[MITgcm-support] query ?

[Martin Losch]

I agree with David, that the problem with the PAOBS data is not what's  
causing these strange heat transports. Trenberth uses the data from  
1979 to 1993 for his analysis, so exactly from the problematic period.
I have now downloaded 56*12=672 monthy averages from
http://iridl.ldeo.columbia.edu/SOURCES/.NOAA/.NCEP-NCAR/.CDAS-1/ 
.MONTHLY/.Diagnostic/.surface/
I averaged the monthly data into yearly averages and computed the  
meridional heat transport for each year. The result is attached: Until  
1975 the heat flux is northward, from 1975 to 1993 its southward then  
northward again. (and never really balanced). When you average over the  
entire period you get the second plot that I have attached. In the same  
plot an average over 1985 to 1989 (red curve). If you compare that to  
Trenberth etal (2001,Climate Dynamics), figure 13 you'll see that this  
curve looks very much like what Trenberth calls "NCEP model" in his  
figure 13 (middle panel). I guess I still haven't understood what the  
difference between "derived fluxes" and "model fluxes" are. But I think  
we know now that, what I have been using and what in the repository for  
global_ocean.90x40x15 and similar experiments are the "NCEP model  
fluxes" which have a dramatic bias!

Martin



-------------- next part --------------
A non-text attachment was scrubbed...
Name: oceanic_hflx.png
Type: image/png
Size: 16191 bytes
Desc: not available
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20050604/f85b4e3c/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: oceanic_hflx2.png
Type: image/png
Size: 21510 bytes
Desc: not available
URL: <http://mitgcm.org/pipermail/mitgcm-support/attachments/20050604/f85b4e3c/attachment-0001.png>
-------------- next part --------------


On Jun 3, 2005, at 4:34 PM, David Ferreira wrote:

>
>
> Dimitris Menemenlis wrote:
>
>>> PS. I am currently having a closer look at the data to find out why
>>> the  ncep heat flux is so different from what, e.g., Trenberth+Caron
>>> (2001,  J.Clim.) find.
>>
>>
>> Martin, there is an old story (rumor?) that NCEP reanalysis got some 
>> key data set reversed by 180 deg in the Southern Ocean.  Could that 
>> play a role?  Has anyone else heard or know more about this rumor?
>>
>> Dimitris
>
> That is actually true.  Trenberth, Caron and Stepaniak (Clim. Dyn. 
> 2001) gives all
> the gory details of the NCEP and ECMWF re-analysis  (Be careful it's 
> not pretty).
> But I am not sure it can explain the difference found  by Martin.
> Anyway, the OHT estimates of Trenberth and Caron use some 
> corrected/adjusted
> NCEP surface heat fluxes (for example, to correct for an implied 
> meridional heat
> transport in the continents) over some short period (a few years). 
> These fluxes are
> probably different from the raw NCEP data.
>
> david
>
> -- 
>
> Ferreira David               Tel : 617 253 7967 EAPS Room 54-1515,
> Massachusetts Institute of Technology,
> 77 Massachusetts Avenue
> Cambridge, MA, 02139
>
>
>
> _______________________________________________
> MITgcm-support mailing list
> MITgcm-support at mitgcm.org
> http://mitgcm.org/mailman/listinfo/mitgcm-support