Tuesday, October 21, 2008

NPR Steals My Line on Climate, Well Almost

A while ago I wrote a post on coffee shop conversations about global warming and how to convince friends that it is better to have more faith in peer reviewed science than random self appointed experts who pop up on the Internet and pronounce that an entire field of study with a history of thousands of publications is wrong because that one paper in that hard to find journal says so.

NPR no doubt taking my cue has a similar "How To Be Conversant About Climate" conversation with Michael Oppenheimer, faculty in geosciences Princeton University, about how to convince Uncle Sal at the dinner table that global warming is not only real but is primarily caused by human activity. Oppenheimer is of the view that while the debate about whether warming is taking place is over and settled, there are differences which are quite genuine and not as easily dismissed on how to deal with this problem. That problem has economic, political and cultural roots and needs to be confronted without taking a "I am right and you are wrong" stance.

There is a second good talk on the show. This explores the broader question of why people find it hard to believe in science and scientists. Uncertainty in interpreting data and the results of an experiment, dissent and debate is built into the scientific process. It is the nature of the beast. But it is often bewildering to people that even after say 2-3 decades of work, there is no certain answer from scientists on a problem. People take that as a weakness of the system, a signal that science may not give them ready answers. Maybe we want that certainty, maybe we need closure on a problem, maybe we need to be reassured that yes this is the one correct answer, that we all too readily succumb into believing a confidently told but scientifically unsupported story.

And again how easily available information on the Internet can be at once a boon but also a hindrance and sometimes has dangerous consequences when people in responsible positions choose to believe wild assertions. Harry Collins of Cardiff University gives the example of parents refusing to vaccinate children with the MMR vaccine because they fear that it may cause autism in their child, something that has been refuted by solid scientific work.

Then there is the case of Thabo Mbeki, ex-President of South Africa, who after reading articles on the Internet announced publicly that retro-viral drugs to combat AIDS don't have any helpful effect. The media too bears some responsibility of spreading disinformation through their insistence on giving both sides on the story equal weight even though one side is just illogical and is not supported by any body of evidence and simply does not warrant such attention.

Both a good listen.

Friday, October 17, 2008

Indian Geosciences And Thoughts On Bengali Geologists

From ScienceWatch.com, analysis of an increase in scientific output from India measured using various indicators. For the analysis ScienceWatch used the Thomson Reuters database National Science Indicators. Indian scientific output shows improvement in the number of papers indexed by Thomson Reuters since 1981 and also improvement in the number of citations per paper i.e. a count of how many times your paper is being referred to by other researchers. The reasoning is that if you have said something significant then people will refer to it over and over again.

Here are some stats:

In 1985, the number (number of papers indexed by Thomson Reuters) was approximately 12,500, and for the next 15 years the total never much exceeded 14,000. Around the year 2000, however, the number began to tick upwards, rising to nearly 17,000 in 2001, reaching 20,000-plus in 2003, and winding up at more than 27,000 in 2007.

and regarding citation stats...

Physics, as it happens, features prominently in the next set of graphs(2), which plot the nation's relative citation impact (that is, India's citations-per-paper average compared against the world average in each respective field) in 14 main fields in a series of overlapping periods from 1985 through 2007

.
...although the impact of India-based research lags the world average in the fields shown, the nation has been on a discernible upswing since roughly the year 2000, with notable gains in, for example, Geosciences(3), Neurosciences(4), and Biology & Biochemistry.

Geosciences citations have improved from being 30% of the world average to around 50% of the world average.

Source: ScienceWatch

A slight improvement I guess over the last 20 years or so, but the condition of a lot of geoscience departments, especially those in state universities leaves a lot to be desired in terms of quality research output. I don't want to analyze this to death. Yes we are improving, but still far behind the U.S, Europe, Britain, Japan and China. A lot has been written on what ails Indian science. Here is a polemical view. And here is a more detailed and critical analysis.

I want to digress and write on a different matter.

I survey the research literature in sedimentary geology quite a bit and all along there has been only the rare publication by India based geologists in the leading geology journals on sedimentary topics. With one peculiar exception. Bengali geologists mostly based in eastern Indian Universities have been consistently publishing papers on sedimentary geology in internationally recognized journals.

I've often thought about this on and off and a number of reasons come to mind.

