headlines resulted. A few years later there were alternative papers that showed the problems in that work. The same thing is happening with the GW causes polar vortex scare. The same thing happened with the midwest heat wave is a sign of global warming scare. There are a lot of questionable results in the peer reviewed media, but eventually the systems works and the truth comes out. But the media has no patience and headlines some of the most outrageous claims....cliff

Loco Geologo said...

"...did not show using observations that storms are getting stronger or that there was a connection with increased aerosols."

As a hydrogeologist who uses numerical models for groundwater use and chemical transport; not tying in an observable connection makes their model no more than a complicated toy.

In groundwater a modeling report must be complete in that the model can be regenerated by any peer, as well being calibrated to observable data. Is this not possible for atmospheric models?

Placeholder said...

Outstanding item. I hope you will write about the two bullet-pointed topics you mentioned. I've seen commentary about them elsewhere, and think the debate would improve with your input.

sciencefirst said...

Thank you, Cliff, for another well-argued, sober analysis. Your sounding the alarm about prematurely published, supposedly peer-reviewed papers and the subsequent media hype is so important! You will take some flak for this, particularly for revealing the dirty little secret about how papers published in PNAS are "reviewed." Rest assured that those of us who are aware of the problem and are interested in serious science are very grateful to you for taking such a courageous stand.

OldTech said...

"There are a lot of questionable results in the peer reviewed media, but eventually the systems works and the truth comes out."

I hope you are right, but unfortunately I am seeing other scientific areas where established biases filter out the questions.

It appears to me that confirmation bias is rampant in many areas, including climate science. This has gone so far that the believers now label anyone who questions the science a skeptic.

What I find particularly distasteful is when the scientist themselves start thinking in terms of believers and skeptics. These are religious terms not scientific terms.

caveat emptor said...

Hyping the science is unfortunate, but understandable. Scientists naturally think whatever they are working on is important and exciting. One doesn't have to look far for examples. For instance I recall a recent blog post here presenting the revolutionary findings that diesel trains emit diesel particulate matter.

As far as the paper I agree with some, but not all, of Dr Mass's criticisms.

I certainly would have liked to see the discussion section of the paper include some caveats and discussion of the limitations of the study (not just the strengths). That is a pretty glaring omission IMO.

OTOH the fact that the paper is exclusively modelling is not really a valid criticism. Lots of "modelling only" studies are published. Results that emerge from modelling can give hints of things to look for in the observations.

Criticizing it for being "speculative" is a bit sad. I hope scientists never stop "speculating". And including some of that speculation in the "discussion" section of a paper is perfectly valid too.

As far as science by press release, that is the unfortunate reality of how our media works. It is not unique to meteorological/climatological science. If anything it is worse in the medical and social sciences.

Bottom line is that this paper will stimulate more work, which will support or perhaps rebut the author's findings and isn't that kind of how science works? Of course the media will never report on all that "boring" follow up work. Again the unfortunate reality of our media.