[First published on 18 December 2020; Nature India Spotlight on Odisha]

Flooding in Odisha is frequent and disruptive. © Dibyangshu Sarkar/AFP/Getty Images

Cyclones lashing Odisha’s coasts have been severe to very severe in their devastation.

The height of cyclonic catastrophe in recent times came with the notorious Super Cyclone of 29 October 1999, when wind speeds hit a furious 260 kilometres an hour. It was the most powerful recorded cyclone of the century. Bhubaneswar saw torrential rains of up to 43 cm in just 24 hours. The wind was so powerful that the recorder at the coastal town of Paradip broke down after reading speeds up to 160 km per hour. Power supplies failed, crippling radar observations at Paradip, and telecommunication was disrupted for a significant period, so civic administration was paralysed. Official records say 9885 people were killed and property worth billions of rupees destroyed, though unofficial estimates put those figures way higher.

In contrast, later events — very severe cyclones such as Phailin (2013), Hudhud (2014) and Titli (2018), extremely severe cyclone Fani (2019), and super cyclone Amphan (2020) – did not see similar human or livestock casualties, or property losses. Comparing situations pre- and post-1999 reveals that use of scientific and technological interventions adequately prepared the administration to minimize losses.


Timely and accurate forecasting of any weather related disaster is key. However, forecasting is not the only component that helps cope with hazards. There are other important dimensions, such as science-based early warning, disaster preparedness and response systems. Scientific awareness among the public, and their involvement, are also crucial.

Before the 1999 super cyclone, the weather agency India Meteorological Department (IMD) used traditional weather charts and observational products to forecast the event. Accuracy of forecasts depends on the available information, the Numerical Weather Prediction (NWP) model used, and the experience of its forecasters. In the late 90s, cyclones developing over the Indian oceans were usually analysed using weather information from ships, the buoys deployed by the National Data Buoy Program (NDBP), special radar at Paradip, INSAT-1D satellite imagery and an NWP model called the Quasi Lagrangian Model (QLM).

During the 1999 super cyclone, adequate data was unavailable from ships and ocean buoys, except one stationed off Paradip. On the morning of 29 October, the Cyclone Detection Radar (CDR) at Paradip stopped recording. This was an analogue radar with data analysed on the screen by the local meteorologist and transmitted in coded form through a telegram, phone or teleprinter. Its photographic film was damaged and the wind speed recorders at Paradip became unserviceable.

Flooding in Odisha is frequent and disruptive. © Dibyangshu Sarkar/AFP/Getty Images

The most significant limitation of QLM was the lack of empirically generated vortex and steering current data. As observations were limited, crucial information on the exact location and size of the storm, its central pressure, the value of the outermost closed isobar, was compromised. Though the INSAT-lD satellite images were useful in estimating the eye of the system, due to limitations in getting timely observations of the NWP model, meteorologists were only able to provide a 24-hour warning, leaving little preparation time. The IMD and other agencies were ineffective when the cyclone hit.

The public communication system also failed to either swiftly disseminate early warning to people or amp up emergency response. There was inadequate capacity for evacuation. The 480-kilometre coastline of Odisha had only 23 cyclone shelters made jointly by the Indian Red Cross Society, and the Odisha government.

Combatting extreme weather

Odisha is also prone to heatwaves, drought, floods, intense storms, and lightning strikes. The heatwave of 1998 was one of the most severe in recent history, killing 2,042 people. This event, coupled with the 1999 super cyclone devastation sparked a shift of approach by the government, from disaster response to disaster prevention. A new agency, the Odisha State Disaster Management Authority (OSDMA), was created in 1999. Twenty units of the Odisha Disaster Rapid Action Force (ODRAF),have been established. One unit of the National Disaster Rapid Action Force (NDRF) is also stationed at Mundali in Odisha’s Cuttack district.

The state now has 879 multipurpose cyclone and flood shelters, some multi-storey, to house evacuated people and cattle during a calamity. Based on the Disaster Management Act, 2005 of the Government of India and the subsequent National Policy on Disaster Management 2009 (NPDM), the National Disaster Management Plan (NDMP) provides a disaster management framework. NDMP is a dynamic system in line with the United Nations’ Sendai Framework for Disaster Risk Reduction 2015-2030. It encompasses all phases such as mitigation, preparedness, response and recovery in an integrated manner by involving authorities at the national, state, district , and local levels.

