Warm winter ? Cold winter ?
In the recent public opinion survey, members of the public gave higher marks on the accuracy of the Observatory's weather forecasts relative to the past several years. Besides the day-to-day weather forecasts, we also provide longer-range forecasts which predict the average temperature and rainfall for the next season. In view of some recent news reports about the possibility of early bloom of flowers intended for the Chinese New Year celebration, with suggestions that temperatures were warmer than those forecast by the Observatory, we would like to say a few words on how good (or bad) these forecasts really were.
In early September 2011, the seasonal forecast issued by the Observatory predicted that the average temperature for autumn 2011 (September - November 2011) would likely be normal to above normal. The actual average temperature for these three months was 25.3 oC, which was 0.5 oC above the normal of the same period, validating the forecast.
In the last few months of 2011, below-normal sea surface temperatures of the central and eastern equatorial Pacific were observed. Climate models around the world predicted that the situation would likely develop into a La Niña event in early 2012. Past statistics show that the northeast monsoon over southern China is usually stronger during La Niña. Hong Kong's winter (December-February) temperature is likely to be normal or below normal when La Niña is in place [2-3]. Furthermore, the majority of climate models around the world also predicted stronger-than-normal northeast monsoon over southern China for this winter. Based on the La Niña consideration and model predictions, our seasonal forecast for winter 2011/2012 (i.e. December 2011 - February 2012) predicted normal to below-normal temperatures.
Winter (December-February) average temperature of Hong Kong
during La Niña in recent decades
Actually, December 2011 was colder than usual with a monthly mean temperature of 16.9 oC, which was 0.9 oC below the normal figure of 17.8 oC. There were six cold days (daily minimum temperature at 12.0 oC or below) in the month, about two days more than normal. The minimum temperature on Christmas Day was 10.3 oC, which was the lowest since 1984. During the first ten days of January 2012, the average temperature of 15.0 degrees was almost two degrees lower than normal. It was cold on 4-6 January 2012 with minimum temperatures below 12 oC. In other words, up to 10 January 2012, the seasonal temperature forecast for the current winter has so far been on the right track.
Instead of depicting day-to-day variation, the seasonal forecast aims at providing an overall picture of the "average weather" in a season in terms of the average temperature and total rainfall. As the forecast uncertainty and difficulty will increase with the forecasting time range, seasonal forecast is still an evolving service and a challenging task in the meteorological community worldwide. Users of the seasonal forecasts are advised to take note of the limitations, especially in handling possible short-term weather variations within a season (e.g. there may still be rather warm days in a cold winter, and vice versa). Users with special applications are welcome to discuss with us on the correct interpretation and utilization of the seasonal forecasts. We always treasure feedback and suggestions from users to help us further improve our climate prediction services to better serve all walks of life in Hong Kong.
T C Lee and S M Lee
 HKO "What's New" 16 January 2012
 Wu M. C. & W. H. Leung, 2008 : Effect of ENSO on Winter Monsoon Affecting Hong Kong,. Presented in the 4th WMO International Workshop on Monsoons, Beijing, China, 20-25 October 2008, HKO Reprint No. 789
 The Weather of December 2011
 Extracts of Climatological Data for January 2012