The monsoon story

This year, the effect of the north-east monsoon has been worsened by an intense El Nino and a quick succession of rain-causing weather systems.

Figure 1. A schematic showing "normal" and "inverted" troughs. The north-east monsoon is characterised by inverted troughs.

AS FRONTLINE goes to print, Chennai has some respite from the unprecedented deluge that has enveloped and totally crippled the city. The yearly north-east monsoon (NEM), which brings rain to the southern peninsular region of the country during the October-November-December (OND) months, has brought an almost continuous heavy downpour over the last 10 days or so. The city received the season’s heaviest rainfall on a single day on December 1. In the 24-hour period ending at 0030 hrs on December 2, (the meteorological station at the Meenambakkam airport in) Chennai had received a whopping 345 mm of rain. The quantum of rainfall during December 2-3 dropped to about 28 mm and even further to 15.8 mm on December 4. The city alone received only about 5 mm of rainfall.

The weather system that brought these torrential rains was a “trough of low pressure”, which is still hovering over the south-west Bay of Bengal (BoB) but has transformed into a pressure low over the same area about 250 km off the Tamil Nadu coast.

This morphed pressure low is currently anchored south of Chennai and closer towards Sri Lanka. So the episodes of intense precipitation too have moved southward, affecting the central coastal districts of Tamil Nadu and the Union Territory of Puducherry. The districts of Puducherry down south to Karaikal, Cuddalore and Nagapattinam have borne the brunt of the heavy rainfall in the past couple of days.

(A trough of low is an elongated region of relatively low pressure (Figure 1). Actually, during the NEM, there is an “inverted trough”, which is oppositely oriented to the normal trough that is characteristic of the south-west monsoon. While in normal troughs of low (V-shaped in the northern hemisphere), the atmospheric pressure decreases from south to north, in inverted troughs (inverted V shape) the pressure decreases from north to south. So there is an increased flow of north-easterly winds that are guided along the arms of the inverted V and directed towards coastal areas, bringing heavy rains along the coast. Since a trough is a much more extended system than a pressure low, the transformation means that the system has intensified and is more localised. In fact, the genesis of tropical cyclones often occurs under the influence of troughs of low.)

According to the data of the India Meteorological Department (IMD), between December 1 and December 4, Chennai received 1,522.7 mm of rainfall against a normal of 662.6 mm (130 per cent excess); Cuddalore 1,041.5 mm (normal 550.1 mm/89 per cent excess); Kancheepuram 1,681.4 mm (normal 535.1 mm/214 per cent excess); Tiruvallur 1,414.0 mm (normal 496.9 mm/185 per cent excess); and Puducherry 1,406.2 mm (normal 673.5 mm/109 per cent excess).

Wait-and-watch situation According to the estimate of flooding rainfall by the United States’ National Aeronautics and Space Administration’s (NASA) Integrated Multi-Satellite Retrievals for GPM (IMERG), from November 28 to December 4, over 400 mm of rainfall fell over areas south of Chennai (Figure 2). The highest rainfall totals were estimated to be over 500 mm in an area just off the south-eastern coast of India. Mercifully, therefore, the heaviest rainfall occurred over the bay rather than over coastal land. (Global Precipitation Measurement, or GPM, is a constellation of eight satellites that gathers rainfall data at 10 km resolution at hourly intervals.) So, is the worst over for the city? Most probably yes. As the satellite image (Figure 3) shows, the rain belt has moved towards Sri Lanka, which has started receiving heavy rains. However, it may still be a wait-and-watch situation for people in the Chennai region for at least until the weekend of December 5-6. If the slow-moving low-pressure system moves inland west-north-westwards, it could again bring intense rain over the Chennai region until it weakens as it moves westward over land and completely dissipates. If it moves westwards, it will bring rain to interior Tamil Nadu as well. Until then, heavy rainfall will be restricted to the central regions of coastal Tamil Nadu as is the case as of December 4. Chennai may, however, continue to receive the normal rainfall characteristic of the season that is brought by the north-easterlies.

Actually, for a week or so before December 3, it was a double whammy for Chennai and the adjoining areas as the coast was struck by the enhanced north-easterly flow due to the trough and the strong easterlies due to the El Nino currently prevailing over the central Pacific. “This was a battle between the north-east monsoonal flows over the Bay of Bengal and the El Nino triggered easterlies,” said T.N. Krishnamurti, a monsoon expert from Florida State University. “The latter won until December 3, and now the north-east monsoon is paving its way back and the rains over Chennai have abated, the rain belt has slowly migrated to Sri Lanka.”

El Nino factor Meteorologists have cited El Nino, which is the warming of the waters in the central and eastern Pacific, as one of the reasons for the unprecedented rainfall over Tamil Nadu this season. The phenomenon is known to be negatively correlated to the summer, or south-west, monsoon and is one of the leading factors causing seasons of deficient summer rainfall and drought. But it is positively correlated to the winter, or north-east, monsoon that brings rain to the States of south India during the OND months. The five meteorological subdivisions of south India—Tamil Nadu, coastal Andhra Pradesh, Rayalaseema, Kerala and south interior Karnataka—receive about 30 per cent of their annual rainfall in this season. Tamil Nadu in particular receives nearly 50 per cent of its annual rainfall in OND.

