In Other News…

The Arctic is getting so warm in winter that James Hansen had to add a new colour to the standard legend – pink, which is even warmer than dark red:

The official NASA maps – the ones you can generate yourself – didn’t add this new colour, though. They simply extended the range of dark red on the legend to whatever the maximum anomaly is – in some cases, as much as 11.1 C:

The legend goes up in small, smooth steps: a range of 0.3 C, 0.5 C, 1 C, 2 C. Then, suddenly, 6 or 7 C.

I’m sure this is a result of algorithms that haven’t been updated to accommodate such extreme anomalies. However, since very few people examine the legend beyond recognizing that red is warm and blue is cold, the current legend seems sort of misleading. Am I the only one who feels this way?

The Nature of Scientific Consensus

Cross-posted from NextGen Journal

It is common for one to fail to grasp the difference between “consensus” and “unanimity”.

A consensus does not require agreement from absolutely every member involved. Rather, it is a more general measure of extremely high agreement, high enough to accept and base decisions on. It’s stronger than a majority-rules style of democracy, but does not necessarily equal unanimity. In fact, in the area of science, where the concept of consensus is particularly important, unanimity is nearly impossible.

With the exception of pure mathematics, scientific theories cannot be proven beyond a doubt. Every physical process that researchers study has some amount of irreducible uncertainty – because there is always, no matter how small, a chance that our understanding could be completely wrong. Additionally, science is never “settled”, because there is always more to learn, whatever the field. Even a law as basic as gravity is still being studied by physicists, and it turns out that it gets more complicated the more you look at it.

Despite this inherent uncertainty, scientists have developed consensuses around all sorts of topics. The Earth is approximately oblate-spherical in shape. Smoking cigarettes increases one’s risk of lung cancer. HIV causes AIDS. There’s a tiny chance that these statements are incorrect, but researchers can still have confidence in their accuracy. Incomplete knowledge is not the same as no knowledge.

However, when there is room for doubt, there will usually be doubters. Physicist Richard Lindzen continues to dispute the health risks of smoking (a conversation is recounted in a recent book by James Hansen). Peter Duesberg, an active molecular and cell biologist, prominently opposes the link between HIV and AIDS. Believe it or not, the Flat Earth Society was alive and well until the death of its leader in 2001 – and signs of the society’s renewal are emerging.

As these examples suggest, for a layperson to wait for scientific unanimity before accepting a topic would be absurd. When consensus reaches a certain point, the null hypothesis shifts: the burden of proof is on the contrarians, rather than the theory’s advocates.

Another case study that may seem surprising to many is that of anthropogenic global warming. A strong scientific consensus exists that human activity, mainly the burning of fossil fuels, is exerting a warming influence on the planet’s temperature, which is already beginning to show up in the instrumental record. This phenomenon is contested by less than 3% of publishing climatologists, a negligible amount of peer-reviewed scientific studies (so few that not one showed up in a 2004 survey’s random sample of almost one thousand papers), and no major scientific societies internationally. Additionally, scientists who dispute the existence or causes of climate change tend to have lower academic credibility than those who do not. It becomes apparent that this scientific question warrants “consensus” standing: never quite settled, never quite unanimous, but certainly good enough to go by. The mainstream media does not always reflect this consensus accurately, but it nonetheless exists.

As world leaders meet in Cancun this week to discuss a global policy to prevent or limit future climate change – a prospect that looks less likely by the day – science can only offer so much advice. Climatologists can approximate what levels of emissions cuts are required to prevent unacceptable consequences, but only when the governments of the world decide which consequences they are willing to accept. Can we deal with worldwide food shortages? Rising sea levels? What about a mass extinction? Even after we define “dangerous consequences”, scientists are unsure of exactly how much temperature change will trigger these consequences, as well as how much greenhouse gas emissions will need to be cut, and how quickly, to prevent the temperature change. All they can offer is a range of probabilities and most likely scenarios.

But remember, incomplete and uncertain knowledge is not the same as no knowledge. Of one thing climate scientists are sure: the more greenhouse gas emissions we emit, the more the world will warm, and the harder it will be to deal with the consequences. There’s no reason for you and I to doubt that simple correlation any longer.

