Chennai’s long-standing and perennial water problem perhaps requires a radical solution. But the solution that T.N. Krishnamurti, a world-renowned atmospheric scientist from Florida State University (FSU), United States, who in recent years has been in the limelight with his highly accurate and successful hurricane forecasts in the U.S., suggests is so drastic and so out-of-the-box that to some the idea may even seem crazy, and it could also become controversial.
What he proposes is a large-scale geoengineering solution that would require altering the frontal face or the orography of a mountain which is in the path of the water-rich trade winds from the Bay of Bengal that the southern region receives during the north-east monsoon. “The marine layer in the Bay of Bengal, the moist air which eventually rises, has relative humidity close to 90 per cent. If I look at the satellite pictures, I see this army of stratocumulus clouds coming by. You know what I think? Oh! it’s really about 50-100 Cherrapunjis coming. I see a huge amount of water right there but it doesn’t make it. Let me tell you where it goes. Some of it goes over the Eastern Ghats without raining and some of it goes by the side of the Ghats down to the southern hemisphere,” Krishnamurti says somewhat ruefully.
Krishnamurti’s idea is to maximally exploit the potentially rich meteorological conditions, which present themselves year after year, by providing the optimal geophysical conditions on the ground for intense precipitation to occur locally and that water, he believes, will more than suffice to meet the city’s requirements. It would be a huge one-time engineering effort but could result in a permanent solution.
The average annual rainfall in Chennai is 1,276 millimetres. Chennai receives about 985 million litres per day (mld) from various sources, of which 830 mld is supplied by Chennai Metropolitan Water Supply and Sewerage Board (CMWSSB), against the required amount of 1,200 mld. The sources of water for the CMWSSB include the fresh water reservoirs and lakes at Poondi, Sholavaram, Red Hills, Chembarambakkam and Veeranam, whose storage levels depend critically on rainfall. Owing to the generally poor rainfall in the region, the total storage in these is usually about 30 per cent of their capacities.
The Telugu Ganga Project, or the Krishna Water Supply Project, is supposed to augment this supply by 300 mld. But this is crucially dependent on the storage in the Srisailam reservoir in Andhra Pradesh (500 kilometres from Chennai). This again is a function of the rainfall situation there, which usually is not very different from that around Chennai. The desalination plants at Minjur (commissioned in 2010) and Nemmeli (commissioned in 2013) add another 100 mld each. At present, the shortage of water in Chennai is about 175 mld.
The daily demand is, however, projected to increase to 2,100 mld by 2031. The various schemes by the Government of Tamil Nadu have, however, not really been able to meet the growing demand. As a result, one may see an increased push towards more desalination plants along the coast. This may not be the ideal option, given the high capital costs and environmental issues with regard to the highly saline discharge back into the ocean, not to mention importation of technology and plant operation and maintenance issues.
Enter Krishnamurti’s idea, to which the State government is apparently open. He is at present working with the faculty of the Department of Civil Engineering of Anna University, Chennai, to prepare a proof of concept proposal to the government, but it is not anywhere near planning for implementation, he stresses. “Till now we have only been talking and having email exchanges,” said K. Premlatha, Professor and Head of the Division of Soil Mechanics and Foundation Engineering of the Department. “It is not possible for me to comment right now; the idea and concept are good. But we are yet to do the feasibility study,” she added.
Notwithstanding the potential criticisms—scientific and non-scientific—that Krishnamurti’s proposal could attract, the science behind his idea is interesting in itself and, therefore, merits discussion. His paper that discusses the technical aspects of his proposal will soon be published in the American Meteorological Society’s Journal of Atmospheric and Oceanic Technology .
Waialeale, the inspirationKrishnamurti’s inspiration for the proposal comes from Waialeale, a mountain about 20 km wide and a kilometre in height in the northernmost Kauai Island in Hawaii. It is one of the wettest spots on the earth averaging about 1,167 centimetres of rain a year, which is greater than what Cherrapunji receives (1,140 cm). But only a few kilometres away the rainfall falls dramatically down to 25 cm rainfall a year.
Since 2006 or so, the data on rainfall there are being regularly collected through automatic rain gauges. The most interesting part of this, according to Krishnamurti, is that all this rain comes (about 70 per cent) on undisturbed days. Disturbed days (when one has strong winds and turbulent weather typical of a rainy season) alone could not bring so much rain since Hawaii has very few such days.
