摘要 : On average, about two tropical storms or cyclones enter or form in the Mozambique Channel between the African mainland and Madagascar each year. Their impact can be devastating. The tropical cycloneIdai, which hit land in Mozambiq... 展开 On average, about two tropical storms or cyclones enter or form in the Mozambique Channel between the African mainland and Madagascar each year. Their impact can be devastating. The tropical cycloneIdai, which hit land in Mozambique in 2019, was one of the deadliest storms on record in the Southern Hemisphere. Previous studies have found that the tracks and strengths of tropical storms and cyclones are difficult to predict more than a few days ahead. An extension of this forecast horizon would be crucial for enabling authorities to take precautionary actions. Here, the ability of the European Centre for Medium-range Weather Forecasts (ECMWF) state-of-the-art ensemble prediction model to predictIdaiand 38 other tropical systems is assessed. Local wind speed maxima and sea-level pressure (SLP) minima were predicted by about 25% of the ensemble members at lead times of 1-3 days but only about 5% of the members at 7-9 days' lead times. However, several variables showed potential as precursors of tropical storms and cyclones, and some of these could be predicted at long lead times. In particular, SLP anomalies, which were significantly lower than normal multiple days before the peaks of the storms, were skilfully predicted by the ensemble mean at lead times of at least 3 weeks. Specific humidity at 500 hPa and divergence at 200 hPa were higher than normal several days before the peaks, and these potential precursors were also forecasted with skill up to about 2 weeks in advance. The inclusion of forecasts of these variables as predictors in hybrid statistical-dynamical forecasting systems could potentially extend the time horizon for prediction and early detection of possible tropical storms and cyclones in the region. 收起