Be Critical of Critics

Cross-posted from NextGen Journal

Of all the inane arguments made against the phenomenon of anthropogenic global climate change, the strangest – in my opinion – are the conspiracy theories.

Yes, scientific fraud does happen, but on the scale of one author, not an entire multi-disciplinary field stretching back for over a century. Imagine the scale of fabrication that would be necessary for this to be true, and the amount of journal editors, expert reviewers, and students who would have to be in on the conspiracy. Scientists are just not that organized.

And for what reason would they do this? Yes, there’s the old Communist-overthrow-of-the-world argument that Fox News pundits argue would somehow result from using capitalist market strategies to put a price on carbon…but many scientists who fully accept the reality of climate change are self-proclaimed Independents and Republicans.

Even if they don’t claim out-and-out fraud, many politicians, journalists, and citizens believe that scientists’ conclusions on climate change are influenced by the lure of grant money. This position shows a fundamental misunderstanding of the way scientific grants work. As Dr. Andrew Weaver, top Canadian climatologist, argued in his fantastic book Keeping Our Cool: Canada in a Warming World, if scientists were purely interested in grant money, it would be more beneficial for them to claim uncertainty in their work. Research dollars give preference to areas of science that remain fundamentally mysterious, not those that are just ironing out the details of well-understood basic processes. Additionally, as meteorology professor Scott Mandia recently showed, the grants that scientists receive for their research don’t actually influence their salaries.

Incredibly, even the Intergovernmental Panel on Climate Change (IPCC) – a highly credible organization that painstakingly summarizes the scientific literature on climate change – is comprised of volunteer scientists. Even the chair, Rajendra Pachauri, doesn’t receive a cent for his work with the IPCC.

Of course, the prospect of a global warming fraud isn’t impossible. Nothing is. But remember, fraud is a criminal charge, and should not be thrown around lightly. Climate scientists, just like anyone else, have the right to be presumed innocent until proven guilty. They shouldn’t have to endure this endless harassment of being publicly labelled as frauds without evidence.

Here’s an example, from a retired American physics prof named Harold Lewis. He recently resigned from the American Physical Society because he didn’t think his views on climate change were being taken into account in the society’s statement. His resignation letter reads almost like satire:

It is of course, the global warming scam, with the (literally) trillions of dollars driving it, that has corrupted so many scientists, and has carried APS before it like a rogue wave. It is the greatest and most successful pseudoscientific fraud I have seen in my long life as a physicist. Anyone who has the faintest doubt that this is so should force himself to read the ClimateGate documents, which lay it bare…I don’t believe that any real physicist, nay scientist, can read that stuff without revulsion. I would almost make that revulsion a definition of the word scientist.

The scientists involved in ClimateGate, the scandal that wasn’t, have been cleared by five independent investigations to date. While some reasonable issues about data archival and sharing have been raised, absolutely no science was compromised by the contents of the stolen private correspondence. For Lewis to say otherwise and fail to provide evidence for this potentially libelous accusation is unduly irresponsible.

I disagree that the definition of scientist is “someone who feels revulsion and jumps to the conclusion of fraud from emails that show, at most, that climate scientists are not always very nice”. On the contrary, I would expect that a scientist would assess media coverage of these emails with a critical eye, examine the context in which they were written, and read the published work of the scientists in question – many of the so-called damning phrases (“hide the decline”, “lack of warming at the moment”) had already been discussed at length in the literature (Briffa et al 1997 and Trenberth et al 2009 respectively).

On a side note, where did he get “literally trillions of dollars” from? The world GDP, according to the World Bank, is approximately $61 trillion. It seems staggering to imagine that a minimum of 3% of the world economy is devoted to climate change research alone.

Let’s see what else Harry Lewis has to say:

In the interim the ClimateGate scandal broke into the news, and the machinations of the principal alarmists were revealed to the world. It was a fraud on a scale I have never seen, and I lack the words to describe its enormity. Effect on the APS position: none. None at all.