1) Proximity to sediments. Jadhavpur University Kolkata, Presidency College, Kolkata, Indian Statistical Institute, Kolkata, IIT Kharagpur and School of Mines, Dhanbad, where a lot of Bengali geologists are based are all within a days driving distance from a variety of sedimentary terrains. These include the Ganges-Brahmaputra fluvial- estuarine-delta complex, the Mio-Pliocene Himalayan foreland and the late Paleozoic-early Mesozoic Gondwana continental rift basins. But proximity cannot be the only reason. Plenty of other Indian universities to the west and south and north sit right on top of Proterozoic and Mesozoic basins.

2) Institutional support. The east is where the systematic study of geology began in India. Kolkata was the seat of British administration until the capital shifted to Delhi in the early 1900's. The Geological Survey of India was established in Kolkata in the late 1800's. There is a long tradition of geology in this part of the country reflected in well funded geology programs in various universities.

3) Economic incentives. There is coal in the eastern rift basins and oil was discovered early in the eastern state of Assam. There was a requirement for expertise in sedimentary geology and eastern universities developed strong sedimentary geology programs to provide it.

4) But Universities are only as good as the people who work in them. Geology frankly has never been a field which attracts the brightest students in India. Parental, social and economic pressures lead to Medicine, Engineering and Management being preferred. Among sciences, Math, Physics and Chemistry suck in the best students. But there has been strong social and cultural support in Bengal for pursuing any kind of an intellectual career. So maybe many more bright Bengali students become geology researchers than elsewhere in India.

I remember my first ever meeting with my graduate adviser in the U.S. "Meet me at 8.00 am sharp", he barked on the phone. And so I went to his office, braving a bitterly cold mid-west January morning. He was a Bengali sedimentary geologist and a fine one too. Winner of the SEPM paper of the year award for 1986 for his excellent work on alluvial sandstones from the late Paleozoic-early Mesozoic rift basins of eastern India.

Maybe there is a germ of truth to my thesis :-)

Monday, October 13, 2008

Some Very Strange Maps

Once in a while I stumble upon, not StumbleUpon, just stumble upon a blog that makes me press the subscription button at once.

Strange Maps is one of these sites. A wonderful collection of weird, creative and humorous ways of depicting the world in "maps".

Having spent long late afternoons in Appalachia sipping sweet tea after a day's fieldwork this one I could relate to all too well; the north to south distribution of McDonald's selling unsweet and sweet tea in Virginia.



Source: Eightoverfive.com via Strange Maps

I am not sure how many geo-bloggers have this site on their blogroll. It's a refreshingly different spot. You can't really categorize it as a geoscience blog although there are posts on geology related topics. But it's just the spatial depiction of information that caught my interest. Maps always give you a different perspective, makes you ponder over previously unrecognized connections. You can create your own story of places visited, geology observed, people you have met. Definitely worth a regular visit.

Hat Tip: Geology News

Saturday, October 11, 2008

Are Most Papers in Nature Likely To Be Wrong?

A friend emailed me this article from The Economist:

In Economic theory the winner’s curse refers to the idea that someone who places the winning bid in an auction may have paid too much. Consider, for example, bids to develop an oil field. Most of the offers are likely to cluster around the true value of the resource, so the highest bidder probably paid too much.

The same thing may be happening in scientific publishing, according to a new analysis. With so many scientific papers chasing so few pages in the most prestigious journals, the winners could be the ones most likely to oversell themselves—to trumpet dramatic or important results that later turn out to be false. This would produce a distorted picture of scientific knowledge, with less dramatic (but more accurate) results either relegated to obscure journals or left unpublished.

Marginal Revolution has a more detailed debate in the comments section.

I don't read much geosciences in Nature or Science. But should I now pay more attention to papers in Carbonates and Evaporites instead of Journal of Sedimentary Research?!

Apologies to the Editors.

More importantly, is some of my unpublished PhD work really the real deal!?

Tuesday, October 7, 2008

Was September 08 Earthquake Month?

The September earthquake story has developed nicely with just the right amount of coincidences to cause excitement in the astrology community. Triumphant comments have been left on my earlier posts on earthquakes claiming "prediction" success. I noticed throughout the month of September quite a bit of traffic coming to my blog from search queries of the like " September astrology earthquake prediction" and so I gather there is a lot of curiosity about whether astrologers can really predict earthquakes.

Just a quick recap for those who have not been following this thread. Some astrologers claim that they can predict big earthquakes based on the alignment of the moon with the sun and other planets in the solar system. Mr. Amit Dave who lives in Mumbai is one such astrologer. He claims that certain alignments of the moon and sun (and other planets) exert tidal stresses on earth. No one doubts this. Nothing new here. The relevant question is whether these tidal stresses acting alone i.e. without input from internally generated tectonic stresses are enough to cause big earthquakes. The answer is clearly no.

