Did you notice the recent anti-IPCC article?

You may have missed the latest attack on the IPCC, because the mitigation sceptics did not celebrated it. Normally they like to claim that the job of scientists is to write IPCC friendly articles. Maybe because that is the world they know, that is how their think tanks function, that is what they would be willing to do for their political movement. The claim is naturally wrong and it illustrates that they are either willing to lie for their movement or do not have a clue how science works.

It is the job of a scientist to understand the world better and thus to change the way we currently see the world. It is the fun of being a scientist to challenge old ideas.

The case in point last week was naturally the new NOAA assessment of the global mean temperature trend (Karl et al., 2015). The new assessment only produced minimal changes, but NOAA made that interesting by claiming the IPCC was wrong about the "hiatus". The abstract boldly states:

Here we present an updated global surface temperature analysis that reveals that global trends are higher than reported by the IPCC ...

The introduction starts:

The Intergovernmental Panel on Climate Change (IPCC) Fifth Assessment Report concluded that the global surface temperature “has shown a much smaller increasing linear trend over the past 15 years [1998-2012] than over the past 30 to 60 years.” ... We address all three of these [changes in the observation methods], none of which were included in our previous analysis used in the IPCC report.

Later Karl et al. write, that they are better than the IPCC:

These analyses have surmised that incomplete Arctic coverage also affects the trends from our analysis as reported by IPCC. We address this issue as well.

To stress the controversy they explicitly use the IPCC periods:

Our analysis also suggests that short- and long-term warming rates are far more similar than previously estimated in IPCC. The difference between the trends in two periods used in IPCC (1998-2012 and 1951-2012) is an illustrative metric: the trends for these two periods in the new analysis differ by 0.043°C/dec compared to 0.078°C/dec in the old analysis reported by IPCC.

The final punchline goes:

Indeed, based on our new analysis, the IPCC’s statement of two years ago – that the global surface temperature “has shown a much smaller increasing linear trend over the past 15 years than over the past 30 to 60 years” – is no longer valid.

And they make the IPCC periods visually stand out in their main figure.


Figure from Karl et al. (2015) showing the trend difference for the old and new assessment over a number of periods, the IPCC periods and their own. The circles are the old dataset, the squares the new one and the triangles depict the new data with interpolation of the Arctic datagap.

This is a clear example of scientists attacking the orthodoxy because it is done so blatantly. Normally scientific articles do this more subtly, which has the disadvantage that the public does not notice it happening. Normally scientists would mention the old work casually, often the expect their colleagues to know which specific studies are (partially) criticized. Maybe NOAA found it easier to use this language this time because they did not write about a specific colleague, but about a group and a strong group.

Figure SPM.1. (a) Observed global mean combined land and ocean surface temperature anomalies, from 1850 to 2012 from three data sets. Top panel: annual mean values. Bottom panel: decadal mean values including the estimate of uncertainty for one dataset (black). Anomalies are relative to the mean of 1961−1990. (b) Map of the observed surface temperature change from 1901 to 2012 derived from temperature trends determined by linear regression from one dataset (orange line in panel a).

The attack is also somewhat unfair. The IPCC clearly stated that it not a good idea to focus on such short periods:

In addition to robust multi-decadal warming, global mean surface temperature exhibits substantial decadal and interannual variability (see Figure SPM.1). Due to natural variability, trends based on short records are very sensitive to the beginning and end dates and do not in general reflect long-term climate trends. As one example, the rate of warming over the past 15 years (1998–2012; 0.05 [–0.05 to 0.15] °C per decade), which begins with a strong El Niño, is smaller than the rate calculated since 1951 (1951–2012; 0.12 [0.08 to 0.14] °C per decade)

What the IPCC missed in this case is that the problem goes beyond natural variability, that another problem is whether the data quality is high enough to talk about such subtle variations.

The mitigation sceptics may have missed that NOAA attacked the IPCC consensus because the article also attacked the one thing they somehow hold dear: the "hiatus".

