Are Hong Kong’s typhoons becoming stronger? Expect longer, more unpredictable storms, Observatory chief says
- Hong Kong Observatory director Chan Pak-wai says proportion of strong typhoons will increase, while they will move slower and affect areas for longer
- Effects of global warming may be partly to blame, according to Chan, who highlights ‘frequent emergence of extreme weather’
Observatory director Chan Pak-wai on Saturday explained the pattern of typhoons had changed and extreme weather would become more frequent, but maintained an earlier forecast that the city could expect four to seven tropical storms this year.
“The proportion of strong typhoons will increase, and they will have slower movement so they will affect an area for a longer period of time, and also move more northerly,” Chan told a radio programme.
Typhoon Talim kicked off Hong Kong’s storm season last month, bringing most of the city to a standstill under a No 8 warning signal. Tropical Storm Doksuri then missed the city but battered mainland China’s southeast coast and brought rare floods to Beijing.
The typhoon has already skirted past the western shores of Okinawa’s islands but is expected to move east and strike the prefecture again over the weekend, before turning north early next week and potentially reaching Kyushu island.
Chan said both Doksuri and Khanun had “considerable variables” in their paths, making forecasts more challenging.
He also revealed the Observatory was seeking government funding to incorporate artificial intelligence (AI) software into its forecasting, following in the footsteps of its mainland counterpart.
The Observatory’s director argued that such tools could outperform traditional methods when predicting the paths of tropical storms.
The China Meteorological Administration in late July released a seven-year road map for developing forecasting models based on machine learning and said it aimed to become a world leader in the use of such AI software.
An article from weekly British scientific journal Nature last month reported that Chinese researchers had developed an AI-powered weather prediction model and claimed it could make early stage predictions of typhoon tracks more accurately than other systems available on the European Centre for Medium-Range Weather Forecasts’ website.
Huawei-developed Pangu-Weather was trained using 40 years of global data and was recently made available to the public via the website.
The system takes less than two seconds to produce a 24-hour global weather forecast at a speed said to be 10,000 times faster than traditional numerical methods, according to the journal.
Director Chan said Hong Kong was expected to have warmer-than-normal temperatures this year, with the Observatory in March predicting the city’s annual mean temperature could be among its top 10 warmest on record.
He predicted the increase in extreme weather would lead to a rise in rainfall rates, with more heavy downpours in some parts of the region.
The Observatory was also planning to step up warnings for extreme temperatures, heavy rain and typhoons, he added.
Temperatures soared last month as the city felt the effects of a subtropical ridge, a large belt of high pressure, that dominated southern China.
July’s mean temperature reached 30.1 degrees Celsius (86.2 Fahrenheit), with an average maximum of 33 degrees and minimum of 28 degrees – all were the third highest on record for the month, according to the forecaster.
The Observatory in Tsim Sha Tsui recorded a maximum temperature of 36.1 degrees on July 27 under the effects of Doksuri, one of the highest figures for the month on record.
Last year was the sixth hottest in Hong Kong since records began in 1884. There were 11 record-breaking weather events, including the warmest autumn ever. July was the hottest month on record, notching up 10 days where temperatures exceeded 35 degrees.
There were 15 days in 2022 that had a maximum daily temperature equal to or above 35 degrees, marking a record for the city. The number beat the six days, previously the highest annual figure, which were recorded in 2016.
Additional reporting by Kahon Chan