“Since there were five-six years of data now, I decided to model this rain using a very high resolution cloud resolving model,” says Krishnamurti. “Undisturbed trade winds come, and they rise over this peculiar-looking mountain that was formed about 6.7 million years ago from a volcanic eruption. The volcano tore its front half (east facing) completely, giving it an armchair-like shape, or a cone with one part of it hived off which makes it conducive to the rising of clouds. Radar images show the passage of streams of stratocumulus clouds and towering cumulus clouds interacting with the Waialeale mountains. That part between cloud base [lifting condensation level in meteorological parlance] to cloud top [base of the inversion level], which is roughly from about 500 m to a kilometre in height, is also interesting from the modelling point of view.”
Krishnamurti used the model to make daily forecasts of the rainfall and the string of forecasts matched the time series of rain gauge data very well. “There is large-scale flow, there is no disturbance and the air climbs the mountain. I could make the forecast almost without errors,” he says. “We could also model a lot more details, such as what type of clouds, what type of buoyancy and what happens to buoyancy. It turns out that buoyancy is stretched by the vertical motion resulting in huge buoyancy that makes very high clouds producing huge rains.”
These undisturbed large-scale flows that rise over a mountain depend on two hydrodynamic parameters, he points out. One is called the Baines number. If the Baines number is < 1, then the flow will go over the mountain; and if it is > 1, then it will only go around the mountain (Fig. 1). The other is a parameter defined as a product of three meteorological factors such as the slope of the mountain from the cloud base to the cloud top level, the speed of the wind impinging on the mountain (which has to be above 8 m/s) and the vertical gradient of moisture. If this triple product has a large positive value, it can result in the production of heavy rain around the mountain, according to Krishnamurti.
Having succeeded in the meteorological modelling of Waialeale rainfall, Krishnamurti made the computer search for other parts of the tropics and subtropics where similar Baines number, similar relative humidity near the cloud base and similar stable weather conditions prevailed. The search found many places that were meteorologically similar but orographically not quite right. “Stratocumulus clouds would come in but produce no rain on undisturbed days,” he says.
“We found one such place near Chennai too, which is the Jawadhu hills [about 200 km away], which is actually an isolated mountain but part of the Eastern Ghats flanked by smaller mountains, where the cloud base and the cloud top are similar and stratocumulus clouds form too. The winter monsoon supplies most of the rains in southern India and yet on undisturbed days there is hardly any rain in the area around the Jawadhu hills. The orientation of the Eastern Ghats is not quite perpendicular to the north-easterly winds. So a lot of the flow slides past it rather than go over it. The Baines number here is much larger than one.”
Virtual geoengineeringKrishnamurti then did some virtual geoengineering with the Jawadhu hills. The total width of the Jawadhu hills is about 25 km but the effective width that the north-east monsoon sees is about 12 km and its height is about 1 km. What he did was to take the front face of Waialeale, place it over the frontal face of the Jawadhu hills and turn it around to make it perpendicular to climatological north-east monsoon flows. “I put it there and computed the triple product. It was more than satisfactory. I also found that, on disturbed days, the Baines number was less than 1, resulting in an uplift of air and formation of heavy rain-bearing clouds,” says Krishnamurti. He did a numerical simulation of rainfall with the altered shape profile of the Jawadhu hills and found that it began to produce 2-4 cm of rain per undisturbed day of the year. When he ran a large number of numerical experiments, he started producing even 5 cm of rain a day (Fig. 2). Also, like in Waialeale, he found that the heavy rain was localised and a little distance away there was no rain.
From a geoengineering point of view, it is not the entire mountain but only the part between the cloud base level and the cloud top level (500 m – 800 m) that would be involved, according to Krishnamurti. “The shape of the mountain between these two levels is most important for the vertical growth of clouds and heavy rains,” he points out. Now north-easterly winds come making an angle of 15° with the eastern direction and impinge on Jawadhu’s face at nearly 35° from the east. This is what necessitates the rotation of the face.
Now, the geoengineering that is envisaged in Krishnamurti’s proposal is to build this front face first and fill the gaps between the old Jawadhu’s eastern face and the newly built rotated front face, with cement or concrete or whatever. The extension in the horizontal direction where geoengineering would have to be done, as suggested by the numerical experiments of heavy rains at Jawadhu, will be roughly 6 km on either side of the tallest mountain, according to Krishnamurti. From the orographic effect resulting from the reshaped contour of the hill’s front face, even with 2.5 cm of rain a day over a 16 square km area around, if harvested entirely, would provide about 412 mld, according to Krishnamurti, which can go to meet a substantial part of the water deficit that Chennai faces today.