I must admit that I originally thought that the emphasis the mitigation sceptics put on the "hiatus" was because they mainly value annoying "greenies" and what better way to do so than to give your most ridiculous argument. Ignore the temperature rise over the last century, start your "hiatus" in a hot super El Nino year and stupidly claim that global warming has stopped.

But they really cling to it, they already wrote well over a dozen NOAA protest posts at WUWT, an important blog of the mitigation sceptical movement. The Daily Kos even wrote: "climate denier heads exploded all over the internet."

This "hiatus" fad provided Karl et al. (2015) the public interest — or interdisciplinary relevance as these journals call that — and made it a Science paper. Without the weird climate "debate", it would have been an article for a good climate journal. Without challenging the orthodoxy, it would have been an article for a simple data journal.

Let me close this post with a video of Richard Alley explaining even more enthusiastic than usually
what drives (climate) scientists? Hint: it ain't parroting the IPCC. (Even if their reports are very helpful.)

Suppose Einstein had stood up and said, I have worked very hard and I have discovered that Newton is right and I have nothing to add. Would anyone ever know who Einstein was?

Tweet

Further reading

My draft was already written before I noticed that at Real Climate Stefan Rahmstorf had written: Debate in the noise.

My previous post on the NOAA assessment asked the question whether the data is good enough to see something like a "hiatus" and stressed the need to climate data sharing and building up a global reference network. It was frivolously called: No! Ah! Part II. The return of the uncertainty monster.

Zeke Hausfather: Whither the pause? NOAA reports no recent slowdown in warming. This post provides a comprehensive, well-readable (I think) overview of the NOAA article.

How climatology treats sceptics. My experience fits to what you would expect.

References

IPCC, 2013: Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 1535 pp, doi: 10.1017/CBO9781107415324.

Thomas R. Karl, Anthony Arguez, Boyin Huang, Jay H. Lawrimore, James R. McMahon, Matthew J. Menne, Thomas C. Peterson, Russell S. Vose, Huai-Min Zhang, 2015: Possible artifacts of data biases in the recent global surface warming hiatus. Science. doi: 10.1126/science.aaa5632.

Boyin Huang, Viva F. Banzon, Eric Freeman, Jay Lawrimore, Wei Liu, Thomas C. Peterson, Thomas M. Smith, Peter W. Thorne, Scott D. Woodruff, and Huai-Min Zhang, 2015: Extended Reconstructed Sea Surface Temperature Version 4 (ERSST.v4). Part I: Upgrades and Intercomparisons. Journal Climate, 28, pp. 911–930, doi: 10.1175/JCLI-D-14-00006.1.

Rennie, Jared, Jay Lawrimore, Byron Gleason, Peter Thorne, Colin Morice, Matthew Menne, Claude Williams, Waldenio Gambi de Almeida, John Christy, Meaghan Flannery, Masahito Ishihara, Kenji Kamiguchi, Abert Klein Tank, Albert Mhanda, David Lister, Vyacheslav Razuvaev, Madeleine Renom, Matilde Rusticucci, Jeremy Tandy, Steven Worley, Victor Venema, William Angel, Manola Brunet, Bob Dattore, Howard Diamond, Matthew Lazzara, Frank Le Blancq, Juerg Luterbacher, Hermann Maechel, Jayashree Revadekar, Russell Vose, Xungang Yin, 2014: The International Surface Temperature Initiative global land surface databank: monthly temperature data version 1 release description and methods. Geoscience Data Journal, 1, pp. 75–102, doi: 10.1002/gdj3.8.

Climatology is a mature field

Apparently I wrote something controversial in a comment at Judith Curry's place (Climate Etc.).

Judith Curry wrote:

The point is that you can’t neutralize plausible alternative interpretations of the available evidence from diminishing the scientific ‘consensus.’ It only takes one such argument, and one person making it (but in fact there are numerous arguments and a substantial number of people making them).

To which I replied:

In practice it will likely take more than one Galileo or study. Also the refutation of classical mechanics by quantum mechanics and relativity did not change many things we already understood at the time. It allowed us to study new things and ask new questions. That was the revolution.

Climatology is a mature field and new findings will more likely change the complete picture only little. The largest uncertainties are in the impacts, improving our understanding there will have to be done one impact at a time. And more likely, one aspect of an impact at a time.