Data to the rescue

Climate science has grown in the last decade with improvements in observational systems and developments in NWP models. The magnitude and direction of wind at the ocean surface are used as key parameters for tracking of cyclones. Wind data across the globe are gathered by the scatterometer payload in the OCEANSAT-2 satellite. India’s Ministry of Earth Sciences has modernized its equipment. The IMD now has 1350 automatic raingauge stations, 722 automatic weather stations and 575 manned surface observatories. The IMD also has 46 upper air pilot balloons and GPS sonde observations.

Twenty buoys assist with effective cyclone prediction. With state-of-the-art infrastructure, three weather satellites provide pictures every 15 minutes, and a series of 24 Doppler radars along the coast, the IMD is capable of sharp predictions five days in advance of cyclones. The Doppler radars provide real-time information about the signature of cyclogenesis in the Arabian Sea and the Bay of Bengal. Two Doppler radars function at Paradip and Gopalpur. In total, 37 automatic weather stations, 28 manual surface observatories, 177 automatic rain gauges and 314 rain gauge stations of the Odisha government provide data to predict weather extremes.

Odisha also has four pilot balloon observatories at Bhubaneswar, Gopalpur, Balasore and Jharsuguda and three GPS sonde stations at Bhubaneswar, Gopalpur and Jharsuguda. The National Knowledge Network (NKN) facility and a separate broadband connectivity to the IMD by the National Informatics Centre (NIC), a display facility of digitized weather charts at synoptic hours, and forecast of different NWP models using three monitors simultaneously, allow real-time predictions.

Scientists are now forecasting weather events at different time scales, from short to long-term climate projections, with a number of NWP models. India uses the multi-model ensemble technique to make forecasts closer to actual observations on the ground. Impact based forecasts are also useful for planners. Speedy communication networks help in evacuation of vulnerable people on priority basis to multipurpose cyclone-flood shelters. Farmers are given timely advice from scientists at the Orissa University of Agriculture and Technology (OUAT) to keep harvested crops safe, and to refrain from applying fertilizers or pesticides.

The 1998 heat wave, which largely affected the poor people, farmers and workers, prompted formulation of a Heat Action Plan by OSDMA, in collaboration with the IMD and the Indian Institute of Public Health (IIPH).

Flood hazards

Flooding during the monsoon months, especially in the rivers Mahanadi, Brahmani, Baitarani, Budhabalanga, Subarnarekha, Vansadhara and Rusikulya, has traditionally been a cause of concern. Odisha’s entire coastal belt is also prone to storm surges. According to the National Remote Sensing Centre (NRSC), about 8.96% of land in Odisha was affected by floods during 2001-2018. Analysis of a large number of satellite images available during this period has enabled the Odisha government and NRSC to make a flood hazard atlas to guide state management.

The IMD has employed ‘now casting’ techniques for thunderstorm and lightning. OSDMA uses sophisticated technology from Earth Networks USA for lightning detection. The private Siksha ‘O’ Anusandhan (SOA) university in collaboration with Indian Institute of Tropical Meteorology, Pune, is issuing lighting warnings by WhatsApp messages.

Odisha has set a great example of public involvement in effective management of disasters in two cyclone affected villages. The Intergovernmental Oceanographic Commission of the UNESCO has recently declared two sites – Noliasahi and Venkatraipur – as tsunami-ready.

With advanced understanding of extreme weather events, and availability of sophisticated instruments, it will become possible to read a 3D picture of the atmospheric situation for any given place. Artificial intelligence is increasingly being used to detect model errors. These tools are useful in delivering timely early warnings. The future looks safer with the availability of apps like Mausam, Meghdoot and Damini from IMD.

Although it is anticipated that climate change will cause more frequent extreme weather events, demanding increased alertness, scientific tools provide a silver lining. Even so, Odisha, like the rest of the world must be mindful of the need to mitigate the forces behind extreme weather by embracing green energy and development practices.

*Former Professor and Head, Centre for Atmospheric Sciences, IIT Delhi, India. ** Director, Centre for Environment & Climate, ITER, Siksha ‘O’ Anusandhan (SOA), Bhubaneswar, Odisha, India.