13.3
Are extratropical seasonal anomalies more predictable during El Niņo than La Niņa ?
Prashant D. Sardeshmukh, CIRES/Univ. of Colorado and NOAA/ERL/CDC, Boulder, CO; and G. P. Compo and C. Penland
There is a common belief in the climate research community that seasonal atmospheric anomalies are more predictable during El Niņo than during La Niņa events. The basis for this is that El Niņo is generally associated with stronger anomalous latent heat release over the central equatorial Pacific than is La Niņa, which forces a stronger global atmospheric response. This stronger signal is presumed to imply a stronger signal-to-noise ratio during El Niņo, and therefore greater predictability. This view will be questioned in this talk on the basis of very large (180-member) ensembles of AGCM seasonal integrations made with prescribed El Niņo, La Niņa, and neutral SST conditions in the tropical Pacific. The different ensemble members were generated with different initial atmospheric conditions. The seasonal El Niņo signal in 500 mb heights was defined as the difference between the ensemble means of the EL Niņo and neutral runs, and likewise for La Niņa. The seasonal noise in each of the three ensembles was defined as the standard deviation of the 180 simulated seasonal averages around their ensemble mean. The most surprising result in this GCM experiment was that although the extratropical 500 mb height signal was indeed generally stronger for El Niņo than for La Niņa SST forcing, the 500 mb height noise was also generally stronger. Extratropical seasonal anomalies may therefore not necessarily be more predictable during El Niņo than during La Niņa events. Note that accurate estimation of ENSO-induced changes of seasonal noise had not been possible previously because of the large ensemble sizes required to estimate such changes. Possible physical reasons for the greater noise during El Niņo will also be discussed in the talk.
Session 13, Seasonal Prediction (Parallel with Sessions 11 & 14)
Wednesday, 17 January 2001, 3:30 PM-4:30 PM
Previous paper Next paper