It is a pity that people did not respond to my claim that large scientific changes (paradigm changes) tend to enlarge the scope of science, rather than to invalidate more practical previous findings. The claim that climatology is mature, however, was too much for many.

Judith Curry returned:

1. Climate science is NOT a mature field. Stay tuned for more and more surprises . . .

Hard to answer such a argumentative heavy weight comment.

There are also some sensible people there, for example Michael answered Curry:

The second [surprises] is not in any way ruled out by the first [maturity].
There are plenty of new discoveries, even in mature fields.

It it interesting that Curry welcomes surprises and uncertainty so much. The surprises are what worry me the most. I guess that is my conservative side. People have worried about climate change through the ages because the climate is so important to us. Now we are messing with the climate as if we know exactly what we are doing and it's gonna be great.

Fitting to the level of the "debate" in the comments at Climate Etc., dlb wrote:

Perhaps he meant a manure field?
Glad to know I wasn’t the only one whose jaw dropped after reading that.

Most others also just expressed their disbelieve. What comes closest to an argument is: scientists (economists, politicians) have been wrong before, thus climate science is not mature. Okay, maybe I am too generous calling that almost an argument. I hope I did not miss any arguments, the "discussion" was somewhat derailed by people arguing against the greenhouse effect. Also fitting to the level of the comments at Climate Etc. I have no idea how an improvement in scientific understanding should result from that blog.

What makes a scientific field mature?

So what is a "manure" scientific field? Three important aspects are probably: mass, time and networking.

But let me first explain what maturity is not. Calling a field mature does not mean that no discoveries will be made any more, it also does not mean that predictions are perfect and confidence levels are zero. Science is not religion, if you cannot handle uncertainty, you should not be debating science.

Let's contrast a mature with a young science. For example the beginning of discovery of the greenhouse effect. From Wikipedia: "The existence of the greenhouse effect was argued for by Joseph Fourier in 1824. The argument and the evidence was further strengthened by Claude Pouillet in 1827 and 1838, and reasoned from experimental observations by John Tyndall in 1859, and more fully quantified by Svante Arrhenius in 1896." These single scientists could easily have been completely wrong. The years indicate that they did not communicate with each other about the topic. There was thus no competition for who was smarter. At that time it could have happened that someone found that the arguments of Fourier were nice, but the observation do not show the effect. That the measurements of Tyndall were carefully performed, but his instrument does not measure what he thought it did. That there is a greenhouse effect, but contrary to the findings of Arrhenius CO2 is completely irrelevant. And at the time, it would have been possible that negative feedbacks are so strong, that any additional radiative forcing by CO2 does not influence the surface temperature. And a Galileo might have found a stupid calculation error in the works of any of these scientists.

Mass. There are thousands of climate researchers and in I just found 112,598 articles on the Web of Science about "global warming" or "climate change". These article will not all be studying climate change, but a large part will. That alone makes a stupid calculation errors nearly impossible. Having multiple people working on the same topic also provides a sparring partner to discuss and compete with. The weight of the evidence is huge, not comparable to previous times the media or scientists called alarm or called for more research.

Networking. Mass allows some scientists to specialize. For example, some people work on radiative transfer for most of their career. They study the importance of various assumptions, try out various methods of solving a problem, solve various problems with radiative transfer (greenhouse effect, passive and active remote sensing) write textbooks, collaborate with scientists working on radiative transfer in other fields, build joint validation programs for radiative transfer codes, and so on. A specialist is less likely to make mistakes as a newby, the networking weaves a scientific field into the the complete network of scientific theories, methods and tools. (Nothing against newbies, a growing science will have lots of them and they can help with fresh ideas.)

If the physical basis of climate change were found to be wrong, this would likely affect many other sciences via this network. Anything is theoretically possible, but this sure makes it much less likely. The first thing many climate change dissenters learn about is the greenhouse effect itself. Unfortunately many get stuck with this very well connected theory. It is much more likely to find problems with aspects that are specific to climatology, the climatological response of the oceans, vegetation and clouds, for example. What will happen with extreme weather, and all the various impacts? These are difficult questions, but they are not questions that are solved by one Galileo paper. If you want to be that guy, please have a look at such topics, not the greenhouse effect and make the "debate" at least a little more intelligent. Even better, take an objective look at these topics, that increases the chance you will contribute to our understanding of climate change.

