Joint Seminar: How predictable is the equatorial Atlantic?

The equatorial Atlantic is subject to pronounced interannual variability that is akin to ENSO in the equatorial Pacific. In difference to ENSO, however, skillful predictions of equatorial Atlantic variability remain largely elusive, with most GCMs matched or even outperformed by persistence or statistical forecasts. This poor skill is often attributed to the severe mean state biases in the tropical Atlantic that continue to plague most GCMs to the current day. Here we present several lines of evidence suggesting that GCM performance may not be the crucial factor limiting prediction skill. 1) Sensitivity tests with several GCMs indicate that internal atmospheric processes contribute a large portion of equatorial surface wind variability during the developing phase of warm and cold events. This means that a large portion of the forcing is not due to coupled air-sea interaction and therefore inherently unpredictable. 2) GCM experiments in which various unrealistic mean state SSTs are prescribed indicate that GCMs are quite capable of reproducing observed surface wind and precipitation anomalies, so long as the prescribed SST anomaly patterns are realistic. 3) Both observations and model results indicate that the seasonal migration of the Atlantic ITCZ cuts short the development of equatorial events because the Bjerknes feedback essentially ceases once the ITCZ moves away from the equator. Thus there is a limited window (typically two months) for coupled feedbacks to amplify events. Our results suggest that it is inherent predictability, rather than mean state biases, that limit prediction skill in the equatorial Atlantic.




13:30 Uhr


Bundesstr. 53, room 022/023
Seminar Room 022/023, Ground Floor, Bundesstrasse 53, 20146 Hamburg, Hamburg


Ingo Richter, Japan-Agency for Marine-Earth Science and Technology (JAMSTEC), Yokohama, Japan


Bjorn Stevens

Zur Übersicht