Friday, July 6, 2018

Business India, Talking To Prof Roddam Narasimha


Business India - the magazine of the corporate world, July 2-15, 2018

Talking To

“Instead of monsoon India followed the fashions of the West”
Prof Roddam Narasimha
(Courtesy ICTS)

Prof Roddam Narasimha, (84), FRS, is a distinguished aerospace
scientist, and among the first few Indian engineers to be elected to
several prestigious international academies like the Royal Society,
UK; the US National Academy of Sciences, US; National Academy of
Engineering, US and the American Academy of Arts and Sciences.
He has contributed enormously to the development of aeronautical
and space sciences in India. He is presently at the Jawaharlal Nehru
Centre for Advanced Scientific Research, Bengaluru. One of his
major areas of research today is cloud evolution and dynamics, a
subject of great relevance to the Indian monsoons and global climate
change. These are excerpts of a conversation between Shivanand
Kanavi and Roddam Narasimha

In the U.S., every tv channel will
tell you weather predictions for
the next 24, 48, 72-hours and
they are fairly accurate. What is
it that makes weather prediction
complicated in India?

The dynamics at higher latitudes of
Europe and North America are simpler
than the dynamics in the tropics.
It is not that the British and Americans
are better at predicting the monsoons
than us. Of course we make mistakes
but so have they.
We have to be the foremost experts
in the world on monsoon. But the best
minds in India have not devoted their
time to the study of monsoon and they
have followed the fashions of the West.
And the West will study what is important
to them. We can say with the state
of knowledge existing in the world
today, not just in India, predicting the
monsoon is a tough job. The Indian
response has been to borrow unknown
American computer codes and so on,
but our experience with them is not
any better. Americans themselves say,
“We can’t predict monsoons”.

Can we compare the actual weather
data with traditional Panchanga
predictions and see any correlation
and how accurate the predictions
are?

If we are talking about rainfall, the correlations
are poor. Panchanga depends
on the longitude, etc. So, there is
no way that it can be precise for the
whole of India or even a large region
of India. Nakshatras for rains are basically
empirical and those are correct
but they vary from region to region.

What about El Nino and so on?

El Nino was first discovered by a British
head of the India Meteorological
Department (imd) around 1915-1920.
Sir Gilbert Walker was a very bright
man. When he came here he didn’t
know any meteorology but he was
fascinated by it. But after a while
he decided that the monsoon was
too erratic – and he didn’t see how
one could understand the dynamics
completely. He said, “I will try and
do statistical predictions” and introduced
statistical predictions to the
Met department.
Till today, Walker’s philosophy, not
precisely his formula but his philosophy,
is what governs predictions in
our imd. It is based on some correlations.
There is Vasant Gowarikar’s
model that took many more parameters
than Walker’s did, but the philosophy
was exactly the same. One
of the things that Walker found was
that monsoon rainfall in India had a
link with what happened in Australia
– which is ironic. Walker didn’t live
to see that a couple of years after he
died, the connection was established
by much better measurements. It left
no doubt. That phenomenon is what
has grown into El Nino. The connection
that Walker saw between Indian
monsoon rainfall and what happened
in Northern Australia is one signal of
El Nino.

ISRO has launched many
meteorological satellites, have they
helped in any way?

Yes, in a big way. To give you a striking
example, look at our cyclone predictions
now. In fact our Met department
had also predicted the cloud burst that
happened in Kedarnath, Uttarakhand.
But no government department took it
seriously. And it led to a major disaster.
So, we are getting better at that
kind of relatively short-term prediction.
Doppler radar too helps a great
deal in short-term predictions whereas
our satellites help a great deal over the
life cycle of a cyclone.

Has computer modelling helped? Is
it easier in upper latitudes?

Computer modelling has helped, but
not as much as one would have liked.
By and large it is easier in the upper
lattitudes. The clouds are
what stump us in the computer models
of the tropics. Northern latitudes
do not have these kinds of cumulus
clouds. They are largely static clouds,
spread clouds. So that is why clouds
are very important to India and that is
how my interest in clouds started.

Is global warming a certain
thing now?