This is one of the important reasons why the IMD forecast a significantly above normal rainfall this year when there is a very strong El Nino prevailing. The forecast (issued on October 16) was 111 per cent of the long-period average (LPA) for the southern peninsula and 112 per cent above the LPA for Tamil Nadu in particular. The LPA rainfall for Tamil Nadu for the base period 1951-2000 is 438.2 mm. Already, the total rainfall for Tamil Nadu from October 1 until December 4 is 592 mm (a 60 per cent excess over the LPA of 369 mm for the period up to December 4; Figure 4). The El Nino this year has been very intense, perhaps the strongest in recent history, with a sea surface temperature (SST) anomaly of nearly 3 °C in the central Pacific region called Nino 3.4, which is known to be strongly correlated to the monsoon regime over India (Figure 5). “Lower tropospheric sustained easterlies emanating from the El Nino stretch well across the Bay of Bengal, which lead to storms over the warm bay,” Krishnamurti said (Figure 6). “These storms form from what we call shear flow instabilities supported by organised convection. That was a storm that produced the heaviest rains on December 1 and 2 over Chennai,” he said.

Chennai has received very heavy rainfall in normal years, El Nino years and La Nina (opposite of El Nino when there is a cold anomaly over the Pacific) years, and while 345 mm of rain is indeed huge, the rainfall that Chennai received on November 25, 1976, was much greater at 452.4 mm, which was a normal year. “El Nino is only one of the factors affecting rainfall during the north-east monsoon,” said S.R. Ramanan, Director, Area Cyclone Warning Centre at IMD (Chennai Division). “What really matters is the frequency, amplitude [the spatial extent] and intensity of the rain-causing weather systems that form within this large monsoon flow. This year, we have had systems forming one after the other in quick succession and it is the cumulative effect of that,” he added.

Five weather systems During the NEM, in a month, the average number of such weather systems is three, according to L.S. Rathore, the IMD Director General in Delhi. “In November, we have had five systems one after the other, four of which were strong and one weak,” he said. It began with an upper air cyclonic (anticlockwise) circulation over the equatorial Indian Ocean and the adjoining south-west BoB towards the end of October, under whose influence a “low-pressure area” formed around November 5 bringing with it rains to coastal Tamil Nadu. On November 7, it intensified a little more and became a “well-marked low-pressure area” causing heavy rainfall. And on the same day it intensified further into a “depression” over the south-west BoB about 460 km south-east of Chennai and 440 km east-south-east of Puducherry.

On November 8, it moved west-north-westwards and intensified into a “deep depression” and in the forenoon lay centred over the south-west BoB, about 60 km east-south-east of Puducherry and 150 km south-east of Chennai. It crossed the north Tamil Nadu coast near Puducherry on November 9 evening. It moved slowly westwards, and on November 10, weakened into a depression over Tamil Nadu about 80 km east of Tiruppattur and 60 km south-south-east of Vellore. It further weakened into a “well-marked pressure area” and further into a “low-pressure area” over Tamil Nadu and adjoining southern Karnataka on November 11, bringing heavy rains all along its trajectory (Figure 7).

Immediately after that, on November 12, a trough of low formed over the south Andaman Sea with an upper air cyclonic circulation and moved westwards and lay over the south-east BoB and neighbourhood on November 13. Under its influence a low-pressure area formed over the same region on November 14. On November 15, it transformed into a “well-marked low-pressure area” and lay over the south-west BoB and adjoining Sri Lanka. On November 16, the system moved slowly and located off the north Tamil Nadu coast, with an upper air cyclonic circulation extending up to mid-tropospheric levels and tilted south-westwards on November 17 off the south Andhra Pradesh-north Tamil Nadu coast. This system brought with it prolonged heavy rains. In fact, on November 16, Chennai received 246 mm of rain, according to NASA’s GPM-based IMERG data.

Short-lived let-up Then what seemed to be a let-up in the rains was short-lived as a prolonged trough of low formed on November 26 centred over the south-west BoB and adjoining Sri Lanka coast with cyclonic circulation. Another trough of low that formed over the south-east BoB and the adjoining south Andaman Sea moved westwards and merged with the former, resulting in the weather system that brought the torrential rains in end November and early December, the heaviest rainfall in 24 hrs over Chennai on December 1-2.

All these systems caused heavy to very heavy rainfall, particularly over the coastal regions. “The north-east monsoon is characterised by slow-moving low-pressure weather systems,” Ramanan said. “Unlike a cyclone or even a deep depression, which moves quickly across over land causing intense rainfall for a few hours, slow-moving systems hover over a region for a long time causing heavy rainfall for two to three days. This year, we have had those persisting even for four days and they have come one after the other. What we are seeing is the cumulative effect of all these,” he said.

