Indian predictions of the likely peak wind speeds of Hudhud at landfall are lower than those by other weather agencies. The why and the how of it.
BY G B S N P VARMA
On October 12 Cyclone Hudhud is expected to make landfall before noon near Visakhapatnam in Andhra Pradesh. While destruction is guaranteed, state and national agencies are moving in meet the zero-loss-of-life goal, but a minor storm is brewing alongside, currently lodged between international and Indian forecasters, on the eventual strength of the cyclone.
Writing in Slate’s Future Tense blog, meteorologist Eric Holthaus, says, “India’s Meteorological Department (IMD) is notorious for underestimating the strength of tropical cyclones in official advisories and was sharply criticised (by me and others) for doing so last year as Phailin was approaching shore.
“At Phailin’s peak, the IMD estimated the storm’s sustained winds at a whopping 30 mph less than most international agencies. Phailin’s central core weakened substantially just before landfall, which also reduced the danger it posed and helped IMD justify its forecasts.”
Holthaus’s contention is that IMD is at it again with Hudhud. While Hurricane Weather Research and Forecasting (HWRF) (a model developed by US agencies) predicts Hudhud will be strong Category 3 at landfall, IMD characterises it as “Very Severe Cyclonic Storm (VSCS)”. A VSCS has an associated wind speed between 119 kmph and 220 kmph, while a Category 3 cyclone has wind speed between 111 mph and 130 mph, i.e. 178 kmph and 209 kmph.
Winds, according to the IMD bulletin issued Saturday at 8.30 a.m. will be “squally, wind speed reaching 50-60 kmph gusting to 70 kmph would prevail along & off north Andhra Pradesh and south Odisha coasts during next 12 hrs. The wind speed would gradually increase to 140-150 kmph gusting to 165 kmph around the time of landfall along & off north Andhra Pradesh (East Godavari, Visakhapatnam, Vizianagaram and Srikakulam districts) and 80-90 kmph along and off adjoining districts of south Andhra Pradesh (West Godavari, Krishna districts) and south Odisha (Ganjam, Gajapati, Koraput and Malkangiri districts).”
The international agencies and forecasters, on the other hand, are estimating the winds to be in the range of 180 kmph to 220 kmph, with the higher figure coming from the HWRF. Now, a high-resolution model, specifically for cyclones, Holthaus says in a phone conversation, is predicting the wind speed to be 220 kmph. The model is still in test phase, one or two years old, and worked well during a storm prediction in July this year. The consensus is winds are going to top out around 200 kmph. “The models, including Global Forecast System (GFS) and European Centre for Medium-Range Weather Forecasts (ECMWF), work well with all other oceans. “Maybe the Indian Ocean is different. But my guess is it’s not.”
Forecasting is based on feeding variables associated with an event into computer models, and reading the possibilities. Models are one part of the forecasting process. It takes information like satellite data, prior knowledge of the region, and local geography and topography. Other factors such as rainfall, storm surge, and the shape of the coastline play an important role in arriving at a correct prediction. The international agencies don’t have the wealth of local knowledge which the IMD has.
Experts in forecasting say the IMD does have access to international models. However, it’s not clear if IMD is making use of them. It is possible that different models lead to different interpretations which is a factor of the quality and nature of information fed into it. Some people have attributed the differences in estimation, Holthaus says, to the fact that IMD has its own studies on cyclones in the Bay of Bengal and the Arabian Sea. “I have never been to IMD offices and talked with people to learn about their methods of forecasting. What is clear, though, is that the IMD does it differently from other countries,” he says. That said, he says, if there is an underestimation of the cyclone’s strength, people need to know the other possibilities. Holthaus says we cannot measure exact strength and wind speed because in this case no planes equipped with instruments are flying into the storm.
IMD’s forecasting head, Dr. M. Mohapatra, who signs off on its bulletins, concurs. “Forecasting is a subjective exercise, not objective,” he says. “Forecasting depends on the type of observations you make, the type of standard operational procedures you’re following, human expertise, tools and technology.”
Mohapatra explains that there are two types of variations: variation in current situation and variation in forecast. If there are variations in current situation, they add up in the end. If there are variations in initial estimates, those too add up in the end. Why, he poses the question himself, is there variation in initial conditions in the first place? He explains: it arises due to the variation in estimation of intensity. Intensity estimations for tropical cyclones are done all over the globe using the Dvorak technique, which is “an empirical technique and a subjective technique.”
By studying the patterns of clouds in a satellite picture, pattern distinctions in the clouds, patterns of cloud distribution, by knowing cloud features, by analysis–intensity number or code for a cyclone is fixed or arrived at. “The interpretation and conclusion depends on the person who is analysing,” Mohapatra says.
There are objective techniques too. He considers them to be erroneous, often overestimating the intensity. Another factor, he says, is the wind associated with the storm. “‘How do you measure the wind speed?’ You can measure it for one second–the instantaneous averaging wind–you can measure it for one minute averaging period, for two minutes, for three minutes and so on.”
While IMD measures wind at the averaging period of three minutes, the US does it for one minute, Japan for five minutes, and China for two minutes. Naturally, he says, the average of wind speed is higher at an averaging period of one minute, than at an averaging period of three minutes. This is another source of difference in estimating a cyclone’s strength.
Mohapatra says structural engineers and civil engineers in India know that IMD estimates wind speed at an averaging period of three minutes. “Based on this, they calculate what the force is and act accordingly.”
Holthaus says it is necessary to make sure people know the strength of the storm too. International agencies are saying there is a stable upper level air mass over the cyclone and it’s free to grow into a rapid intensification system. Also, it’s moving over the warm waters of the Bay of Bengal. Conditions seem conducive even for a category-5 cyclone.
Mohapatra contends that sea surface temperature is not as important as thermal energy in the ocean. In October and November, the sea surface temperature is between 300C and 320C, with or without cyclones. Most days, it’s more than 30 degrees. “That doesn’t mean we have rapid intensification.” What is important is thermal energy in the ocean. If you measure that energy, he says, it’s not sufficient to support rapid intensification.
An important factor in this case, he says, is wind shear (Changes in wind velocity over short distances). He says it’s moderate to high, 15 knots to 20 knots. High wind shear doesn’t lead to rapid intensification. If it does happen, he intends to update the storm’s status depending on how the system evolves.
Whatever the evolving and eventual situation, could we forecast whether the cyclone is going to weaken, as happened in the case of Phailin, and whether there will be a rapid intensification of the system? “There’s a way to forecast weakening in that situation, but it’s difficult… just like forecasting rapid intensification is difficult. As a storm nears land, it usually weakens as the circulation is disrupted by hills and the land itself,” says Holthaus.
In the last 15 years, India only has three cyclones of this magnitude. “In these kinds of things,” Eric Holthaus says, “it is better to assume that it’s going to intensify, and gear ourselves for that. You end up saving a lot of lives.”
(G B S N P Varma is a freelance journalist based in Andhra Pradesh)
The elevation of mahant Adityanath, head of the Gorakhnath temple, as Uttar Pradesh chief minister shows exactly how far we have come on the road to Hindutva since May ...
The initial announcement that rendered their savings useless came as a shock and spelled disaster to the powerless poor, with no bank accounts and debts to usurers. But ...
Number of issues 12 nos | Yearly cover price INR 600 | Yearly offer price INR 400