Tropical cyclones (TCs) occur over the warm tropical oceans of north Atlantic, western North Pacific, eastern North Pacific, northern Indian Ocean (NIO), south Indian Ocean and the south Pacific Ocean. The average combined frequency of occurrences of TCs is about five in a year over the Bay of Bengal (BoB) and the Arabian Sea (AS) of which about four form over the BoB and one over the AS. The phenomenon is the same over all tropical ocean basins though geographically different names are assigned. Indian ships have been traversing over the BoB from the beginning of the Christian era to 10th century AD and narratives about TCs in the BoB have always been common in Indian folklore.
When Europeans started sailing the Indian Ocean in the 16th and 17th centuries, many ships were lost to TCs in the south and northern Indian Ocean basins. Accounts left behind by mariners were well utilised by the East India Company to win a war with France by effectively saving their ships from a TC moving towards Chennai, in the early part of the 18th Century, which led the British to control India. In 1768, within a decade after the East India Company took control of Bengal, a very severe TC hit Kolkata taking a large toll of human lives—about 3,00,000 as per one estimate. To Henry Piddington, an employee of the East India Company in Kolkata, we owe the earliest scientific studies about TCs in the NIO. Based on accounts of ships experiencing TCs, Piddington published a series of about 50 papers during 1839 to 1858 in the Journal of the Asiatic Society of Bengal. He had coined the term cyclone from a Greek word ‘Kuklos’—meaning coiling of a snake, and the word has since been used to describe the TCs in the region. In 1864, Piddington authored the ‘Sailors Horn book of Storms’, which guided sailors on how to tackle a TC during a voyage.
Establishment of cyclone warning system
A severe TC struck Kolkata in October 1864, resulting in over 80,000 deaths and a colossal loss to shipping. It was followed by another severe TC that struck the Odisha Coast. The two events prompted the Bengal Chamber of Commerce to recommend the establishment of a regular storm warning system over BoB. British India consequently established the first storm warning system in 1865 for Kolkata Port followed by creation of the India Meteorological Department (IMD) in 1875. The storm warning system has undergone a lot of improvements and revamping since then. At present, India’s storm warning system comprises four-stage warnings and is organised into (i) Area cyclone warning centres at Kolkata and Chennai for the BoB, (ii) Area cyclone warning centre, Mumbai for the AS, and (iii) Cyclone warning centres at Bhubaneswar (Odisha), Visakhapatnam (Andhra Pradesh) and Ahmedabad (Gujarat). The whole system is coordinated from IMD, New Delhi by the Cyclone Warning Division.
Growth in research
Sir John Eliot, who discovered that TCs form on the advancing and retreating edges of the southwest monsoon wind system, initiated research on TCs on the NIO between 1875 and 1903 and discovered that the motive force that maintains them is the latent heat, which is released in copious rainfall accompanying a cyclone. In the period between 1903 and 2014, Indian researchers made significant contributions to the knowledge of TCs, though several of their ideas followed work done in developed countries. Researchers in USA and Japan acquired advanced original knowledge on TCs rather rapidly due to their access to superior technology and tools like weather radar, aircraft reconnaissance, satellite monitoring, computing and enhanced research fund allocations to universities and research organisations. India, which was once at par with the developed world until 1940, was left lagging behind for over 70 years until recent times.
In the last decade, advanced observational systems and high speed computing has provided India an impetus. Indian researchers could also establish the similarity of TCs in the NIO with hurricanes over the North Atlantic basin, when working with American scientists on the International Indian Ocean Expedition (IIOE).
Advancement between 1940 and 2010
Observational systems: Weather satellites are being used in the USA and India since 1960. The automatic picture transmission system was established in 1965 at Mumbai. At present India has its own satellite INSAT in geo-stationary orbit—36,000 km above the earth, and also uses polar satellites, ocean satellites and others to get specialised information on TCs. For visible and infra-red bands, a two decade long research in the USA, between 1965 and 1984 helped finalise a scheme for estimating the intensity of a TC based on cloud pattern. This, called the ‘Dvorak technique’, is operationally used in India to determine cyclone intensity. The in built uncertainty level in this technique in determining the location of the centre of a cyclone is about 25 km. Since this uncertainty is embedded, the track prediction takes it into account. Coastal radars are used to track the movement and intensity of TCs. India established overlapping weather radars between 1970 and 1980, and these have been recently converted into Doppler radars along the eastern coast of India. A radar picture of a cyclone shows the characteristic eye, spiral bands, vertical extent of clouds, and rainfall rates in clouds.
Dynamics of TC formation: It was in the mid-1960s when two groups of US scientists, working independently, discovered the theoretical basis for the intensification of a weak tropical weather disturbance into a TC. They found that under ordinary conditions, tropical atmosphere supports convection on a small scale (1-10 km). However, a cooperative mechanism works when there is cyclonic vorticity at the top of the atmospheric boundary layer such that the latent heat released on the small scale supports the warming of the overlying atmospheric column of the large scale (1000 km). Favourable interactions between the large and small-scales lead to intensification. When this theoretical concept was used over a warm tropical ocean under a rotating atmosphere, a cyclone-like vortex was simulated and relevant equations solved using a computer. Between 1965 and 1980, several research groups in USA, Europe and Japan could simulate TC structures realistically, and solve the major problem of tropical cyclogenesis through computer simulation. By the 1990s, scientists in advanced countries had begun using high horizontal resolution atmospheric models (50 km lesser resolution) for weather prediction. These models, when fed with observed data, could predict formation of a TC-like structure. This was the beginning of real-time realistic dynamic prediction of TCs over warm tropical oceans. By 1990, scientists in USA had also developed meso-scale regional models (grid resolution 25 km and lower) vastly improving the prediction and movement of TCs.
Since 2006, Indian scientists have been using high resolution global and regional models to operationally predict the formation and movement of TCs. Research in the last two decades has further shown that there is a third scale meso-scale involved in TC formation. Thus, TC formation is a complex interaction of three weather scales— the small, the meso and the large-scale environment.
Advances in prediction: By the 1930s, the method used for TC prediction depended on synoptic data with judicious combination of climatology. The cyclone was supposed to be steered by a combination of horizontal winds of 6–14 km height. There were large errors in such forecasts and they were not usable beyond 48 hours. Until 1978, objective techniques based on persistence, analogue and climatology were used with some success for up to 48 hour forecasts. The non-divergent barotropic model developed in USA in the 1950s continued to be used for TC prediction in India until the 1970s. With increased computing power, Indian researchers and operational establishments now use a suite of models varying from meso-scale to global scale to predict TCs. India has launched a major research-cum-operational programme known as the Forecast Demonstration Project of the Bay of Bengal TC Experiment (FDP-BoBTEX) since 2008 in which observational systems are enhanced to provide data for the prediction of TCs. Statistics, collected over the period 2003-2013 has shown that the uncertainty in the prediction of TC track is now reduced to about 125, 200 and 270 km for 24, 48 and 72 hour forecasts, respectively. IMD uses multi-model ensemble forecast using predictions from advanced countries’ models. However, with aircraft reconnaissance, track errors could decrease further.
From the early definition based on the increasing sustained wind speed criteria, a more precise definition of a TC has been adopted. A TC can be now defined as a warm core system of at least 5o diameter which extends vertically through most of the troposphere with a little vertical slope and a cyclonic relative vorticity (a measure of the rotation of the vortex) of at least 10-4 per sec in the inner core region (about 100 km radius from the centre).