In two earlier posts I exposed Mr. Dave's lack of understanding about geology and earthquakes and explained just why tidal stresses acting on their own are incapable of causing big earthquakes on earth. He never gave me a single substantive reply to my questions but kept making foolish claims about his "predictions" in the comments sections of my posts. So in this post I will try to explain what is a scientific earthquake prediction and whether Mr. Dave's "predictions" measure up to this standard.

The main confusion among people is between making the rather mundane claim that earthquakes will occur on any particular day against the more meaningful exercise of predicting a earthquake. Astrologers are unable to grasp this distinction or more cynically ignore it since the laypersons they are trying to reach certainly don't make this distinction.

When geologists make a prediction about a earthquake it is in this format:

An earthquake of magnitude ...6.5-7 magnitude (range).. will strike (with a probability) ...time range..(usually in the range of the next few years) at location... Longitude X and Latitude Y due to ... hypothesis of how and why an earthquake is due /imminent in the declared region.

Astrologers ignore the third requirement of the location and are puzzled by the fourth. Why give a hypothesis of how and why an earthquake is imminent.

Here's why. Geologists studying earthquakes and working in a particular area have a hypothesis on how geological stresses may cause an earthquake. This hypothesis describes for example the rates at which strain is accumulating along a particular fault based on say the motion of the earth's plates. Or take tidal forces. Ultimately they can only act or trigger earthquakes by imposing stresses on the crust. So even a theory of earthquakes like Mr. Dave's which relies solely on tidal forces needs to explain in terms of the imposed stress and accumulating strain how , why and where these tidal forces are likely to cause a rupture. If an earthquake does occur in that region, this hypothesis can then be validated against what actually happened. It gives geologists a chance to test how good their understanding of earthquake mechanisms is.

Here is a great example of this process from the September issue of Geology in a paper that studies the big Kashmir earthquake of 2005:

.... We thus conclude that the 2005 event did not occur on the plate boundary megathrusts, but on intraplate active faults within the Sub-Himalaya. Consequently, the accumulated elastic strain around the complex northwestern margin of the Indo-Asian collision zone has not been significantly released by the 2005 earthquake.

What this means is that the earlier hypothesis that a big earthquake in Kashmir will likely take place along a major boundary fault zone was wrong. The earthquake instead took place on a subsidiary fault system. This however warned geologists that strain is still accumulating along the major fault and another big earthquake in that region is likely.

Science is a self correcting process. If we are to improve our understanding of earthquakes, a prediction must include a declaration of a hypothesis which can then be validated against what actually took place.

Alright, on to Mr. Dave's "predictions". Here it is

1) 7th September 2008 -- 2020GMT ---- 7.2 On Richter scale at 12 Degrees East or/168 Degrees west ( Precise location both longitude and latitude is not possible as of to day)2) 10th September 2008 --1950 GMT ---7.3 on Richter scale-at 66 Degrees east or 114 degrees West3) 15 th September 2008 --0910 GMT----More than 7.5 on Richter scale -at 147 Degrees East or 33 Degrees west4) 16 th September 2008 ---2015 GMT---- 7.5 on Richter scale -at 87 Degrees East or 93 Degrees West

The original prediction he gave me did not have even the incomplete location information. It is only when the USGS insisted that a location is necessary (he forwarded me a mail from the USGS) that he put in just the longitude.

But that's not it. These dates are way too specific and Mr. Dave then provides a hedge.

As per my theory the dates of September 2008 are already published in my blog. However one more thing required to be noted is that , the dates shown are the peak force dates ,and have maximum potential of triggering the quakes. However , the quakes follow a sinusoidal pattern. As the dates mentioned approaches nearer the quake intensity and frequency keep on increasing. After the dates have passed the tidal force slowly fades away.

And then not satisfied that his sinusoidal wave will cover enough ground, a few more dates:

Thus , there are fair chances of quakes of 6.5 to 6.8 Richter scale on following dates also. places not known.27 th August 2008----0415 Hrs IST31 th August 2008-----0115 Hrs ISTAlso 22 nd and 23 rd September 2008----up to 7.0 on Richter scale
So, he has covered so much time with his spread of dates that he has ensured himself a high degree of "prediction" success.

