Where the West went wrong

Print edition : November 15, 2013

THE Dvorak technique of intensity estimation with satellite imageries is a manual technique, so there is subjectivity involved. Although automated techniques of objective estimation exist, they are not used operationally because there are limitations, according to M. Mohapatra of the IMD’s Cyclone Warning Division. The National Hurricane Centre (NHC) of the U.S. ultimately depends only on the manual technique. Some synoptic data, which are available to all the meteorological agencies of the world in real-time through the global dissemination system called GTS, are also used in this estimation. So why was there a difference?

“We have found from our research that, in general, intensity estimation based on the Dvorak technique by the U.S. agencies such as the NOAA and the NHC is generally higher than the IMD’s estimates,” said Mohapatra. Apparently, it is not just true of the North Indian Ocean (NIO) only but also of the Pacific. According to Mohapatra, for Pacific cyclones, the Japan Meteorological Agency (JMA) has shown that U.S. estimation is always higher. But this is only a general observation and estimations may not differ in specific cases, such as cyclone Thane of December 2011.

Said Mohapatra: “One of the reasons for the difference is that even though U.S. agencies monitor events globally they have no expertise with regard to local features of the NIO basin. The second reason is the convection during the day is not the same. There is diurnal variation in convection. Convection becomes maximum around 12 hours in the day time and at 21-24 hours at night, and this diurnal variation may not have been taken into account in their initial estimates based on imageries alone. So there is bias in their estimation. This variation has to be factored in to estimate the correct T number. The T number cannot increase now and decrease three hours later.”

One could, of course, argue that this diurnal variation must be known to the U.S. agencies as well because they too have hurricanes; in fact, more than the NIO region. “Local conditions are different for different basins and also the local times are different,” pointed out Mohapatra. “So, they may not have complete information on the magnitude of the diurnal variation over Bay of Bengal,” he added.

As the website of the Bureau of Meteorology of Australia says, “Tropical convective systems, especially those at sea, undergo large diurnal variations because of the differences in the shortwave and longwave radiation budgets of clear and cloudy regions. Tropical cyclone structure changes, on the other hand, show little substantiated diurnal preference. Because many forecasting methods use trends in satellite imagery, it is critical that forecasters be aware that cloudiness, and therefore the satellite signatures of tropical systems, also show a marked diurnal variation unrelated to structural changes [emphasis added]. Short-term trends in satellite imagery can be very misleading and cause major forecast errors.”

So how does one determine which estimate is correct? If you have ground truth observations then you can tell. However, for systems over sea only in situ aircraft measurements can provide the ground truth and Indian meteorological service agencies do not have any. But when the system comes within the Doppler radar range, the intensity estimate can be verified. And when it comes closer to the coast, coastal observations, which provide pressure and wind data, can further authenticate that. Based on pressure and wind, the intensity can be determined.

“We have verified that,” said Mohapatra. “At the time of landfall we have verified, based on pressure at Gopalpur, outer pressure [of] 941 millibar. So at the system it would be a little bit less, say 940 mb. Then the pressure drop at the centre compared to the surroundings was 66 mb. For 66 mb, if you calculate the intensity based on the known relationship between pressure and wind speeds, it comes out to be 110 knots, which is about 205-210 km/hr. In fact, they ultimately reduced their estimates to whatever we had given earlier,” he added.