Longrange forecasts

Three areas of advance deserve attention. Predictions of the number of Atlantic tropical storms and hurricanes and of the number of days on which these occur have been developed from statistical relations with the El Niño state, mean April to May sea-level pressure over the Caribbean and the easterly or westerly phase of the stratospheric tropical winds at 30 mb (see pp. 27 and 273). Cyclones in the following summer season are more numerous when during the spring season 30- and 50-mb zonal winds are westerly and increasing, ENSO is in the La Niña (cold) mode and there is below-normal pressure in the Caribbean. Wet conditions in the Sahel appear to favour the development of disturbances in the eastern and central Atlantic. An initial forecast is made in November for the following season (based on stratospheric wind phase and August to November rainfall in the western Sahel) and a second forecast using information on nine predictors through July of the current year.

At least five forecast models have been developed to predict ENSO fluctuations with a lead time of up to twelve months; three involve coupled atmosphere-ocean GCMs, one is statistical and one uses analogue matching. Each of the methods shows a comparable level of moderate skill over three seasons ahead, with a noticeable decrease in skill in the northern spring. The ENSO phase strongly affects seasonal rainfall in northeast Brazil, for example, and other tropical continental areas, as well as modifying the winter climate of parts of North America through the interaction of tropical sea-surface temperature anomalies and convection on mid-latitude planetary waves.

Summer monsoon rainfall in India is related to the ENSO, but the linkages are mostly simultaneous, or the monsoon events even lead the ENSO changes. El Niño (La Niña) years are associated with droughts (floods) over India. Numerous predictors of monsoon rainfall over all India have been proposed, including spring temperatures and pressure indicative of the heat low, cross-equatorial airflow in the Indian Ocean, 500

and 200 mb circulation features, ENSO phase, and Eurasian winter snow cover. A key predictor of Indian rainfall is the latitude of the 500-mb ridge along 75°E in April, but the most useful operational approach seems to be a statistical combination of such parameters, with a forecast issued in May for the June to September period. The important question of the spatial pattern of monsoon onset, duration and retreat and this variability has not yet been addressed.

Rainfall over sub-Saharan West Africa is predicted by the UK Meteorological Office using statistical methods. For the Sahel, drier conditions are associated with a decreased inter-hemispheric gradient of sea-surface temperatures in the tropical Atlantic and with an anomalously warm equatorial Pacific. Rainfall over the Guinea coast is increased when the South Atlantic is warmer than normal.

Continue reading here: SUMMARY

Was this article helpful?

0 0