Bringing science and development together through news and analysis

  • Computer simulation models can predict tropical cyclone formation

[NEW DELHI] Advanced computer simulation models can predict cyclone formation in the north Indian Ocean a week in advance, an international conference heard.

Peter Webster, professor at the Georgia Institute of Technology's school of earth and atmospheric sciences, US, told a meet on Indian Ocean tropical cyclones and climate change, which began this week (14 February) in New Delhi, that such models used multiple computer simulations.

Webster’s team tested a European Centre for Medium Range Weather Forecasting (ECMRF) model, using an ensemble of computer simulations to predict weather, on data from 2007—2010.

The team found the ECMRF model showing ‘moderate to high’ probabilities of predicting correctly that a cyclone would form in the north Indian ocean, with lead-time of one to seven days.

For a five-day advance forecast, the model could detect 65 per cent of cyclone formations, Webster’s colleague James Belanger reported. Retrospective analysis of data showed the model picking up Cyclone Nargis — that devastated Myanmar in 2008 — seven days before its formation, he said.  

Cyclones devastate countries along the Bay of Bengal and Arabian Sea with densely populated coastal plains. The region experiences only five per cent of global storms, but a massive 95—98 per cent of damage in terms of lives and property, speakers said. 
 

A report by Webster and colleague James Belanger submitted to Weather and Forecasting that "eight of the ten deadliest TCs (tropical cyclones) of all time have occurred in the Bay of Bengal and the Arabian Sea, with five impacting Bangladesh and three making landfall in India."

"These challenges highlight the need to provide governments and populace in the region as much advance warning as possible," it added.

Dev Raj Sikka, former director of the Indian Institute of Tropical Meteorology, cautioned that there were differences between ‘potential prediction ability’ of models and ‘realisable prediction’.

"Experience on the ground shows us that the performance of most prediction models is far less than their theoretical potential," Sikka told SciDev.Net.

A World Meteorological Organisation (WMO) team had warned in the March 2010 issue of Nature Geoscience of increased intense TC activity in some regions due to global warming.

Thomas Knutson, one of the authors, told the conference that TCs could impact weather dynamics in countries in the mid-latitudes and even polar regions, and this needs to inform global weather prediction models.

Inputs from the conference, organised by the WMO and the Indian Meteorological Department, will feed into the fifth assessment report of the United Nations Inter-Governmental Panel on Climate Change due for release in 2013.