Regionally flux-adjustedUnified Model |
|
Centre for Global Atmospheric Modelling
The University of Reading |
|
How to set up a regionally flux adjusted integration of HadCM3
The first step to setting up a flux-adjusted version of HadCM3 is to run a coupled version with Haney forcing, almost exactly as described in the "how to..." page for regional coupling .
The one difference between using Haney forcing for regional coupling and using it to generate flux adjustments is that the timescale of relaxation will be different. In the regional coupling framework, the relaxation is very strong indeed so that the SSTs never drift away from the climatology in the specified region. When performing flux adjustment you will generally want the relaxation to be much "softer". The flux adjustments are intended to keep the SSTs in the right sort of range, but some variability is still allowed. For instance, you may want the seasonal mean SST in the model to be about right, whilst still allowing variability within the season. So instead of using a Haney forcing parameter of 1000 W per m**2 per K as in the regional coupling case, use a value of 163.76 W per m**2 per K. This is the value traditionally used by the Hadley Centre in the HadCM2 model which required flux adjustment to remain stable. This gives a relaxation time scale of about 2 weeks.
When you perfrom the Haney forced integration, you must include the ocean STASH diagnostic for anomalous heat flux (STASH code 30203) as a monthly mean value. This is the additional heat flux that the Haney forcing had to apply to the ocean to relax the SSTs back towards the climatology you specified. You will use this value to create the annual cycle of flux adjustments.
Click on the buttons on the left for the remaining steps in the process.
|
