Hello Chris and Caroline,
I think having the albedo in the file is a good idea. I limitted initially the cost to 100 in the output because I thought that anything larger than 10 is regarded as failed anyway? Sorry if this is wrong.
Von: ORAC Developers [mailto:[log in to unmask]] Im Auftrag von Caroline Poulsen
Gesendet: Dienstag, 22. Mai 2012 16:10
An: [log in to unmask]
Betreff: Re: NetCDF output files - thoughts?
Hi Chris, yes clearly cost not being written out correctly is a bug I will look at that and I agree it is useful to add the albedo.
I will fix up the cost if you can add the albedo that would be great. Give me a call if you want advice on how to do that.
From: ORAC Developers [mailto:[log in to unmask]] On Behalf Of Chris Arnold
Sent: 22 May 2012 10:37
To: [log in to unmask]
Subject: NetCDF output files - thoughts?
I have two quick thoughts about the NetCDF output files for ORAC.
First - an issue (I think)
At present only costs in the range 0-100 are written to the output file (and everything either side of that range is assigned the fill value).
> This causes significant difficulties in phase discrimination/cloud masking (particularly over difficult scenes such as sea-ice).
> I think the cost should be written out for ALL retrievals that converge?
I guess a solution would be to output log(cost) on the basis that we are likely to be much more interested in the low costs than the high costs?
Second - and related
Would it be worth adding the surface albedo information to the secondary output files? It is a key part of the error budget and thus an important consideration when looking at the diagnostics?