Premium
The impact of mean state errors on equatorial A tlantic interannual variability in a climate model
Author(s) -
Ding Hui,
Keenlyside Noel,
Latif Mojib,
Park Wonsun,
Wahl Sebastian
Publication year - 2015
Publication title -
journal of geophysical research: oceans
Language(s) - English
Resource type - Journals
eISSN - 2169-9291
pISSN - 2169-9275
DOI - 10.1002/2014jc010384
Subject(s) - thermocline , equator , climatology , momentum (technical analysis) , mode (computer interface) , forcing (mathematics) , environmental science , coupled model intercomparison project , flux (metallurgy) , climate model , geology , climate change , oceanography , latitude , geodesy , computer science , economics , metallurgy , operating system , materials science , finance
Observations show that the Equatorial Atlantic Zonal Mode (ZM) obeys similar physics to the El Niño Southern Oscillation (ENSO): positive Bjerknes and delayed negative feedbacks. This implies the ZM may be predictable on seasonal timescales, but models demonstrate little prediction skill in this region. In this study using different configurations of the Kiel Climate Model (KCM) exhibiting different levels of systematic error, we show that a reasonable simulation of the ZM depends on realistic representation of the mean state, i.e., surface easterlies along the equator, upward sloping thermocline to the east, with an equatorial SST cold tongue in the east. We further attribute the differences in interannual variability among the simulations to the individual components of the positive Bjerknes and delayed negative feedbacks. Differences in the seasonality of the variability are similarly related to the impact of seasonal biases on the Bjerknes feedback. Our results suggest that model physics must be enhanced to enable skillful seasonal predictions in the Tropical Atlantic Sector, although some improvement with regard to the simulation of Equatorial Atlantic interannual variability may be achieved by momentum flux correction. This pertains especially to the seasonal phase locking of interannual SST variability.