Time. Time is important first of all as the amount of time a specialist spends on a topic. Secondly, science takes place in the scientific literature. Performing a study, writing it up and getting it published can easily take a year or longer. While the main ideas might be known from conferences, an idea can only be fully tested after publication. And the response again easily needs a year. Thus science takes time. Also creativity takes time. The longer no new idea turns out to change the main picture much, the less likely it becomes that that happens. And also for creativity mass is important, more people have more funky ideas.

How does a mature scientific field work?

Discoveries. That does not mean that no new discoveries are made. In my own field homogenization we have made spectacular progress the last decade. Modern homogenization methods are now twice as good as traditional ones. Better methods have increased scientific confidence that the temperature trend is robust, but did not change the trend much. (This may be different for daily data, used to study changes in extreme weather, where non-climatic changes are expected to be more important.)

While the homogenization of the annual means is a "manure" science, the homogenization of daily data is in its infancy. Some of my colleagues object to me calling annual homogenization mature, because they still have many ideas to improve it. I hope that after this post, people will understand that I do not see that as a contradiction. Science is never finished, but that does not mean that we know nothing.

Bias. In the beginning, when there is just one person or a few groups working on a problem, they may be tempted to exaggerate the importance of their problem. Most scientists are rather conservative when it comes to making strong claims, but scientists are also humans and some may be tempted by external incentives, although disingenuous pseudo-skeptics like to exaggerate this problem. Calling something a problem helps attracting more people and funding. Exaggeration is relatively easy in this stage as the uncertainties are high and are high due to ignorance.

When a field gets larger, every speciality has an incentive to exaggerate the importance of their speciality. A solar physicist is tempted to claim it is the sun, but hindered by the evidence. I went into science to understand the world a little better and maybe to show off my skills. Were I driven by monetary incentives, as a naive economist might expect, I would be tempted to claim large uncertainties due to non-climatic changes, that would make my field more important. Even if Anthony Watts thinks otherwise, it would be bad for my career to claim that climate data is fine. Making unscientific claims WUWT-style would hurt my career even more and take the fun out of being a scientist.

Because of the incentives of the solar physicist to claim that global warming is due to the sun and my incentive to claim it is non-climatic, the chance of a bias in the big picture is much reduced for a "manure" science. Our understanding will keep on improving and estimates will change, but at this stage I would no longer expect any biases in the basic science: the changes will go in any direction.

Judith Curry likes the word uncertainty. Towards other scientists she can claim that she intended it the way science uses the word: we do not know the exact value, it lies in a confidence range. It could be higher, it could be lower. To her audience that sounds like: they know nothing, they are not sure about climate change, maybe there is no problem after all. Curry's audience hears bias! And is shocked that someone dares to call climatology a mature science.

Let me close with an interesting tweet on the topic of terms used differently inside and outside the scientific world. (The figure comes from Communicating the science of climate change by Richard Somerville and Joy Hassol. h/t Lars Karlsson)

Imp.of inverted triangle for sci.comm - What scientists say & what public hears. http://t.co/XoypoQvr95 via@LauRob85 pic.twitter.com/jdt0JWzKWd
— Alix Cage (@alixcage) April 29, 2014

* Some of the comments have been edited for readability.
** Tip: do not search for "mature" Flickr images.

Falsifiable and falsification in science

"You keep using that word. I do not think it means what you think it means."

In a recent post, Interesting what the interesting Judith Curry finds interesting, I stated that "it is very easy to falsify the theory of global warming by greenhouse gasses." The ensuing discussions suggest that it could be interesting to write a little more about the role of falsifiable hypotheses and falsification in science. The main problem is that people confuse falsifiable and falsification, often do not even seem to notice there is a difference, whereas they have very different roles in science.

