Until a few decades ago, natural disasters were thought of as anomalies; freak occurrences that one could at best prepare for, but not predict.
A 39-year-old Danish theoretical physicist, Per Bak, came up with something called the sandpile model. (Read on for a bit more on what that is.) Two Indians, Deepak Dhar and Ramakrishna Ramaswamy, took that theory and built a mathematical model that proved it worked. He returned to India in 1978, to work at the Tata Institute of Fundamental Research (TIFR) in Mumbai. In conversation with Kanika Sharma, the 74-year-old talks about his pathbreaking model, time and ageing, and how Sheldon Cooper really does get a lot of things right. Bak was about 40 when he visited TIFR and spoke about this theory. In 1989, when we published our paper, it was the first analytical solution of this class of models. If I say a large earthquake will occur somewhere in India in the next 100 years, that is too weak a prediction to be useful. On the other hand, work on highly difficult problems and you might work for 30 years and still not arrive at a solution. I suppose it is an achievement to survive this long… I would prefer to be 35, but that’s not going to happen. At 30 or 40, I was on edge much more. I don’t look forward to being 90, you know. FACT FILE: DEEPAK DHAR * Dhar, 74, grew up in love with science. “I would prefer to be 35, but that’s not going to happen. I don’t look forward to being 90.
The laws of probability didn’t seem to work here. When it came to certain systems — storms, earthquakes, meteors — there was too much at play; too many layers, each one too complex and shaped by too many variables. Who could model or account for that? An answer emerged in the 1980s. This year, Dhar became one of four physicists awarded the Dirac Medal by the International Centre for Theoretical Physics. (Previous recipients include Stephen Hawking.) Dhar’s Abelian sandpile model has, over the years, been applied to earthquakes and avalanches, traffic and financial markets. Dhar grew up in love with the science magazines his father bought him. After graduating from University of Allahabad and earning a Master’s at the Indian Institute of Technology-Kanpur, he moved to the California Institute of Technology (Caltech) for a PhD. There, he worked with the legendary physicist and Nobel laureate Richard Feynman and served as his teaching assistant. Excerpts from the interview. * What can you tell us about the sandpile model? Think of a sandpile with sand being added to it slowly. It will eventually reach a point of instability when an avalanche of sand grains will occur. Now this instability is the cumulative effect of small events: grains being added bit by bit. Still, a big event is inevitable. Bak argued that this happens in many natural processes, such as fluctuations in the River Nile, the extinction of dinosaurs, solar flares, and so on. He was just a few years older than I was. He was very charismatic and a powerful speaker. My colleague Ram Ramaswamy and I were not fully convinced, and the two of us actually thought we could prove him wrong. To test his theory, we created a variation of the mathematical model that Bak had proposed. He had not managed to solve it analytically, only studied it by computer simulation. We managed to solve this variation exactly, using his theory. So, it was very useful in advancing the subject and convincing people that this was the right area to study. * How accurate is it? If you apply the same general idea of sandpiles to solar flares, for instance, it helps predict the frequencies at which large events will occur, and helps predict their sizes. But a prediction also has to be useful. Our predictions are usually on an intermediate scale. Can you tell us what it’s like being inside the head of a theoretical physicist? How much do these characters get right? I would say Sheldon Cooper, in The Big Bang Theory, though of course a caricature, represents a quirky socially maladapted scientist quite well. He’s a bit difficult to live with, but people around him are kind and accommodating. He is passionate about his work. I think a lot of us also have our fixations. In the end, hopefully, the result is some progress for science, and for mankind. Our value system is also a little bit different from everybody else’s, in the sense that the things that are important to everybody else, like money and power, are not so important to us. That can be a major source of discomfort to others. Then again, some of our friends say, “If you’re so smart, why aren’t you making more money? What can I say? I’m content doing research, solving sudoku puzzles and playing card games in my spare time. * What would you say is a crucial ingredient in this kind of work? Some people think of research in science as sort of detective work. The importance of looking at a problem with a fresh eye is something I learnt from Professor Feynman. He used to say: “Are the questions you ask understandable to a high school student? They may not understand the answer, but at least they should be able to explain what the question is. If you cannot explain it to someone who is not an expert, then perhaps you have not yourself understood the question well enough. Feynman was a legendary figure. One could learn a lot just by watching him. Of course, he knew who he was, but he was not pompous. He was kind, especially with students. He was quite polite and gentle, explaining ideas without making them feel dumb. * What do you feel most strongly, looking back on your years and your work? You learn to adjust to time. I find that I am not so nervous all the time. I think I’m now a bit more relaxed. Living simply becomes a bit harder, as time goes by. He studied it at University of Allahabad, then the Indian Institute of Technology-Kanpur, and the California Institute of Technology (Caltech), where he worked with the legendary physicist and Nobel laureate Richard Feynman. * He returned to work at the Tata Institute of Fundamental Research (TIFR) in Mumbai, then the Indian Institute of Science Education and Research (IISER), Pune. He is currently a distinguished professor at TIFR’s International Centre for Theoretical Sciences (ICTS-TIFR) in Bengaluru. * What does it feel like to look back on such a life?
Because many other people are already working on them, that’s what the sandpile model makes possible. * You like to joke that you work on an intermediate timescale too… Well, I don’t like working on very fashionable or pressing questions,. Because there are so many TV shows about marvellous geeks and nerds today, it can sometimes feel like we have some idea,.
It was proposed by the Danish theoretical physicist Per Bak in 1987, linked to what he called “self-organised criticality”, which was his general theory of how nature works. Any system is evolving, with multiple factors making it complex, he said. And he said his model could help estimate the tipping points at which such events might occur. * So you set out to build a mathematical model based on his theory? And he said: This is how nature works. The “goodness” of prediction will naturally depend on the quality of the data, and the algorithm used. I work on problems of the “middle path”, of intermediate difficulty. (Laughs) * That’s exceedingly modest. “I suppose it is an achievement to survive this long,” says Dhar, chuckling.
And making a very specific prediction — like an earthquake in a certain city tomorrow — is much harder. Of course, you cannot tell what will happen tomorrow, but the expected daily payoff decreases as you grow older. .
