Farecasting the cloudburst
(appeared in Oct 2017)

(link to main website)

Extreme climate events have been happening more often, but we could hope to get early warning, says S.Ananthanarayanan.

Loss and destruction by floods and heavy rains may not be unusual in India, but sudden downpours that paralyse whole districts appear to have become frequent. Each year gives us instances that are worse than those of previous years and this trend, of growing extreme events, has been underlined by the IPCC, by writers and researchers.

M.K. Roxy, Subimal Ghosh, Amey Pathak, R. Athulya, Milind Mujumdar, Raghu Murtugudde, Pascal Terray and M. Rajeevan, at the Institute for tropical meteorology, Pune, IIT-Mumbai, at Kochi, University of Maryland, the Sorbonne and the Ministry, Govt. of India, report in the journal, Nature Communications, a study of how extreme events are increasing despite falling average rainfall. The study discovers features that could help forecast these events as early as two to three weeks before they occur.

The losses caused by floods that are attributed to extreme rainfall events in India, during the last decade, were about US$ 3 billion a year, which is a whole 10% of the losses worldwide, the paper says. There were 268 flooding events reported in India during the 65 years from 1950-2015, affecting 825 million people, and a good part of the flooding events took place over central India, the paper says.

The paper displays the trend of rainfall events in India in a panel of graphs. As the first picture in the panel shows, it is over central India that there is much rainfall, and hence agriculture that depends on rainfall. The next picture shows that it is in this belt that we have had most of the extreme rainfall events. The two graphs in the middle show that these extreme rainfall events have been getting more frequent, in fact, a threefold increase, since the year 1950. And yet, the next graph shows, the average rainfall, since 1950, in this region where 60% of agriculture depends on rain, shows a decline of 10-20%. The fall in average rainfall, the paper says, has been attributed to factors that include warming of the Indian Ocean, El Nino events, air pollution and changes in land use. A suggestion that the extreme rainfall events during the period are linked to global warming is still not established. Some studies, however, suggest that high moisture content resulting from warming of the ocean is responsible and there is an understanding that implicates the low-pressure zones that form in the Bay of Bengal.

The evidence, however, as the last figure in the panel shows, is that the low-pressure events have been fewer during the period and there has been less humidity. This raises critical questions, the paper says: “How is the increase in the frequency of extreme events sustained despite a weakened monsoon circulation and a decrease in the number of depressions over central India? Is the moisture content increasing despite a declining monsoon, and if so, what ensures the availability of moisture to foster these extreme events? Is it that the depressions are bringing in more moisture even though their frequency has declined?”

The study has revealed some features that point to answers to these questions. For one, it is found that it is not local conditions, like temperature, that are linked to extreme events. In fact, the paper says, there is a ‘negative correlation’ between the number of extreme rainfall events and the surface temperatures over central India. On the other hand, it is found that there is a connection between the trend of such events and the temperature in other areas, particularly over the Indian Ocean. It has been suggested that rise in the temperature over the Indian Ocean may lead to rise in water vapour in the atmosphere and hence more moisture over central India. The humidity levels over the Indian Ocean, however, are not statistically linked with extreme rainfall events. And it is seen that the warming and increased moisture over the ocean has not led to more rainfall over central India. The dynamics involved are obviously complex, the paper says.

To track down where the moisture for extreme rainfall events was coming from, the authors of the paper turned their attention towards what they call ‘widespread extreme rainfall events’, or when extreme events occur over a sufficiently large area and can cause largescale floods. As the study also called for high resolution data, the study made use of the satellite data available after 1982. And the study consisted of examining the daily maps of moisture levels and winds during the days preceding extreme events.

The results were clear evidence of a build-up of humidity over the northern Arabian Sea and then being blown by winds from the west towards central India. The unusual humidity over the Arabian Sea rises till 6 days before an extreme event and is then blown eastward, leading to heavy rainfall. Analysis of where the moisture that ends up as rain in central India comes from leads to the Arabian Sea being the largest source, followed by local evaporation, while the Bay of Bengal and the Indian Ocean bring up the rear.

The extreme events over central India are hence seen to be preceded by rising sea surface temperatures in the Arabian Sea, about 2-3 weeks before the events. While this leads to rising air columns and higher humidity, there is also warming of the land mass in Pakistan and north-western India. The result is a south-to-north pressure gradient, and thanks to the west-to-east rotation of the earth, there arise winds that blow from the west to the east. The timing of the rise of sea surface temperature is seen to correspond to the westerly wind and rain events that occur 16-19 days (2-3 weeks) later over central India.

This trend of rising Arabian Sea temperatures preceding rain events is borne out historical data, for instance, the central Indian floods in 1989 and 2000, Mumbai floods in 2005, South Asian floods in 2007. There are other studies of the Mumbai floods of 2005 that linked the intense rainfall with an active moisture flow from the Arabian Sea, consistent with the present study, the paper says.

The paper concludes with the observation that anomalies in the surface temperature of the Arabian Sea are likely to increase, and hence the extreme rainfall events over central India. The correlation of these events, which occur uniformly over a wide area, with rise in sea temperature, however, makes it possible to make forecasts with a 2-3-week lead time. The paper notes that the cost of collecting data to implement early warning is miniscule, compared to the losses that arise because of the floods.

------------------------------------------------------------------------------------------

Do respond to : [email protected]