These are my comments for the PDS
review of the Student Dust Counter (SDC) on the NASA mission
NEW HORIZON, products version 3.

This is the review for the data sets: 

nh-j-sdc-3-jupiter-v3.0 and nh-x-sdc-3-launch-v3.0


- Download went fine.  
- My data set review was to check whether I could be able to: 
 
extract events, with utc time and impact charge
extract quality data for each event
sort out dust impacts and get geometry

The analysis was done with IDL. 


Data analysis
--------------
1- Comparison of two archived version of the same event:
nh-j-sdc-3-jupiter-v1.0/data/003939/sdc_0039393723_0x700_sci_1.fit
and
nh-j-sdc-3-jupiter-v3.0/data/20070420_003939/sdc_0039393723_0x700_sci_1.fit

I used the same IDL routine as during my first review: MRDFITS

Each .fit is now 9 columns by 30 rows, was 6 columns by 30 rows in the previous version.


The new charge table contains:
   UTC_TIME        STRING    '12-APR-2007 16:45:32'
   MET             DOUBLE           38702250.
   CHANNEL         INT              4
   CHARGE          DOUBLE           766346.38
   MASS            DOUBLE       1.5352672e-12
   MASS_THRSH      DOUBLE       1.2303292e-12
   M_SIGPLUS       DOUBLE       1.3900990e-11
   M_SIGMINUS      DOUBLE       1.3900560e-11
   QUALITY_FLAG    STRING    'OK'
The past version was:
   TIME            DOUBLE           38702250.
   CHANNEL         DOUBLE           4.0000000
   THRESHOLD       DOUBLE           594042.45
   CHARGE          DOUBLE           755059.81
   SIGPLUS         DOUBLE           89144.439
   SIGMINUS        DOUBLE           82340.201
 

At first glance, the new version contains values that can be more easily understood by the user.
Calibrated values (mass and error factor) are directly provided, which is a great improvement compared
to the previous version. 
I note that the CHARGE value has changed by a few percents compared to the previous version. 
This is probably the result of the new calibration ?

I focused the review on understanding how the mass is obtained.
Using the sdc_cal_matrix.fit content, one gets from DN numbers to the number of electrons. 
The formula for calibration is the one given in Horanyi et al. sdc_ssr.pdf
Ne = 3.8e17m[g]^1.3v[km/s]^3
I am assuming here there is nothing more recent or updated, and this formula is presented in the
DOC section, so I guess it applies.

To obtain the mass, one needs the impact velocity of the particle (as SDC is sensitive to the
momentum of the impactor only). 

Maybe I overlooked it, but it would be useful to have the theoretical impact velocity
of particles on heliocentric circular orbits onto the detector, so the user could see directly
where the mass value comes from.

I tried to find the spacecraft (heliocentric) velocity: the table document/nh_trajectory.tab in principle
contains it [XYZ component in J2000] but note that the lbl does not say with respect to what body the
velocity is computed. The label says, for example, 
'Cartesian X component of the velocity vector of the New Horizons spacecraft in the J2000 ecliptic frame.'
J2000 is a reference frame, not attached to any specific body.
This sentence is incomplete as it does not indicate the body with respect to which the velocity is computed. 

(I think that this should be corrected.)

I therefore gave up trying to get the velocity from the data set at this point 
and computed the impact velocity of particle on circular heliocentric orbits myself by:
- downloading the spk kernels for new horizon [went to naif.jpl.nasa.gov]
- using the impact time to compute the spacecraft velocity AND the theoretical impact velocity that a particle
on a heliocentric circular orbit would have on the spacecraft at the event time.

Then: I compare the number of electron (charge) obtained assuming the indicated mass value with the impact velocity
I computed (using the formula as in sdc_ssr.pdf) with the number of electron indicated in the data.

Result: I tested randomly in different .fit data files, and I get a systematically higher number of electrons 
(factor 1.2 to 1.3). 
The same test done assuming the number of electrons from the data set yield obviously similar differences on the mass
value.


UPDATE: I JUST FOUND SOME NOTES FROM MY PREVIOUS REVIEW, D. JAMES TOLD ME THAT THE FORMULA NEEDS
TO BE UPDATED: THE COEFF IS NOT 3.8E17 BUT 3E17. USING THIS COEFF, I GET SIMILAR RESULTS. 
THE CORRECT FORMULA THAT GIVES THE N_E FROM MASS AND VELOCITY SHOULD BE GIVEN IN THE .LBL. 


Minor points: 
The quality flag is  a big improvement. 
The corresponding description in the .LBL is good.