The power of science and falsification are beautifully illustrated in this video by asking normal people on the street to discover the rule behind a number sequence (h/t U Know I Speak Sense).

Falsifiable

Karl Popper only asked himself what distinguishes a scientific hypothesis from an ordinary idea.
Popper's beautiful thesis was that you can distinguish between a scientific and a non-scientific statement by asking oneself if it can be falsified. If it cannot, it is not science. Thus the worst one can say about an idea that is supposed to be scientific is that it is not even wrong.

Important side remark: Please, note that also non-scientific ideas can be valuable, Popper's philosophy itself is not science, just like most philosophy, political ideas, literature and religion.

And please note that wrong hypotheses are also scientific statements; that they are wrong automatically shows that they can be falsified. Even falsified hypothesis are still scientific hypothesis and can even still be useful. An good example would be classical mechanics. This illustrates that Popper did not think about whether hypothesis were right or wrong (falsified), useful or not, but whether a statement is scientific or not scientific.

To be falsifiable, falsification is only needed to be possible in principle. It does not matter whether falsification would be hard or easy for the question whether it is science. This is because the main value of the criterion is that it forces you to write up very clearly, very precisely what you are thinking. That allows other scientists to repeat your work, test the idea and build upon it. It is not about falsification, but about clarity.

That also implies that the daily job of a scientist is not to falsify hypothesis, especially not solid and well-validated ones. Scientists are also not writing down new falsifiable hypothesis most of the time, in fact they rarely do so. Those are the rare Eukeka moments.

The terms scientist and science are clearly much broader and also much harder to capture. The ambitious William M. Connolley set out to define science and what a scientist does in a recent post. Definitely worth reading, especially if you are not that familiar with science. Disclaimer: not surprisingly, the aim was not completely achieved.

Psycho analysis

A classical example for Popper of a non-scientific hypothesis would be Freud's psycho-analysis. The relationship between the current psychological problems of a patient and what happened long ago in the patients childhood is too flexible and not sufficiently well defined to be science. That does not mean that what happens to a child is not important, there are many modern findings that point into that direction (Joachim Bauer, 2010). If someone else would succeed in making Freud's ideas more specific and falsifiable, it would even be a valuable contribution to science. It also does not mean that psycho-analysis does not help patients. Finally, it also does not mean that it is wrong, rather it means that it is not even wrong. It is too vague.

Morphic fields

Another example is the idea of Rupert Sheldrake about morphic fields. Sheldrake claims that when an idea has been invented before, it becomes easier to reinvent it. He has a large number of suggestive examples where this seems to be the case. Thus there is a lot of information to validate his idea.

The problem is, it is impossible to falsify the idea. This idea is, again, too vague and if you do not find the effect in an experiment, you can always claim that the effect is smaller, that the experiment was not sensitive enough or not well executed.

When I was studying physics in at Groningen University, Sheldrake gave a talk and afterwards naturally got the question whether his ideas were falsifiable. He dogged the question and started about the science philosophy of Thomas Kuhn on paradigm changes that shows that in practice it can be hard to determine whether an idea is falsified. However, whether an idea is falsifiable is clearly another question as how falsification works, which will be discussed below. Then Sheldrake started fueling tribal sentiments, by complaining that only physicists would be allowed to have hypotheses with fields, why not biologists? Discrimination! As the climate "debate" illustrates, adding some tribal conflict is an effective way to reduce critical thinking.

This does not mean that the ideas of Sheldrake may not turn out to be valuable. The list of examples that validate his ideas is intriguing. This may well be a first step towards making a scientific hypothesis. That is also part of the work of a scientist, to translate a creative, fresh idea you got during a hike into a solid, testable scientific idea. Morphic fields are, however, not yet science.

Anthropogenic global warming

The hypothesis that the man-made increases in the concentration on greenhouse gasses leads to an increase in the global mean temperature can be falsified and is thus a scientific hypothesis. There is no need to go into details here, because Hans Custers just wrote an interesting post, "Is climate science falsifiable?", which lists ten ways to falsify the "AGW hypothesis". One would have been sufficient.

