Assessment (as of 28 February 2017):
- Sea surface temperatures of central and eastern equatorial Pacific maintained a gradual warming trend in the past two months and were near normal in January 2017. According to forecasts by a number of climate models around the world, ENSO-neutral conditions are expected to persist in spring 2017.
- Under the influence of global warming and local urbanization, spring temperatures in Hong Kong reveal a significant long-term rising trend. Climate models around the world forecast a weaker northeast monsoon over southern China for this spring, and hence normal to above-normal temperature over the region.
- Climate models around the world do not clearly indicate any trend for spring rainfall over southern China in 2017. Taking results from statistical forecast methods into consideration, the spring rainfall in Hong Kong is expected to be normal to above-normal.
- Climate prediction centres around the world generate seasonal forecasts, employing a variety of methods including
dynamical models, statistical methods, expert judgment and combinations of them. Predictions from different centres do
not always agree and large discrepancies can occur at times. The Observatory adopts an ensemble approach to formulate
its seasonal forecast for Hong Kong, taking into consideration available products from major climate prediction centres
and the Global-Regional Climate Model (G-RCM) operated in house. Thus,
the resulting forecast issued by the Observatory may be different from the forecast given by G-RCM.
- The Observatory gratefully acknowledges the Tokyo Climate Center of the Japan Meteorological Agency and the National
Centers for Environmental Prediction of National Oceanic and Atmospheric Administration, USA for providing dynamical
prediction model forecast and hindcast data to support the formulation of seasonal forecast for Hong Kong.
- Forecast for summer 2017 (June to August 2017) will be available around 1st June 2017.
Seasonal forecast charts provided by G-RCM and major climate centres: