Making sense of typhoons by flying into the eye of the storm
Observatory scientists fly into the heart of tropical cyclones in order to better understand and predict their paths
Late in the afternoon of August 13, while the No 3 signal was issued in Hong Kong, the Observatory meteorologist Wong Wai-kin was in a small plane over the South China Sea, flying at 2,400 metres, just above the main clouds swirling around Severe Typhoon Utor.
The centre of the storm was nearby, and the plane began descending through dense clouds. Rain pummelled the windows. Visibility came down to almost zero, as if in thick fog. In the violently swirling air, the plane was tossed about like a sampan on two-metre waves.
After about 15 minutes, the plane emerged from the cloud, entering clear air, with a calm sea below and blue sky above: it was in the eye of Utor. For Wong, it was a precious moment: he was now the first of the Observatory staff to fly into the eye of a severe typhoon. It was also a significant milestone in research on typhoons, the west Pacific's mighty, sometimes devastating, storms.
Observatory director Shun Chi-ming cites three of the most lethal: "More than 10,000 people died during storms in 1906 and 1937, and in 1874, more than 2,000 died."
In 1874, some weather-wise locals had an inkling that a typhoon was approaching, but there was no official warning as it was still nine years before the Observatory would be established. In 1906, a warning was issued just 20 minutes before gales began blowing. By 1937, forecasts had improved, yet although more than a day's notice was given about an approaching typhoon, it arrived overnight and the onslaught took many by surprise.
From 1966, the Observatory began using satellite images to help with routine forecasts. Satellite data is now pivotal to analysing tropical cyclones like typhoons. Besides light in the visible spectrum - useful only in the day - it uses microwaves and infrared radiation to obtain estimates of upper-air cloud movements and temperatures, as well as wind speed and direction near the surface of the sea based on signals scattered by waves.
This wealth of information, along with observations from weather stations, ships, ocean buoys and radar data, is then used in computer models to predict how storms will develop and where they will head. There are several of these models, some spanning the globe, others with a regional focus.
The Observatory has a model that covers east Asia and the west Pacific. It runs on a fast computer system housed in a row of black towers the size of tall filing cabinets, with racks of processing units that can together perform 7.7 trillion operations per second.
"The computer weather model is based on an operational model system from the Japan Meteorological Agency, and every three hours, we run a three-day weather forecast," says Wong. The system can also process products for specialised users and applications, and create images for television broadcasts.
Although models undergo continual development, they still struggle to accurately predict typhoon tracks and their intensities. Problems include forecasting the existence and strength of a high-pressure system that tends to guide typhoons westwards, while if it weakens, they turn north and then northeast. "This recurvature is very difficult to predict, and storms will move either west-northwest or northeast - the forecast can't be in the middle," says Shun.
When a typhoon appears set to impact Hong Kong, at least four forecasters man the Observatory's forecasting room round the clock. In the room, monitors display radar and satellite images, wind measurements and more. The director, assisted by an assistant director, is stationed at a table with a built-in computer screen showing a range of computer forecasts. He uses these forecasts alongside real-time observations and experience to decide which warning signals to issue, and when.
With typhoons posing threats to life and property, the Observatory team is always seeking ways to improve its forecasts. One way is by using reconnaissance flights to gather data from within storms. The United States relies heavily on flights to assess hurricanes, but in this region, there have been few recent flights other than occasional research missions flown by Taiwan, which do not reach the South China Sea near Hong Kong.
Other than a trial flight in 1973 by late director Gordon Bell, the Observatory neither oversaw nor participated in such flights until two years ago.
In 2009, an Observatory team equipped a Government Flying Service plane with a weather probe, a small torpedo-like object mounted below a wingtip. This was used to gather data on flight paths around the airport, which was then used to support the development of a wind shear and turbulence alert service. "I talked to Michael Chan, the Government Flying Service controller, and asked if they could fly into typhoons," recalls Shun. "He said yes."
So in June 2011, beginning with Tropical Storm Haima, pilots flew missions into storms south of Hong Kong. Wong and his colleagues then added data obtained from the probe to the computer model, finding that it helped to improve forecasts.
In Wong's recent mission into Utor, he made three small circuits within the eye of the typhoon, searching for the lowest air pressure he could find. "I remember former director Lam Chiu-ying asking if the minimum air pressure zone of a tropical cyclone was a point or an area," he says.
Then, pilot Keith Ma headed north and into the eye wall, returning homewards at an altitude of just 910 metres - keeping as low as reasonably possible to record powerful winds.
After processing the data, Wong noted that winds were at least gale force until he reached 290 kilometres from the centre of Utor. This was consistent with satellite data, helping Shun come to his decision to issue the No 8 signal as Utor neared Hong Kong.
Successes like this mission spur the continuation and improvement of Hong Kong's typhoon reconnaissance flights. Later this year, the Government Flying Service will begin operating a new plane with better weather-monitoring equipment relaying information directly from the satellites. With Wong or a colleague on board, this will also release dropsondes - cylinders packed with sensors that descend on parachutes - transmitting information on wind speed and direction, air pressure, humidity and temperature until they hit the sea and are lost.
Martin Williams is a Hong Kong-based writer who specialises in conservation and the environment. He has a PhD in physical chemistry from Cambridge University