And rightly so. Even if, for the sake of argument, the CRU emails had discredited all of the research group’s data and publications, their conclusions about the current planetary warming have been independently replicated by multiple land- and satellite- based databases. In the United States alone, there is NASA GISS, NOAA NCDC, RSS, and UAH. All show the same global warming that CRU detected. Some, due to complexities in the measurement of Arctic temperatures, show even more.

The scientific literature fully supports the general premise of the APS statement on climate change: the world is warming, humans are causing it, and unless we reduce carbon emissions quickly and dramatically, it’s going to be bad. No alternative explanation for the situation has been able to withstand the scrutiny of peer-review.

Science is about looking at all sides of an issue, but it’s not a free-for-all. If someone can’t back up a claim, they don’t have an inherent right to get it published regardless. Unfortunately, in the Internet age, that doesn’t matter – if what they’re looking for is media attention, not scientific accountability.

There’s a difference between “lacking the words to describe the enormity” of a so-called fraud, and lacking the evidence to support such an accusation. As scientists (and prospective scientists, such as myself!), we need to be critical in our assessment of all claims – including the claims of critics and contrarians. Being objective isn’t always the same as being neutral.

Global Surface Temperature Change

I really enjoyed reading “Global Surface Temperature Change“, by James Hansen and his team at GISS. Keep in mind that it’s still in the draft stages – they haven’t submitted to a journal yet, but they certainly plan to, and it’s a very credible team of scientists that will almost definitely get it published.

The paper is mostly about the methods of global temperature analysis. It’s more of a review paper than an account of a single experiment. However, their main discussion point was that even by using the same data, problems can be addressed in different ways. The two main problems with temperature analysis are:

  • “incomplete spatial and temporal coverage” (sparse data)
  • “non-climatic influences on measurement station environment” (urban heat island effect).

The authors explain the methods they use and why, and explore the impacts that different methods have on their results.

GISS measures anomalies in the temperatures, largely because they are much smoother and more consistent, geographically, than absolute temperatures. In 1987, they determined that anomalies could be safely extrapolated for a radius of 1200 km from a station and still be accurate. GISS smooths the whole map out by extrapolating everything and averaging the overlapping bits.

Extrapolating is also very useful in areas with very few stations, such as the polar regions and parts of Africa. In this map, grey indicates missing data:



The Arctic is particularly problematic, not only because its data is so sparse, but also because it has the largest anomaly of any region in the world. If you have incomplete coverage of an area that is warming so dramatically, it won’t pull its full weight in the global trend, and your result will almost certainly be too low.

This difficulty with the Arctic is the reason that GISS says 2005 is the warmest year on record, while HadCRUT, the team in England, says that 1998 is. GISS extrapolates from the stations they have, and end up getting pretty good coverage of the Arctic:

They’re assuming that areas with missing data have the same anomaly as whatever temperature stations are within 1200 km, which, as they determined in 1987, is a pretty fair assumption.

However, HadCRUT doesn’t do this extrapolating thing. When they don’t have data for an area, they just leave it out:

This might sound safer, in a way, but this method also makes an assumption. It assumes that the area has the same anomaly as the global average. And as we all know, the Arctic is warming a lot more and a lot faster than the global average. So it’s quite possible that GISS is right on this one.

Another adjustment that NASA makes is for local, anthropogenic, non-climatic effects on temperature data. The most obvious of these is the urban heat island effect. As an area becomes more urban, it gets more pavement, less vegetation, and its albedo goes down – it absorbs more heat. This often makes cities substantially warmer than the surrounding rural areas, which can obviously contaminate the temperature record. However, there are ways of eliminating urban influences from the data so we can see what the real trend is.

The first step is determining what stations are considered urban. The obvious way to do this is through population, but that’s actually not very accurate. Think of somewhere like Africa, where, even if there are thousands of people living in a small area, the urban influences such as concrete, absence of vegetation, or exhaust aren’t usually present. A much better indication is energy use, and a good proxy for energy use, that’s easy to measure, is lights at night-time.