Why is that? Because of the high frequency of moderate to big earthquakes, something most people are unaware of. 7-8 magnitude earthquakes occur at a frequency of 1-2 per month, 6-7 magnitude earthquakes occur 10-12 per month, 5-6 magnitude earthquakes occur 2-4 per day, and 4-5 magnitude earthquakes occur 20-25 per day.

As expected Mr. Dave claimed earthquakes of varying magnitudes on dates before, between, on and after his peak potent period. Not a single 7+ earthquake occurred on any of these days, but Mr. Dave blatantly misstated the magnitudes of these earthquakes. It is a comical list of claims, but one that speaks to his disingenuity. He seems incapable of reporting accurately even the most basic of facts.

You can check for yourself all this earthquake information by going on the USGS earthquake search.

But my main quibble is not with these reporting inaccuracies. I put that down to over-excitement. My objections are over procedures.

What is the justification for claiming that tidal forces caused the earthquake that was claimed and not one of the dozen others that took place on that day? All Mr. Dave is doing is choosing post facto from a list of earthquakes that occurred on the day, the earthquakes that is closest to the magnitude he gave or one he deems significant in some way.

I like to call this retrospective coronation.

Again, how does he know that tidal forces caused the bigger earthquake he has claimed and not a smaller earthquake that took place some where else? He doesn't know, but you see without a prior declaration of a location and a hypothesis it is impossible for anyone to verify his claim. And that suits the astrology community just fine.

That is why astrologers never give a location.

Mr. Dave will deny that he chooses earthquakes after they occur, but here is an example which gives his game away. He had "predicted" that an earthquake of magnitude 7.5 or so will occur on September 16th. Unfortunately for him there was no earthquake larger than 6 on that day. So what did he do? He claimed a 5.0 magnitude earthquake (he gives the mag. as 4.9) that occurred in the Satara district of Maharashtra as proof of his "predictions"!

It is stupid claims like this one that lays bare the intellectual bankruptcy of these bogus earthquake prediction schemes astrologers concoct from time to time.

Remember earthquake frequencies? There were 26 earthquakes on September 16 of magnitudes 4-6. There were 4 earthquakes of magnitude greater than the one he choose. How does he know that tidal stresses caused the Satara earthquake and not any one of the other 25? He doesn't. The only reason he choose to claim this particular earthquake is because thousands of people in the Mumbai-Pune area felt it. Humans react to and remember events that occur close to the place they live and Mr. Dave knew he could fool people into believing that he actually predicted that earthquake.

So astrologers are not really predicting a earthquake. They are simply just highlighting after the fact one out of the many that occur on any given day.

Location and a hypothesis, Location and a hypothesis, Location and a hypothesis. Without these any claim to a prediction is meaningless.

Did I mention even the incomplete location information given by him was wrong for all the earthquakes he has claimed?

I find this non-disclosure of location information most revealing! Mr Dave and other astrologers argue that they too are trying to save lives by coming up with a prediction system. That is as strange an argument as I have even come across. How on earth can you save lives if you don't specify where an earthquake is going to strike?

One last requirement of a scientific prediction is the complete disclosure of negative results. Mr. Dave has never admitted that his location information is missing or wrong in all the cases. Another type of negative result is earthquakes that occur on days other than those given. Ambuj pointed this out sometime back. Mr Dave claims that big earthquakes occur solely due to tidal stresses and there are potent days when they will occur. Since his September prediction which appeared on my blog on July 13 to the date of this post, leaving aside the entire month of September, there were 25 earthquakes of magnitudes 6 and above. How does he explain these earthquakes which all fell on days when the heavenly bodies were not engaged in a destructive tango?

In fact how does he explain not just these biggish earthquakes but the hundreds of smaller earthquakes that occur every day?

Here is my suggestion for what Mr. Dave's next prediction should look like:

Magnitude X (range) , Time range , location in Longitude and Latitude and a hypothesis of how and why tidal stresses will cause strain to build up and rupture rocks in the region he has indicated. This is to be followed by a full disclosure of negative results.

If he does this I am sure the seismology community will follow his claims with interest. But just rolling the dice and jumping up in excitement whenever an earthquake happens to coincide with a full moon, new moon or quarter moon day is not acceptable.

I predict Mr. Dave will not accept this qualifying criteria.

I am sure because:

Gustav ,Hanna and Ike three major storms in first week of September08Vanuattu quake of 6.5 on 8th September 2008 ,at 0305 Gmtkeep counting pleaseDave
Astrologer / Engineer

Yes hurricanes too. With this mindset Mr. Dave will continue to fool himself and other ignoramuses who are already convinced that their destiny has been decided by the alignment of distant planets.