Dear Joan,

Thanks for your reply. However, you evaded the crucial point about
what background you subtracted. If you subtracted the background
very close to the loop edges then you should be fine, but if you took the
background away from the loop then your background-subtracted fluxes 
are heavily contaminated by the background corona. I wish your paper 
would quantify what you did. All the paper says is "... we choose 10 pixels 
from a CLEAN background area...". Since the loop fluxes typically make up
only about 10% of the total flux, the background level is at 90%.
If you choose a background in a dark part of the image away from the loop
(perhaps this is what you mean with a CLEAN background area ?),
say at a 50% level, your background-subracted loop flux is 50% instead of
10%, and then the DEM is dominated by the background
corona rather than by the loop-associated flux. This is why I want to know
what background you subtracted, so that I can judge whether the resulting
DEM is pertinent to the loop or to the corona. It makes all the difference !

Looking forward for a clarification. Thanks !

Regards,
Markus



On Nov 23, 2010, at 11:10 AM, Joan T Schmelz (jschmelz) wrote:

Dear Markus,
 
(2) What are the total fluxes F_tot and background fluxes F_bg that you
used in your DEM modeling of the loop fluxes F_loop=F_tot - F_bg 
in the 6 coronal AIA filters? If the reader knows these 12 numbers, 
one could verify or reproduce your DEM modeling.
 
>>We’ve already used two quite different DEM methods on these data and gotten very similar results, and personally, I do not want a repeat of what happened after your re-analysis of our CDS loop data. (If anyone is reading this and does not know what I’m talking about, that’s probably a good thing.) I think what might be of greater interest is an entirely independent analysis of your own.
 
Also, you don’t just need 12 numbers; you need the uncertainties as well. These are extremely important, especially for loops that are not visible in all the filters. We always want high- and low-temperature constraints, but we also need to know how strong those constraints are.
 
(3) What are the cross-sectional flux profiles F_tot(x), F_bg(x), F_loop(x)
you inferred at the analyzed loop segments. If you show these profiles to the
reader, one could inquire or judge whether the same loop or multiple loops
are seen at the analyzed location of the loop, and what the background
contamination in the loop flux could be.
 
>>Same as (2) above.
 
Regards,
Joan

____________________________________________
Dr. Markus J. Aschwanden
Solar & Astrophysics Laboratory
Lockheed Martin Advanced Techology Center
Org. ADBS, Bldg. 252
3251 Hanover St., Palo Alto, CA 94304, USA
Phone: 650-424-4001, FAX: 650-424-3994
_______________________________________
____________________________________