The pattern in 2005 too, he said, was similar with pressure lows, well-marked lows, troughs, a deep depression and many weather systems forming one after the other. The cumulative rainfall for the city of Chennai that year was about 2,110 mm. And, interestingly, 2005 was a neutral/La Nina year, not an El Nino year. In 1997, the strongest El Nino year until the present one—it had an anomaly of 2.8 °C—the total rainfall up to December 2 was 1,505 mm. This year, as of December 2, it was 1,558.9 mm against a normal of 634 mm until that date. The (redefined) normal for Chennai district as a whole for the entire season (up to end December) is 790 mm; for the city alone, it is 856 mm.

So, while El Nino is indeed a major factor for the NEM, it can be offset by other meteorological factors that influence rainfall. In El Nino years, the warm pool over the Pacific elongates and extends into the eastern Indian Ocean and also, as mentioned earlier, the easterlies are much stronger in the Indian Ocean. The two phenomena together lead to conditions favourable for the formation of many low-pressure systems that can grow into cyclones or strong low-pressure systems with widespread rainfall. This is what seems to have happened this year.

According to Krishnamurti, the El Nino-NEM link this year is particularly different because of the presence of a very warm SST over the BoB (Figure 8).

“The SSTs this year are near 30 °C. That helped convection within the vortices induced by shear flow within the easterlies to develop more. That feature is not seen in all El Ninos and this is significant in the context of NEM rainfall over coastal Tamil Nadu,” he said. (The panel in Figure 9, based on GPM measurements, depicts the movement of the precipitating elements from the Pacific to the BoB over the week November 28-December 4.) Also, according to some recent studies by scientists of the Indian Institute of Tropical Meteorology (IITM), Pune, the El Nino-NEM link has become stronger in recent years.

While the above describes the formation of the pressure lows, what about the formation of the troughs of low that have added to the adverse impact of the monsoon? According to Rathore, the formation of the north-south (inverted) troughs of low is because of the prevailing westerlies in the north (which this year have not moved significantly southward) and the thermal difference between the warm waters over the BoB and north India. The meeting of these two fronts results in the development of troughs of low.

Although 2005 was similar, it was not quite the same. There were always gaps of four to five days between successive episodes of intense rainfall. Indumathi, a professor at the Institute of Mathematical Sciences (IMSc) in Chennai, distinctly remembers that the IMSc was completely flooded that year. “I have never seen such rains before. My kids have not gone to the school for nearly a month now. In 2005, the schools were shut for a maximum of a week or so. This year, it is something else,” she said. Indeed, the enormity of the rainfall impact this year is much more as is clear from the fact that while Frontline did not do a cover story on the rain that year it is doing so now.

Therefore, although Chennai has had intense rainfall in the past during the NEMs, including a much greater quantum over a 24-hr period, the unusual feature this time is rain-causing weather systems forming in quick succession. The question whether all this is due to ongoing global warming and climate change is moot. “We cannot say this until we run climate model simulations for 100 years (with and without CO ₂ increase),” said J. Srinivasan of the Indian Institute of Science, Bengaluru.

Potsdam study A recent study at the Potsdam Institute for Climate Impact Research, Germany, has found a clear upward trend in the past few decades towards more unprecedented daily rainfall events, particularly short-term torrential rains that can lead to high-impact flooding. In particular, an analysis by M. Rajeevan and others at the IITM has found an increase in intense rainfall events associated with the NEM over the past six decades (Figure 10). The average worldwide increase, according to the Potsdam study, is 12 per cent, which is consistent with rising global temperatures.

“The accumulated heavy downpour from the sequence of events has saturated the soil,” Srinivasan said. In fact, the diminishing floodplains available for overflows from the Adyar and Cooum rivers that run through the city because of encroachments and construction on those plains and the poor drainage and sewerage system are two of the major reasons for the city being drowned in water. Also, the management of water supply to the city from the four lakes that are its main water sources leaves much to be desired (see Frontline, December 11).

‘Not rocket science’ “What is sad is that we don’t get tired of seeing the same movie again and again. Every time there is a flood, it is the same scenario: heavy rains initiate them but they are made disastrously worse by poor management of lakes and reservoirs, which are opened at the last minute to dump lots of water, creating catastrophic floods, far more than what would have occurred from just the rains,” said Balaji Rajagopalan of Colorado University, an expert in hydrology and water resources. “My work is extensively on the Colorado river basin, which is semi-arid and has similarities to water resources of the Indian subcontinent. Several lessons can be learned,” he said.

“The management of the lakes and reservoirs requires a decision support system coupled with weather and hydrologic forecasting systems,” he added. “What frustrates me is that this is not rocket science and is routinely done in the U.S. and other places. Weather forecasts in conjunction with hydrological models can provide forecasts of inflows to the reservoirs. This information is then used in a decision support system to make releases. This is adapted as the forecasts evolve. Of course, urban planning, maintenance of drainage ways will significantly mitigate the impacts. Cities should plan for such scenarios by simulating such events to see the impact so that they can plan disaster responses,” Rajagopalan said.