A clear example is that if the average world temperature drops one degree, back to values before 1900 and stays there for a long time without there being other reasons for the temperature decrease (e.g. volcanoes, sun, aerosols) the theory would be falsified. To get to ten ways, Custers has to come up with rather adventurous problems that are extremely unlikely because so many basic science and experiments would need to be wrong.

Seen in this light, the climate ostriches are almost right, it is highly unlikely that the theory of man-made global warming will be refuted, that would require highly surprising new findings and in most cases it would require basic physics, used in many sciences, to be wrong. However, just because it is highly unlikely in practice that the hypothesis will be falsified because there are so many independent lines of evidence and the hypothesis is well nested into a network of scientific ideas, that does not make it theoretically impossible, thus AGW is falsifiable.

Falsification

It doesn't matter how beautiful your theory is, it doesn't matter how smart you are. If it doesn't agree with experiment, it's wrong. Richard P. Feynman

This quote is a favorite one of the climate ostriches. Unfortunately, falsification is a little more complex in practice.

On consensus and dissent in science - consensus signals credibility

Since Skeptical Science published the Pac Man of The Consensus Project, the benign word consensus has stirred a surprising amount of controversy. I had already started drafting this post before, as I had noticed that consensus is an abomination to the true climate ostrich. Consensus in this case means that almost all scientists agree that the global temperature is increasing and that human action is the main cause. That the climate ostriches do not like this fact, I can imagine, but acting as if consensus in itself is in bad thing in itself sounds weird to me. Who would be against the consensus that all men have to die?

Also the Greek hydrology professor Demetris Koutsoyiannis echoes this idea and seems to think that consensus is a bad thing (my emphasis):

I also fully agree with your statement. "This [disagreement] is what drives science forward." The latter is an important agreement, given a recent opposite trend, i.e. towards consensus building, which unfortunately has affected climate science (and not only).

So, what is the role of consensus in science? Is it good or bad is it helpful or destructive, should we care at all?

Credibility

In a recent post on the value of peer review for science and the press, I have argued that one should not dramatize the importance of peer review, but that it is a helpful filter to determine which ideas are likely worth studying. A paper which has passed peer review, has some a-priory credibility.

In my view, consensus is very similar, consensus lends an idea credibility. It does not say that an idea is true; if formulating carefully a scientist will never state that something is true, not even about the basics of statistical mechanics or evolution, which are nearly truisms and have been confirmed via many different lines of research.

The value of peer review for science and the press

The value of peer review keeps on producing heated debates. An interesting example was the weekend that physics professor Richard Muller wrote an op-ed in the New York Times. Some claim that Anthony Watts halted his blog for 2 days and released a scientific manuscript and an accompanying press release on the same weekend to steal attention away from Mullers op-ed. Both the op-ed and the press release were about scientific claims that had not passed peer review. Thus the Washington Post asked: Is it okay to seek publicity for a work that is not peer reviewed?

Watts et al. manuscript The eventful weekend at end of July 2012, resulted in two worthwhile blog post in the New York Times (Andrew C. Revkin at dotEarth) and the Washington Post (Jason Samenow). The manuscript was clearly released prematurely and had serious methodological problems. A few days after the press release and the blog reviews, Anthony Watts still wrote: "I’m hoping to post up a revised draft, addressing many of those comments and corrections in the next day or two." And he opened a "work page" for the manuscript, which is so quiet you can hear crickets. Just when no one expected it any more, the zombie manuscript came back from the undead; this March Watts wrote about this manuscript: "we are preparing a paper for submission". I am not a native speaking. May I ask, if you write "we are preparing", that indicates an ongoing action, right? Is there any lower limit on the intensity of this action?

The other side of the question about seeking the press before peer review is should a journalist only write about peer-reviewed studies? Further questions that came up since are: Is it unscientific to cite non-reviewed studies? Should the IPCC limit itself to reviewing only the peer-reviewed literature? Is peer review gate keeping? Is peer review necessary?

As often the context is important. What the value of peer review is, depends on who you are? An expert or not, a journalist or a newspaper reader? Another important part of the context is how controversial the finding is.