So GISS put a bit of code into their analysis that singles out stations where nightlight brightness is greater than 32 µW/m2/sr/µm, and adjusts their trends to agree with rural stations within 1200 km. If there aren’t enough rural stations within that radius, they’ll just exclude the station from the analysis.

They did an even more rigorous test for this paper, to test just how much urban influences were contaminating the long-term trend, and it was pretty interesting.

There were enough stations considered “pitch-dark” at night, where they couldn’t detect any light, to run a global analysis all by themselves. The trend that came out was <0.01 °C/century smaller than GISS’s normal calculation, an amount of error that they described as “immeasurably small”.

The result of all this temperature analysis is a graph, with one new point every year, that is “eagerly awaited by some members of the public and the media”:

However, this graph isn’t actually as useful as this one – the 12-month running mean:

“From a climate standpoint there is nothing special about the time  of year at which the calendar begins”, so instead of only measuring January-December, you can also do February-January, March-February, and so on. This way, you get a data point every month instead of every year, and more data means more accuracy. It also solves problems with short-term influences, such as El Nino, La Nina, and volcanic eruptions, that the annual graph was having. These fleeting, but fairly substantial, influences can fall completely into one calendar year or be split between two – so their influence on global temperature could be overestimated or underestimated, depending on the starting month of the calendar. The 12-month running mean is much less misleading in this fashion.

As it is, we just set a new record for the 12-month running mean, and unless La Nina really takes off, 2010 will likely set a new record for the annual graph as well. But the authors argue that we need to start moving away from the annual graph, because it isn’t as useful.

The authors also discuss public perception of climate change, and media coverage of the issue. They say, “Our comments here about communication of this climate science to the public are our opinion…[We offer it] because it seems inappropriate to ignore the vast range of claims appearing in the media and in hopes that open discussion of these matters may help people distinguish the reality of global change sooner than would otherwise be the case.”

They make the very good point that “Lay people’s perception tends to be strongly influenced by the latest local fluctuation”, and use this winter as a case study, where a strongly negative Arctic Oscillation index caused significantly cooler-than-normal conditions across the United States and Europe. Consequently, a lot of people, especially in the US, began to doubt the reality of global warming – even though, in the world as a whole, it was the second warmest winter on record:

The authors also talk about data sharing. GISS likes to make everything freely available to the public – temperature station data, computer code, everything. However, putting it out there immediately, so that anyone can help check for flaws, has “a practical disadvantage: it allows any data flaws to be interpreted and misrepresented as machinations.” Multiple times in the past few years, when there have been minor errors that didn’t actually change anything, GISS was widely accused of making these mistakes deliberately, to “intentionally exaggerate the magnitude of global warming”. They realized this wasn’t working, so they changed their system: Before releasing the data to everyone, they first put it up on a private site so that only select scientists can examine it for flaws. And, of course, this “has resulted in the criticism that GISS now “hides” their data”.

Personally, I find the range and prevalence of these accusations against scientists absolutely terrifying. Look at what has become mainstream:

Scientific fraud is a very serious allegation, and it’s one thing for citizens to make it without evidence, but it’s another thing altogether for the media to repeat such claims without first investigating their validity:

I have been disgusted by the media coverage of climate science, especially over the past year, especially in the United States, and I worry what this will mean for our ability to solve the problem.

However, there is still fantastic science going on that is absolutely fascinating and essential to our understanding of global climate change. This paper was a very interesting read, and it helped me to better understand a lot of aspects of global temperature analysis.

Deniers?

I really enjoyed New Scientist’s Special Report: Living in Denial. What a fascinating phenomenon, and a fascinating batch of articles exploring it.

The denial of science is a growing problem. It’s not restricted to a particular ideology – while denying the harmful effects of smoking or the existence of climate change is typically a position of the far right, vaccine denial and H1N1 conspiracy theories are largely restricted to the left.

It occurs even among the well-educated, or among youth who are still immersed in up-to-date curricula. For example, this year at the university, a student group put up signs saying “Don’t get the swine flu shot – it contains mercury!” The chemistry students got mad, and said that labelling thimerosal as toxic mercury was comparable to saying “Don’t eat table salt, it contains chlorine gas!”

