Persistent discrepancies between modeled and observed trends in upper tropical Pacific Ocean temperature and a few implications for regional climate projections

Montag 05.12.2022, 14:00 → 15:00 Europe/Berlin

Speaker: Dr. Richard Seager, Professor, Lamont-Doherty Earth Observatory, Columbia University, New York, USA

- online colloquium -

Where?    ZOOM meeting room: https://geomar-de.zoom.us/j/7013744802?pwd=M0NvZWlvZVRuQThXdWdPbjJpYUNHZz09
Meeting-ID: 701 374 4802
Kenncode: 59KaFp

Abstract:

The trends over recent decades in tropical Pacific sea surface and upper ocean temperature are examined in observations, an ocean reanalysis and the latest models from the Coupled Model Intercomparison Project Six and the multimodel Large Ensembles archive.  Comparison is made using three metrics of SST trend - the east-west and north-south sea surface temperature (SST) gradients and a pattern correlation for the equatorial region - as well as change in thermocline depth.  It is shown that the latest generation of models persist in not reproducing the observed SST trends as a response to radiative forcing and that the latter are at the far edge - or beyond -the range of modeled internal variability.  The observed combination of thermocline shoaling and lack of warming in the equatorial cold tongue upwelling region is similarly at the extreme limit of modeled behavior.  The persistence over the last century and a half of the observed trend towards an enhanced east-west SST gradient and, in four of five observed datasets, to an enhanced equatorial north-south SST gradient, is also at the limit of model behavior. It is concluded that it is extremely unlikely that the observed trends are consistent with modeled internal variability.  Instead, the results support an earlier argument, supported by simple atmosphere-ocean modeling, that the observed trends are a response to radiative forcing in which an enhanced east-west SST gradient and thermocline shoaling are key.  The argument continues that, due to chronic tropical mean state biases, the latest generation of climate models continue to be unable to simulate this aspect of climate change.  The implications of this erroneous trend in equatorial Pacific SSTs for regional climate projections, such as in western North America, will be briefly discussed.

 

Related work can be found at: https://academiccommons.columbia.edu/doi/10.7916/6wd4-f378