1) Identification_Area - Fix name of comet "103P/Hartley 2" in title: EPOXI MRI-VIS 103/Hartley 2 Encounter Photometry Collection

2) Modification_History (in all labels) - This data set labeled v1.0, but v1.0 was a PDS3 data set by Williams et al.  I recommend v2.

3) Does Modification_Detail "Initial conversion from PDS3 to PDS4" make sense in this dataset's context?  This is a whole new dataset/version.  It is not a conversion from the Williams et al. PDS3 dataset.

  There should be some note that explains this is a new version of the Williams et al. data set.  Is Modification_Detail the right place?

4) overview.txt

  a) character encoding is off: There are UTF-8 characters, but emacs displayed it as iso-latin-1-dos?  I got an error with "mousepad".  Please correct.

  b) Missing discussion of the prior version:

      Williams, J.L., Li, J.-Y., Bodewits, D., and McLaughlin, S.A., EPOXI 103P/HARTLEY2 ENCOUNTER - MRI PHOTOMETRY V1.0, DIF-C-MRI-5-EPOXI-HARTLEY2-PHOTOM-V1.0, NASA Planetary Data System, 2012.

  c) (Here and in the main documentation) All apertures > 20 pixel radius are invalid, and the photometry tables seem to agree.  They should just be deleted, no?  It is a strange way to save data (i.e., not space efficient), but it is fine.  Users might find it confusing, though.  I recommend the reasoning behind the data structure be added to overview.txt (and the main documentation).

    Invalid measurements (i.e.  aperture photometry with r < 3 pixels
    or r > 20 pixels, and azimuthal profile integrals with r<10 pixels
    or r larger than the size of the detector) are labeled -99.

  d) "Confidence Level Overview": This discussion is probably best in the photometry document, right?  The overview should be more brief.  Also quote absolute uncertainties, but largely point the user to the documentation for details.


5) Photometry document

  a) Section 1. Add a statement about and reference to Williams et al. data here (or Section 6?)

  b) Section 3.6. would it be more fair to replace "we" with Williams et al. in the following sentence?

    In order to minimize the contamination from background stars, we
    developed the so-called ‘azimuthally-averaged photometry’ procedure,
    which measures ...

  c) "‘azimuthally-averaged photometry’ procedure, which measures the
integrated flux through the median radial profile of the comet where
stars are filtered out using an outlier-resistant determination of the
mean."
  - median --> mean?  The description that follows does not refer to
    the median.

  d) Aside from celestial sources, the background is dominated by bias residuals and associated noise (Klaasen et al. 2013).  The bias residuals are correlated noise with levels somewhere around 1 to 3 DN.  The background discussion quotes flux densities, but if the bias is the main source, then it should be described in DN first.  The conversion from DN to spectral irradiance depends on the exposure time and filter, so it will vary from frame to frame.  I recommend that this background source be described in units of DN, and converted to W/m2 μm as an example.

     Page 7 the discussion of background should be based on DN first, then converted to physical units as an example.  Furthermore, this would allow one to estimate the background for the C2, OH, GC, and UC images.

  e) Section 3.7: Missing bias correlated noise (striping).  It is described in Klaasen et al. 2013.

  f) Section 4. Known constrains.  This text is largely repeated from 3.6.  It seems that only one is needed and the other should refer to it.  Recommend discuss in 4 and refer from 3.6.

  g) Section 5. Also explain reasoning for aperture columns here (i.e., why so many -99s?).

  h) Section 5. Recommend "see table labels for column details."

  i) Description of Flux and Surface Brightness: Recommend "A value of -99 indicates photometry [was not] computed for the given aperture."  The aperture limits appear to be a choice by the data providers, rather than a physical limitation of the data or procedure.