As Michael Shermer’s article explains, the defining mark of science denial is a refusal to change one’s mind based on evidence. This is easy to identify for something like Holocaust denial, where evidence is abundant in the public sphere.

It gets a little harder for more technical issues like climate change or vaccines. Scientific opinion is overwhelmingly on one “side”, but the average person does not know or understand the evidence to support this consensus. An article about the thermodynamics of the stratosphere won’t sell a lot of papers. Most people unconsciously follow the credibility spectrum and trust what their doctor or NASA scientists say.

However, some don’t realize that scientific credibility is not the same as an appeal to authority, and so express contrarian opinions. Vaccines cause autism. Global warming is nonexistent/natural/inconsequential. The way that the Twin Towers fell proves that it was orchestrated by the US government.

There are two groups of contrarians: the skeptics, and the deniers. The skeptics are the ones who will change their minds based on evidence – they just haven’t encountered that evidence yet. My favourite example of this is from the Friends episode when Phoebe declares she doesn’t believe in evolution. When Ross starts talking to her about fossils, she says, “Oh. I didn’t know there was actually evidence.”

It’s amazing how many insights you can get out of a supposedly “fluffy” sitcom. I could write an entire essay analyzing that clip…..

I have met dozens of very reasonable people who doubt climate change because they don’t know about the evidence for it. People my age throw around the phrase “it’s a natural cycle” a lot, until I explain that the climate doesn’t act like a pendulum. It doesn’t have to compensate for past periods of warming or cooling – it simply responds to forcings. If the forcing is cyclical, then the climate will be cyclical, but some forcings are a different shape altogether. Similarly, I know a teacher who previously thought that natural causation of the current warming was a legitimate scientific theory, due to a presentation from a teacher’s conference….until I did a bit of probing and discovered that this presentation was given by Tim Ball.

These people are very reasonable. They are willing to change their minds based on evidence. They’ve just been unlucky enough to be misinformed by our flawed system of science journalism.

Then there are the deniers. They call themselves skeptics, but they will not change their minds, no matter what evidence you give them. They either move the goalposts, change the subject, or continue to repeat the same claim even after you have rebutted it patiently multiple times. Go check out some YouTube comments to see what I’m talking about.

Often their ideology or worldview is extreme in some way. For many members of the far right, any problem that would be solved by the government (think cap-and-trade or smoking legislation) will be rejected out of hand. On the far left, anything that would benefit corporations (usually vaccines or traditional medicine) will face a similar reaction. As Michael Specter says, “We hate Big Pharma. We run away from Big Pharma….and leap right into the arms of Big Placebo.”

This phenomenon suggests that science communication is not the answer – for deniers. I learned long ago that trying to change the minds of deniers is a complete waste of time. However, I still feel that science communication and the rebuttal of common misconceptions is absolutely vital. The true skeptics need access to the evidence they are lacking, so that they will be more informed, and our population will move farther towards solving the many science-related problems we face.

These skeptics deserve our time, our efforts, and our respect. They are the target audience of my blog, even if my most active commenters and supporters are a different group altogether. The reason that any of us here do all this work in communication, I believe, is for the true skeptics.

Michael Fitzpatrick argues that we shouldn’t use the label “deniers” at all. I wouldn’t want to alienate the true skeptics by coming across as someone who insults others. However, I think that calling deniers “skeptics” is unfair to the skeptics. They are two completely different groups that we must distinguish between. Skepticism is a worthy quality in science, and giving the complimentary title of “skeptic” to someone who doesn’t deserve it is unfair to those who do. We need to cater to the people who are willing to learn and who don’t want to waste our time. Science communication shouldn’t have to be like No Child Left Behind.

Michael Shermer’s second article, similarly, says that we should participate in debates with deniers and give them a chance to be heard. The truth will prevail, he argues, even if the deniers refuse to give in. I would agree with this position if it were a matter of opinion or policy, which is wholly democratic. Yet science is completely different. Science isn’t about free speech and giving equal time for all views. It is about giving time to those who have the most accurate analyses and robust conclusions. In science, you shut up and listen until your ideas are strongly supported by evidence. Then you publish.