The Value of Peer Review for Science

Peer review gives an article credibility. As such peer review is "just" a filter, it does not guarantee that an article is right. Many peer-reviewed articles contain errors, many ideas outside of the peer-reviewed literature are worthwhile. However, on average the quality of peer-reviewed work is better. Thus peer-reviewed work is more likely worthy of your attention.

If you are a scientist and an idea/study is about something you are knowledgeable about there is no reason to limit yourself exclusively to peer-reviewed articles, but it is smart to prefer them. A scientist will only use peer review to preselect, because you simply cannot read and check everything. Life is short and attention a very limited resource. I also see no problem in citing studies that are not peer-reviewed, whether scientific reports or conference contributions. I do feel that by citing such studies, you give them some of your reputation, you become partially a reviewer and should read them as careful as a reviewer would.

Peer review is far from perfect. It is not intended to and cannot prevent fraud. Some bad papers will get through and some good ones will be rejected. This can be annoying for the scientists involved, but given that peer review is just a filter, it is not that bad for science in general. It is only since the second world war that peer review has become the standard. We also had scientific progress before that time.

Investigation of methods for hydroclimatic data homogenization

The self-proclaimed climate sceptics have found an interesting presentation held at the General meeting of the European Geophysical Union.

In the words of Anthony Watts, the "sceptic" with one of the most read blogs, this abstract is a ”new peer reviewed paper recently presented at the European Geosciences Union meeting.” A bit closer to the truth is that this is a conference contribution by Steirou and Koutsoyiannis, based on a graduation thesis (Greek), which was submitted to the EGU session "Climate, Hydrology and Water Infrastructure". An EGU abstract is typically half a page, it is not possible to do a real review of a scientific study based on such a short text. The purpose of an EGU abstract is in practice to decide who gets a talk and who gets a poster, nothing more, everyone is welcome to come to EGU.

An idea to combat bloat in genetic programming

Introduction

Genetic programming (Banzhaf et al., 1998) uses an evolutionary algorithm (Eiben, 2003) to search for programs that solve a well-defined problem that can be evaluated using one number, the fitness of the program. Evolutionary algorithms are inspired by the evolution of species by natural selection. The main difference of this search paradigm, compared to methods that are more traditional, is that is works with a group of candidate solution (a population), instead of just one solution. This makes the method more robust, i.e. less likely to get trapped in a local minimum. These algorithms code their solution as a gene. New generations of solutions are formed using sexual (and a-sexual) reproduction. Sexual reproduction (crossover of genes) is emphasized as it allows for the combination of multiple partial solutions into one, thus using the parallel computations made by entire population, making the use of a population less inefficient as it would be otherwise.

Bloat and evolution

One of the fundamental problems of genetic programming (GP) is bloat. After some generations (typically below one hundred), the search for better programs halts as the programs become too large. The main cause of bloat is generally thought to be the proliferations of introns. Introns are parts of the program that do not contribute to the calculation that is made, e.g. a large calculation that is multiplied by zero in the end, or a section that start with: if false then. Additionally, bloat can be caused by inefficient code, e.g. two times x=x+1, instead of x=x+2.

In genetic programming, the linear genes are typically converted into trees or lists with lines of code. Some other coding methods are also used, see Banzhaf et al. (1998). Performing a simple crossover on the genes would almost certainly result in an illegal program. Therefore, crossover is performed on a higher level, e.g. two sub-trees are exchanged between the sets of genes, or sections of program lines are exchanged. This way the child-program is legal code, but still, it is often bad code, and its results are not very similar to its parents. As a consequence, evolution favors bloat at the moment that it becomes hard to find better solutions. In this case, it is better for the parents to get offspring that is at least just as fit as they are, as getting less fit children. The probability to get equally fit offspring is better in large programs (with a small coding part) than in small ones (most of which is useful code), as in large programs it is more likely that the crossover operator does not destroy anything. The problem of bloat is thus intimately linked to the destructive power of crossover.