It depends on whom you ask. The atmospheric
scientists are pretty convinced.
Just look at how the weather is changing.
The Arctic is melting, nobody
can doubt that, the evidence is very
strong. The frequency of the extremes
in weather is increasing. Basically,
the change in the extremes would be
more noticeable than the change in
the average. So, a degree or two in
mean earth’s temperature might not
change the mean rainfall very much.
But if you look at the monsoons of the
last twenty years, extremes of a certain
kind are more common.
So, atmospheric scientists say it is
dangerous to ignore that this may be
a possibility. By and large earth scientists
are more or less convinced that
this is happening.
But, as I said, there are other people,
businessmen and some pretty
well-known physicists who argue
against it. Their point may be translated
as “there is no clinching evidence
in favour of climate change”.
Which is true but the people who
look at extremes say the evidence is
getting stronger and stronger. It is
stronger than it was ten years ago.

However it is still based on only 100
years of reliable data?

Quite right. So the debate goes on.
Moreover, clouds are the greatest
uncertainty in these climate models,
particularly the interaction of clouds
with radiation. To give a simple example,
if you have an overcast sky during
the day it is a bit cooler because
the solar radiation is not coming down
to us, it is getting reflected back. But
during the night an overcast sky is
warmer. Why? The sun is not there
but the infrared from the ground gets
reflected back from the clouds. So, an
overcast day is pleasant, an overcast
night is warm. This is a spectacular
everyday example of how clouds interact
with radiation.

What fascinates you about
clouds and what are the unsolved
issues there?

The largest uncertainty in the prediction
of climate change has to do
with clouds. In the research done by
all the different groups in the world
about which way the atmosphere may
go as you emit more greenhouse gases
and various other substances into
the atmosphere, every now and then
something happens which doesn’t
agree with what is predicted. The last
ten years is a good example. Then
there are always sceptics who say the
models are no good and that you are
just making alarmist statements. “If
you can’t explain why earth’s temperature
hasn’t gone up in the last 10 years,
how can we believe anything you say
about global warming,” they say.
Actually some physicists are also
sceptics. The general view in the
atmospheric science community is:
“We do not claim to predict every blip
that occurs in it. The blip is not going
to prevent warming of the planet on
the whole, or climate change”.

When did you switch to studying
clouds from problems in aerospace?

Globally, the serious study of clouds
started around the 1960s. Very simple
experiments were made. Unfortunately,
by the 1970s, it was realised
that those simple experiments were
not behaving like clouds at all. We
picked it up where they had stopped.
In the 1980s, I started an experiment
in the Indian Institute of Science, Bengaluru
which turned out to be very
revealing. And so from then on I have
been working on clouds in some form
or other. We are getting closer and
closer and we can reproduce a wide
variety of cloud types in our water
tank. Now a lot of people have started
taking notice. However, it is difficult
to reproduce all the atmospheric conditions
in the lab whereas on the computer,
I have more freedom, provided I
have access to a Supercomputer.
In England, rainfall is not the major
variable, it rains almost every day. It is
a drizzle most of the time, so they are
worried about temperature. In India,
we don’t worry about the temperature,
because it is warm most of the time.
Our variable is rainfall. Rainfall is one
of the hardest things to predict. Temperature
and wind can be predicted
more easily than rainfall. It can be
raining here in Malleshwaram, but not
on the Airport road in Bengaluru.
The last bit of physics required to
tell whether it is going to rain or not
is very hard. Most of these cumulus
clouds that you see in India during
the monsoons, are actually flow
clouds. They are bubbling up from
the ground. And therefore there is a
very complex fluid dynamics associated
with it, it is not just thermodynamics.
It comes as a surprise to
many people that these things that
we see every day, which we admire
and love are not understood. u

Shivanand Kanavi, Former VP of TCS, is Consulting
Editor at Business India and Adjunct Faculty at
NIAS. He tweets as @shivanandkanavi and blogs at
http://reflections-shivanand.blogspot.com./ He can
be reached at skanavi@gmail.com
The Arctic is melting,

Monday, July 2, 2018

Conversation: Prof Roddam Narasimha Global warming, monsoon, cloud research


Conversation with Prof Roddam Narasimha: 