But the sheer scale of geoengineering being envisaged seems to be mind-boggling. Because, in terms of the dimensions involved, it would mean erecting a wall (with concrete perhaps) that is more than 500 m tall and about 12 km wide at the appropriate angle to the front face of the mountain. That would be a structure nearly as long as the longest dam in the world and taller than the tallest building in the world!
“This structure would indeed be more than three times the height of the tallest building of Dubai,” Krishnamurti clarifies. “But this is not to be compared with a dam because the structure one has in mind is not for stopping or slowing of a continual flow of river water. It is not standing on its own feet; it is an extension of an existing hill. The newly built structure would lean on the existing hill and would become a part of it, with the caveat that a front windward facing shape would mimic the Waialeale, especially between the cloud base and the cloud top. The data for the shape of Waialeale is available from USGS [U.S. Geological Survey] at a resolution of 200 m. That is what was used in the numerical modelling. Below the cloud base level the shape is not critical for geoengineering,” Krishnamurti added.
A civil engineer friend of Krishnamurti in the U.S., who has worked with the famous Feather River Project involving a lot of masonry work with concrete, told him that this kind of engineering was possible. He also apparently recommended building a holding tank, stretching across the entire geoengineered front face of Jawadhu (semicircular in shape) and a few kilometres in front, at about 400 m above sea level to harvest the rainfall efficiently. A pipeline running from this holding tank to Poondi lake, which is a distance of 130 km (Fig. 3) but only 40-50 m above sea level will drive the water flow by gravitation alone.
The rough cost of the project, according to the engineer who carried out a typical Indian costing estimate, works out to around Rs.1,000 crore, compared with Rs.600 crore for a 100 mld desalination plant. An Indian engineering consultancy firm based in Chennai, which has been looking at the proposal, has, on the other hand, estimated the cost to be about Rs.2,000 crore. But this will be a one-time cost, compared with Rs.600 crore for the initial capital cost alone for one 100 mld desalination plant. The system can obviously operate only during the four months of the winter monsoon. But is it guaranteed that this will happen every day of the monsoon?
On the basis of his simulation experiments, Krishnamurti believes that this will happen nearly on all undisturbed days. On disturbed days, all reservoirs will get a lot of rain because, according to Krishnamurti, the north-east monsoon is relatively more moist and more robust compared with the south-west monsoon. As in Hawaii, such conditions will last for a couple of days and steady trade regime will re-establish itself, Krishnamurti says.
“We understand so little about the role of mountains in precipitation, I am not sure if one can guarantee that what he claims will happen,” said J. Srinivasan, a well-known atmospheric scientist from the Indian Institute of Science (IISc) in Bangalore. “But I have not studied his proposal to make any further comments on it,” he added. “I am sure there will be plenty of doubts and questions,” says Krishnamurti.
“I am highly sceptical because you have to examine such an idea with different models to see that the results are robust,” says Sulochana Gadgil, a monsoon expert formerly with the IISc, who had heard Krishnamurti give a presentation on this in Bangalore last year. “Simulation with just one model is not sufficient to make these claims or to think of implementing a big meteorological project on that basis. The idea is totally untested anywhere. And I would say Rs.1,000 crore is a huge sum of money to implement something that to me is extremely uncertain to work. And if it doesn’t work, then what?”
“Modelling rainfall itself is not the main issue,” said Ravi Nanjundiah of the Centre for Atmosphere and Oceanic Sciences (CAOS), IISc. “My concern is from a larger ecological impact that such engineered heavy local rainfall can have. For instance, what does trapping all the moisture to a localised region do to areas leeward from there? Will they become more arid? What would happen to the biodiversity of the region? And, of course, there would be local farming practices based on existing rainfall patterns, and one needs to know how they will adapt to a changing ecology. I think a much more holistic, total ecological impact assessment has to be done before any geoengineering is planned,” he added.
Krishnamurti says, “I would like my proposal to be critically examined by scientists. Maybe one should set up a scientific committee which can ask all sorts of scientific questions. We can run numerical experiments on the computer day and night and have the questions answered.”
COMMents
COMMents
SHARE