Advection of the disturbance fields in stochastic parameterisations

A relatively new method to estimate the uncertainties of the weather prediction is the use of ensembles. In this case, the numerical weather prediction (NWP) model is run multiple times with slightly varying settings. A popular choice is to vary the initial conditions (prognostic variables) of the model, within the range of their uncertainty. It seems like this method does not bring enough variability: the ensemble members are still relatively similar to each other and the observation is often still not in the ensemble.

As a consequence, meteorologists are starting to look at uncertainties in the model as an additional source of variability for the ensemble. This can be accomplished by utilising multiple models to create the ensemble (multi-model ensemble), or multiple parameterisation schemes or by varying the parameterisations of one model (stochastic parameterisations). This latter case is discussed in this essay.

Stochastic parameterisations

In parameterisations the effect of subscale processes are estimated based on the resolved prognostic fields. For example, the cloud fraction is estimated based on the relative humidity at the model resolution. As the humidity also varies at scales below the model resolution (sub grid scale variability), it is possible to have clouds even though the average relative humidity is well below saturation. Such functional relations can be estimated by a large set of representative measurements or by modelling with a more detailed model. In deterministic parameterisations the best estimate of, for example, the cloud fraction is used. However, there is normally a considerable spread around this mean cloud fraction for a certain relative humidity. It is thus physically reasonable to consider the parameterised quantity as a stochastic parameter, with a certain probability density function (PDF).

Ideally, one would like to use stochastic parameterisations that were specially developed for this application. Such a parameterisation could also take into account the relation between the PDF and the prognostic model fields. Developing parameterisations is a major task and normally performed by specialists of a certain process. Thus, to get a first idea of the importance of stochastic parameterisations, NWP-modellers started with more ad-hoc approaches. One can, for example, disturb the tendencies calculated by the parameterisations by introducing noise.

Online generation of temporal and spatial fractal red noise

It is relatively easy to generate fields with fractal noise using Fourier, wavelet or cascade algorithms (see also my page on cloud generators). However, these noise fields have to be calculated fully before their first use. This can be problematic in case a 2D or 3D spatial field is needed as input to a dynamical model that should also have a temporal fractal structure. Often the number of temporal time steps is so large, that the noise field become impractically large. Therefore, this essay introduces an online method to generate fractal red noise fields, where the new field is calculated from the previous field without the need to know all previous fields; only the current and the new field have to be stored in memory.
The algorithm involves two steps. The spatially correlated red noise is calculated from a white noise field using Fourier filtering. The white noise field evolves temporally correlated by addition of Gaussian noise. In other words, every pixel of the white noise field represents a fractal Brownian noise time series. Fortunately, the spatially correlated noise field retains the fractal nature of the temporal structure of the white noise field.

On cloud structure

I tried to keep this text understandable for a broad scientific audience. Thus who already know something about fractals, may find the first section on fractals trivial and better start with the second section on why clouds are fractal.

Clouds are fractal

Fractal measures provide an elegant mathematical description of cloud structure. Fractals have the same structure at all scales. That may sound exotic, but fractals are actually very common in nature. An instructive example is a photo of a rock, where you cannot see if the rock is 10 cm large or 10 m, without someone or some tool next to it. Other examples of fractals are commodity prices, the branch structure of plants, mountains, coast lines, your lungs and arteries, and of course rain and clouds.

The fractal structure of a measurement (time series) of Liquid Water Content (LWC) can be seen by zooming in on the time series. If you zoom in by a factor x, the total variance of this smaller part of the time series will be reduced by a factor y. Each time you again zoom in by factor x, you will find a variance reduction by a factor y, at least on average. This fractal behaviour leads to a power law; the total variance is proportional to the scale (total length of the time series) to the power of a constant; to be precise, this constant is log(x)/log(y). Such power laws can also be seen in the measurements of cloud top height, column integrated cloud liquid water (Liquid Water Path, LWP), the sizes of cumulus clouds, the perimeter of cumulus clouds or showers, and in satellite images of clouds and in other radiative cloud properties.

If you plot such a power law in a graph with logarithmic axis, the power laws looks like a line. Thus, a paper on fractals typically shows a lot of so called log-log-plots and linear fits. To identify scaling you need at least 3 orders of magnitude, thus you need large data sets with little noise.