Global Warming, Monsoon Forecasting, clouds and all that

(appeared in http://www.rediff.com/news/interview/rainfall-is-one-of-the-hardest-things-to-predict/20180711.htm

and http://www.rediff.com/news/interview/an-interview-about-rain-you-must-read/20180716.htm )


Prof Roddam Narasimha 
(Courtesy IISc website)

Prof Roddam Narasimha, FRS, is a distinguished aerospace scientist, and among the first few Indian engineers to be elected to several leading international academies like the Royal Society, the US National Academies of Sciences, National Academy of Engineering and the American Academy of Arts and Sciences. He has contributed enormously to the development of aeronautical and space sciences in India. He is presently at the Jawaharlal Nehru Centre for Advanced Scientific Research, Bangalore. One of his current areas of research is cloud evolution and dynamics, a subject of great relevance to the Indian monsoons and global climate change. These are excerpts of a conversation between Shivanand Kanavi and Roddam Narasimha.

Q. What fascinates you about clouds and what are the unsolved issues there?

A. Clouds are central to two major problems today—one is climate change. The largest uncertainty in the prediction of climate change has to do with clouds. In the research done by different groups in the world, about which way the atmosphere may go as you emit more greenhouse gases and various other substances into the atmosphere, every now and then something happens that doesn’t agree with what is predicted. The last ten years are a good example. So there are always sceptics who say 'those models are no good and you are just making alarmist statements. If you can’t explain why earth’s temperature hasn’t gone up in the last 10 years, how can we believe anything you say!'.

Q. Michael Crichton wrote a novel, and half of it is a polemic about climate change. Have you read it, it is quite interesting?

A. Yes, it is quite interesting. Actually some physicists are also sceptics. The general view in the atmospheric science community is: “Yes, every now and then there are blips. We do not claim to predict every blip that occurs in it. But after the blip, we can explain what is happening. This (the blip) is not going to prevent warming of the planet on the whole, or climate change”.

Q. Is global warming a certain thing now?

A. It depends on whom you ask. The atmospheric scientists are pretty convinced. They say forget about the details. Just look at how the weather is changing. The Arctic is melting, nobody can doubt that, the evidence is very strong. Or if you look at the extremes of weather, the frequency of the extremes is increasing. There, the evidence is very strong. Now, that would be consistent with what the models are saying. Basically, the change in the extremes would be more noticeable than the change in the average. So, a fraction of a degree might not change the mean rainfall here very much, but the extremes can be affected. If you look at the monsoons of the last twenty years, extremes of a certain kind are more common.. So, atmospheric scientists say it is dangerous to ignore that this may be a possibility. But, as I said, there are other people who argue against it and there are some pretty well-known physicists who argue against it.

Q. What about earth scientists, oceanographers …?

A. I think by and large earth scientists are more or less convinced that this is happening.

Q. Who is not convinced?

A. A large number of businessmen are not convinced.

Q. Within the scientific community?

A. Within the scientific community there is a small group of physicists who are quite distinguished, but quite sceptical. I wouldn’t say they are people who don’t know anything and it comes from the way they do physics. In physics generally a question has either yes or no as the answer. And so they will want some clinching evidence. Their point may be translated as “there is no clinching evidence in favour of climate change. It is variable but we have always known that it is variable. There is not even strong evidence that the variability has changed". But the people who look at extremes say the evidence is getting stronger and stronger. It is stronger than it was ten years ago.

Q. But, it is still based on just 100 years of reliable data?

A. You are quite right that is a problem. So the debate goes on. However on clouds everybody agrees. What is the greatest uncertainty in these climate models? The greatest uncertainty is clouds, particularly the interaction of clouds with radiation. It is complex. For example, if you have an overcast sky during the day it is a bit cooler because the solar radiation is not coming down to us, it is getting reflected back. But during the night an overcast sky is warmer. Why? The sun is not there but the Infrared from the ground gets reflected back from the clouds. So, an overcast day is pleasant, an overcast night is warm. This is a spectacular everyday example of how clouds interact with the radiation. There are other effects as well. So a lot of uncertainty is because of such interactions.
The monsoons are of great interest to India and increasingly everywhere in the world. Over the last 20 years, people have discovered other monsoon-type climates in the world. Till now, the monsoon meant the Indian monsoon, fpr most people, but now you have to say ‘Indian Monsoon’, because there is also an Australian monsoon also an African monsoon and many others, although they are not of the same scale as the Indian monsoon.

