Coupled models and Climate Variability

Having studied the atmospheric component of the Unified, HadAM3, in some detail, we turned our attention to coupled model integrations using a variety of model runs, including ECHAM and HadCM3.

Key issues investigated over the last year:
1) Where and how is coupling important to climate variability?
2) Do teleconnections undergo low frequency (decadal and beyond) variability in coupled models?
3) How well do coupled models deal with observed teleconnections, such as the delivery of the ENSO signal to Africa?


Issue 1: The importance of coupling
This figure shows the difference between the standard deviation of rainfall in a 50 year control run of HadCM3 and HadAM3 forced with climatological SSTs for 50 years. The tropical Indian Ocean shows up as a region where the lack of coupling in HadAM3 leads to more rainfall variability. The Pacific differences are easier to understand!

H3CON-UM (Climatology SSTs)

Please click on the above thumbnail image to see a larger JPEG version.

Issue 2: low frequency variability.
Do teleconnections undergo low frequency variability in coupled models? The panels below show Nino 3 - Indian Ocean SST (temp over land) correlations for 30 year periods in HadCM3 forced with the A2 scenario. Links to the Indian Ocean weaken substantially during the early half of the 21st century.

Please click on the above thumbnail image to see a larger JPEG version.

Issue 3: Stability of teleconnections.
Running 20-year correlations between Nino 3 and southern African rainfall, for the period 1980 to 2100, in in the same experiment shown above. Note how the correlation changes sign! Much Palaeo work is premised on the idea that ENSO teleconnections remain fixed...

Please click on the above thumbnail image to see a larger JPEG version.

For more information contact: Dr Richard Washington