When papers skeptical of climate change get published (all three per year!), such debates are worthy. The authors passed the test of peer-review, and even if their papers are obviously sub-par and are soon to be retracted, they deserve some debate and discussion. Let’s debate contrarian science when it is actually science – when it is actually published.

By paying close attention to and publicly debating with the authors of blog science, however, we are further confusing the public’s already skewed image of science. “It doesn’t matter whether or not you publish,” we seem to be telling them, “it’s all about free speech.” The scientific process has rules, and if deniers can’t pass the necessary, but not sufficient, condition of peer-review, their work doesn’t deserve to be treated as scientific research, and we shouldn’t give them our attention.

Let’s ignore the people who aren’t worth our time, because we have limited time, and there are people out there who deserve every minute of it.

A Better Credibility Spectrum

It’s been over a year since I wrote The Credibility Spectrum, my first post ever. Since then I’ve learned a lot, and have altered the credibility spectrum in my own mind – so I thought I’d alter it here, too.

This credibility spectrum is sort of split into two: the scientific community, and the non-scientific community. The scientific community starts with scientists, and I want to stress that this category only includes scientists with experience in the issue at hand. Just because someone has a PhD in one area of science doesn’t mean that they are an expert in all areas. For example, it’s very easy for a computer scientist to go through ten years of university without studying any biology at all.  Treating them as an expert in evolution, therefore, would be illogical.

These scientists write peer-reviewed papers, published in journals like Nature and Science, which are another step up the credibility spectrum. Instead of just having the name of an expert attached to them, their methods and conclusions have been evaluated for robustness and accuracy. This is the minimum level of credibility from which I recommend citing scientific claims.

However, as thousands of papers are published every month, and they’re generally studying the frontier of their field, it’s inevitable that some of them will be proven wrong later. As Gavin Schmidt and Michael Mann wisely said, peer review is a necessary but not sufficient condition.

That’s why there are scientific organizations and assessment reports, like NASA or the IPCC, which compile peer-reviewed knowledge which has stood the test of time into consensus statements. Even the top level of the credibility spectrum isn’t infallible, but it sure has a low error rate compared to other sources.

Everyone who isn’t a scientist, which is most of us, falls into the lower half of the credibility spectrum. The category I refer to as “communicators” includes the mainstream media, projects like Manpollo or 350.org, high school teachers, politicians…….They’re not part of the scientific community, so you should always always always check their citations, but they’re held more accountable for what they say than just any random person on the street. If they make glaring errors, people will be more upset than if the same errors were made by individuals – comments on YouTube, discussions with your neighbours – which make up the lowest rung of our credibility spectrum.

Something that I found really interesting  when I put this together was the general flow of information between different sources. In the scientific community, research starts with scientists, and the best research is published in journals, and the best journal articles are picked up by major organizations. As the scientific knowledge progresses through the different sources, the weaker assertions are weeded out along the way. The flow of information is going up the pyramid, towards the narrower part of the pyramid, so that only the best is retained.

However, in the non-scientific community, the flow of information goes the other way. Communicators present information to individuals, which is a much larger group. Information travelling down the pyramid, instead of up, allow rumours and misconceptions to flourish much more easily.

This isn’t to say that, when they come head-to-head, organizations are always right and individuals are always wrong. But given the history of such disagreements, and the levels of credibility involved, you’ll know where to place your bets.

We Have Slides!

After a marathon PowerPoint-session yesterday I finally got my 63 slides out of the way. Here is the presentation for anyone who is interested. The script is written in the notes beneath the slides.

I like to have things fading in and out of my slides, so sometimes the text boxes and images are stacked on top of each other and it won’t make sense until you view the animation.

Researching the median lethal dose of arsenic during my spare at school was really awkward. I had to do a lot of hasty explaining to my friends about how it was a metaphor for small concentrations having large effects, and no, I wasn’t planning to poison anyone.

Anyway, enjoy.

Mind the Gap (12 MB)