Q. Is the monsoon specific to the Indian subcontinent or is it also present in other parts of Asia?

A. They are elsewhere but they are on a smaller scale. Now there is a more general definition of monsoon that it is a switch in circulation and it is a fairly rapid switch in circulation at a certain point of the year and it occurs every year at around the same time. That it is a cycle, so it switches on around late May or early June and it switches off around September in case of the South-West monsoon. So it depends largely on the contrast between land and sea. So if you define it that way, as a seasonal switch in oscillation, you will find more than one.

Q. And the atmospheric circulation, when it brings moisture, you get rain?

A. Correct. When it comes in from the sea, it brings moisture and you get rain. In monsoon regions like India the general interest is not in the temperature or wind except for heat waves and storms; it is rainfall.

Q. So if you look at Cambodia, Vietnam, Laos…

A. They have South-East Asian monsoon, it is the same thing. 
If you live in England, rainfall is not a major variable. It rains almost every day. It is a drizzle most of the time, so they are worried about temperature. In India, we don’t worry about temperature, because it is warm most of the time. Our variable is rainfall. Rainfall is one of the hardest things to predict. Temperature and wind can be predicted more easily than rainfall. Rainfall, as common experience suggests, is very spotty. It can be raining here in Malleshwaram, but not on the old Airport road in Bengaluru. So, in space and time it is very spotty. 

The last bit of physics required to tell us whether it is going to rain or not is very hard.

Q. There is moisture, which you can measure, but under what conditions will it precipitate? Is that they key unknown?

A. Yes, under what conditions will it precipitate, is the key thing, and how much and where. If you ask a physicist about condensation he will say we know the law, it was discovered more than 100 years ago. At each pressure, there is a particular temperature at which water vapour will condense into liquid water. But, that is not enough. Most of these cumulus clouds, or the kind that you see in Bengaluru and most parts of India during the monsoon, are actually flow clouds. They are bubbling up from the ground. And therefore there is a very complex fluid dynamics associated with it, it is not just thermodynamics. This is fluid dynamics and thermodynamics interacting. So you can very crudely say, a cloud starts over a warm patch on the ground, light air rises and if it is moist, at some stage, it will reach the condensation temperature and condense.

Q. Gets heavier and comes down?

A. No, it doesn’t get immediately heavier. All clouds don’t rain. If the convection is sufficiently strong it will just carry it up. If you keep condensing more and more, at some stage all the water vapour has condensed already. For it to fall as rain, the droplets have to get bigger. Big enough that their weight will make them fall. If they are very small drops, they won’t fall, they will be carried with the wind. Only if the drops are large they will fall against the upward flow within the cloud. And there are many other factors that have not been completely understood to this day. Because, clouds interact in a very complicated way with the air outside. Suppose you induct a lot of air from the outside and you have moist air inside, even if it has condensed the air from outside will dilute the cloud. It is dry air. Therefore, clouds and precipitation are still an unsolved problem.

Q. When moist air in clouds keeps rising it will freeze. Will that produce hailstones or snowflakes?

A. So there are two phase changes that could occur. From vapour to liquid and liquid to solid. Each time the phase changes occur, the flow properties in the cloud undergo a change. Why, because you are suddenly putting heat into the flow. One of the things I have been doing is finding out what happens when you put heat the right amount of heat into flow at the right place. You can make a flow look like a cloud.

Q. So that’s what you meant by interaction of thermodynamics and fluid dynamics?

A. Correct. You have to put the right amount of heat. So my view now, based on our experiments is, you can make a cloud flow and a live cloud plume in any fluid you like--Water, oil, mercury—but subject to one condition: the amount of heat you put in must bear a certain ratio to the energy of the flow in the fluid. You do that and you will get something very similar to what happens in the sky. A cloud is a special kind of plume. That is our conclusion. First of all, we say cloud is a transient plume. It has a finite lifespan.
I have been doing a lot of experiments in the lab and on the computer and it is a lot of fun. 
It comes as a surprise to many people that these things that we see every day, which we admire and love are not understood.

