- when  you list the sensitivities for the maps, I was wondering if those are the predicted values or measured values?  If they are the version predicted from HIPE, it might be worthwhile to verify that is the sensitivity we achieve.  I think you must have already done this in measuring the background level and sky noise in each map, so perhaps these are the measured values?

*** These are the measured values.  I note in the text that these are "the achieved" sensitivities (i.e., measured).

- in Section 3.1 paragraph 3 you say the error is "the unbiased statistical estimate of the error on the mean"  and I wasn't sure what that meant.  Is it the error on the mean or did you do something fancier to make it unbiased or maybe the error in the mean is already unbiased and I'm just showing how little I know about statistics?  

*** Hmmm ... took that directly from Helene's blurb.  I'll check with Helene, but I would think it's the usual definition of the error of the mean.

- in Section 3.2 paragraph 1 the 70/160 ratio is probably a typo since it isn't a "SPIRE quantity"

*** Good catch.

- on the topic of aperture corrections, I was wondering if the 8 micron might be a better starting point than the 3.6, since the 8 micron should mainly be tracing dust at least and has only a slightly larger starting psf than the 3.6.  Just a thought.  Also, wouldn't the need for aperture corrections go away if you were to convolve all of the bands to SPIRE 500 or something after the foreground/background & point source subtraction but before doing the aperture photometry?  I guess there would be an aperture correction for the 500 micron still, but at least the correction would presumably be the same at all of the wavelengths.

*** I looked into using the 8um image instead of the 3.6um, for the same reasons you are thinking.  But I saw too many instances of weak/patchy 8um maps, and so I did the trade-off of good S/N vs not-so-sure-it-actually-traces-cold dust.

I believe Gonzalo is carrying out global photometry after smoothing to the 500um PSF.  I prefer to not smooth and incorporate (admittedly small) aperture corrections, especially when targets are faint and may have only a few patches of observed emission.  I worry that smoothing in those cases will make it too tough to distinguish between galaxy emission and foreground cirrus.

- in the last paragraph of section 4.5, I think the last couple of sentences could be reworded. I would say something like: "Figure 8 shows a primary reason for the discrepancy: even when limited to lambda > 100 um photometry, single-temperature blackbody fits overestimate the dust temperature, thus underestimating the dust mass.  The single-temperature model does not account for the contribution of warm dust emitting at shorter wavelengths and the temperatures are driven towards higher values in the attempt to fit both the short and long wavelength far-IR emission." Or something along those lines.

*** I like it -- thanks!

- I was wondering on some of the figures how much of the scatter is real.  Would it be possible to get error bars on Figure 2, 4, 5 and 6?

*** I put representative errors bars in Figure 2