how can i design a appropriate SSTA over a special area

[ok8200@sina_com]

Dear CAM 3.0 users,

I want to run CAM 3.0 by increasing the SST over a specified region (Nino3 area), by 3k and then use it as the input boundary dataset.
but there some questions :
1,when the negative SSTA were put in the model, East Asian monsoon responds to the negative SSTA is not opposite to the positive SSTA,in fact ,the NCEP/NCAR results of compositing analysis are opposive.
2, whetheer the positive or negative SSTA ,when the SSTA put in the model increased,the anomaly size of the atmospheric circulation simulated did not changed linearly follow the SSTA's augmentation.Some results have large difference .
so ,how how can i design a appropriate SSTA over a special area,shoud i do many experiments ,then select an experiment which is similar as the the NCEP/NCAR results of compositing analysis .
Any suggestions are highly appreciated.
Thanks!

 

[rneale]

3K SST anomalies are very large anomalies to add to the SST forcing and one may get neither a linear response to this (the system is not linear. particularly for large perturbations) or the expected sign of the response (certain aspects of the model's system almost certainly have different sensitivites to the observed system).

Obviously, one answer to the questions you are posing is simply that the model does not have the same sensitivity to SST perturbations as in observations (all GCMs have some deficiencies in this regard). You may however gain some further insight into the problem you are addressing by performing experiments with smaller (1K or less SST perturbations), maybe in an ensemble framework to determine the linearity of the CAM3 system. One further caveat to consider is the coupled versus non-coupled modelling framework, which may have differing sensitivities in the responses to SST anomalies (the coupled system is able to damp SST anomalies, the uncoupled system is not)

 

[ok8200@sina_com]

Dear rneale:
thank you very much for you replay,but I still have a problem.I increased the positive SSTA(real SSTA of Sep,Oct,Nov of 1997) over a specified region (Nino3 area),the sst of other months is not charge.then run CAM3.0 to jun of the next year.but if I use the negative SSTA,do the same experiment ,the result almost the same with the result of use positive SSTA.
Maybe there is something wrong in my experiment,but I can't find it.please give me some questions,thanks a lot!

 

[rneale]

How exactly are you adding the SST anomalies? Are you directly changing the correct values in the netcdf SST boundary files for the appropriate months? Confirm that you are doing exactly what you expect by doing an ncdiff on the original and your new boundary file s and checking all the months that you have changed to make sure the SST anomalies exist.
If you have output SST to the history file during your runs then this should tell you as well.
Then beyond that I am not sure what is worng. When you say the results are 'almost' the same, what does 'almost' mean?

 

[ok8200@sina_com]

Dear rneale:
I'm sorry I don't know how to insert image, I send an e-mail to you.
Thank you for your replies! Best wish!