Q. Many people have given up saying it is chaos, butterfly effect, there are too many unknowns etc…

A. You are quite right. The scientific study of clouds started around the 1960s. Very simple experiments were made. Unfortunately, by the 1970s, it was realized that those simple experiments were not behaving like clouds at all. So the attempt to make cloud flows in the lab sort of declined. But you can say we picked it up where they had stopped. 
In the 1980s, I started an experiment in IISc with some students beginning with G S Bhat (who is now Professor at CAOS, IISc) and that experiment turned out to be very revealing. From then on I have been doing clouds in some form or other. We are getting closer and closer to a cloud flow and you can reproduce a wide variety of cloud types in our water tank. At that stage a lot of people started taking notice.
Now I am interested in getting the details right on the computer. In the lab, it is difficult to reproduce all conditions, including water vapour, condensation and so on because the setup in the lab cannot be very big. But, in the computer, I have from one point of view more freedom provided I have access to super computer time. Right now we are making cloud simulations on the computer.

Q. Which computer are you using?

A. That is my big problem. The CSIR Fourth Paradigm Institute (CSIR-4PI) formerly CMMACS (CSIR Centre for Mathematical Modelling & Computer Simulation), Bengaluru acquired a big computer and I made a lot of my computations on that computer. But that is no longer big enough. IITM (Indian Institute of Tropical Meteorology) Pune, has a very powerful computer. IISc has a Cray so, I do some computations there.

Q. I have a stupid question. We had a longstanding traditional Panchanga system, with meteorological applications, and that is what made it important in an agricultural economy. Now that we have 100 years of reliable data from the Indian Meteorology Department, can we compare the actual weather with Panchanga’s predictions and see any correlation and how accurate the predictions were?

A. If you are talking about rainfall, I think the correlations will be poor. In the princely state of Mysuru, there was a Panchanga made with Maharaja’s patronage called the Ontikoppala Panchanga. But, we know that Panchanga depends on the longitude, etc. So, there is no way that it can be precise for the whole of Karnataka. However, this doesn’t mean they knew nothing. They did realize that this rain is cyclical. Varsha (a year) is really rain. Rain is seasonal--once a year. So the word for year and word for rain is practically the same. And they had a rough idea about how long it lasted--3-4 months.

Q.  And, in between cycles also, they have Nakshatras associated with different parts of the rainy season…

A. That is correct. Those are basically empirical and those are correct but they vary from region to region. So I did one thing at IISc though not along the lines you are thinking. I said we should talk to the farmers and see what kind of forecasts they are doing. So we met a few farmers and invited them to Bengaluru at the Institute of Science. They were first of all very pleased that people actually came to them to talk about weather. I had all the Met department statistics, recorded rainfall, converted into Nakshatras which is available on the CAOS (Centre for Atmospheric and Oceanic Sciences, IISc) website. Then we had a long chat, they all said the predictions we get are not good both from Panchanga and the Met department.

Q. Our Met Department doesn’t make detailed predictions. They just say normal, sub-normal, monsoon etc. Whereas in the U.S., every region, every TV channel will tell you 12-hour, 24, 48, 72 hours' predictions and they are fairly accurate. So what makes weather prediction complicated in India?

A. That is a very important question. The dynamics at higher latitudes of Europe and North America are simpler than the dynamics in the tropics. The tropics are basically more complicated. It is not that the British and Americans are better at predicting the monsoons than us. They have made big mistakes time and again. Of course we keep making mistakes but so have they.
We are more interested in the monsoon than anyone else in the world---it affects our economy, it is very important in many ways. So we have to be the foremost experts in the world on monsoon. 
But the best minds in India have not devoted their time to the study of monsoon and they have followed the fashions of the West. And the West will study what is important to them. We can say the Indian scientific community as a whole has not looked at our own problems with the same intensity that the Western World has looked at their problems, although now things are changing. 
What we can say is that with the state of knowledge existing in the world today, not just in India, predicting the monsoon is a tough job. The Indian response has been to borrow unknown American computer codes and so on but our experience with them is not any better. Americans themselves say, “we can’t predict monsoons”.

Q. What about El Nino and so on? Was effect of El Nino on Indian monsoon realized by Indian scientists or they just borrowed it from the West?

A. That is a very interesting question. El Nino was at that time not known by that name at all. The physical phenomenon connected with El Nino was first discovered by a British head of the Indian Met department Sir Gilbert Walker, around 1915-1920. He was a very bright man. When he came here he didn’t know any meteorology but he was fascinated by it. But after a while he decided that the monsoon was too erratic- and he didn’t see how you one could understand the dynamics completely. He said, “I will try and do statistical predictions” and introduced statistical predictions to the Met department.

Till recently, Walker’s philosophy, not precisely his formula but his philosophy, has been what governs predictions in our IMD. It is based on some correlations. There is now Vasant Gowarikar’s model that took many more parameters than Walker’s did, but the philosophy was exactly the same. 
One of the things that Walker found was that monsoon rainfall in India had a link with what happened in Australia. In the days of the British Empire he easily got data from many parts of the British Empire. All the other global networks were not so strong. So he found, for example, what pressure is doing in Port Darwin or Perth has a link with what monsoon is doing in India with a lag. At that time, this idea was unthinkable. Everybody said monsoons come because of the clouds, what does it have to do with what is happening in Perth? In fact, people were critical of him. Which is ironic. Walker didn’t live to see that a couple of years after he died, the connection was established by much better measurements. It left no doubt. That phenomenon is what has grown into El Nino. The connection that Walker saw between Indian monsoon rainfall and what happened in Northern Australia is one signal of El Nino. But it is not only Australia.

Q. So is there a Pacific oscillation just as there is an Atlantic oscillation?

A. This is called El Sona. EL Nino is Southern Oscillation (ENSO). There is an Atlantic Northern Atlantic Oscillation (NAO). More such oscillations have been discovered. If you have an oscillation it is slightly more predictable. Therefore a link with the monsoon can be detected, which is what Walker did.
So you can say it was discovered in India. Not by an Indian, but by a British meteorologist. Of course there were many people who helped him and all the calculations were made by Indians.

Q. The meteorological satellites launched by ISRO to study ocean surface temperatures etc, have they helped in any way?

A. Yes, in a big way. To give you a striking example, look at our cyclone predictions now. Recently, there was this big cyclone over Odisha. Indian Met department got it right because they had overall satellite data. Therefore the cyclone was tracked right from the time it was born. And if you remember in the previous super cyclone thousands of people were killed. This time it was more like 10 or 20 people.
In fact our Met department had also predicted the cloud burst that happened in Kedarnath, Uttarakhand. But no government department took it seriously. And it led to a major disaster. So when this cyclone came to Odisha no bureaucrat and politician wanted to get caught unprepared once again. So they took it seriously and very few people died in Odisha. 
So that kind of relatively short-term prediction we are getting better at. Doppler radar too helps a great deal in short-term prediction whereas our satellites help a great deal over the life cycle of a cyclone. They also help in studying climate change. Now we can track changes that have occurred over the past 20-30 years.

Q. Has computer modelling helped?

A. Computer modelling has helped on the scale of couple of days, but not as much as one would have liked.

Q. Is computer modelling easier in the higher latitudes?

A. By and large it is easier. The clouds are what stump us in the computer models of the tropics. Northern latitudes are not dominated by cumulus clouds; they do not affect atmospheric circulation very much. They are largely static clouds, spread clouds. So that is why clouds are very important to India and that is how my interest in clouds started.

Thanks Prof Narasimha, this conversation has been an education for me.

(Shivanand Kanavi, Former VP of TCS, is Adjunct Faculty at NIAS. He is a theoretical physicist, senior journalist and author. He tweets as @shivanandkanavi and Blogs at http://reflections-shivanand.blogspot.com/  He can be reached at skanavi@gmail.com )