Two SDSS data pulls around Pal5

/home/ch558/palomar5/sdss_result.csv
proper motions
ra, dec, psfmag_g, gmr, pm_ra, pm_dec, pm_ra_err, pm_dec_err
pulled with web interface


/home/ch558/palomar5/sdss_pal5_fehz.csv
Fe/H and z redshift
elodieFeH, elodieZ, elodieZErr, z, zErr, ra, dec, g, r
pulled with SQL, i.e.:

SELECT s.ra, s.dec, s.psfmag_g, (s.psfmag_g - s.psfmag_r) as gmr,

    pm.pmra, pm.pmdec, pm.pmraerr, pm.pmdecerr

FROM star s

    JOIN propermotions pm ON s.objid = pm.objid 

WHERE 

   s.ra BETWEEN 239.38 AND 243.38

   AND s.dec BETWEEN 5.2 AND 7.2

   AND CLEAN=1

Both data files cover the same areas, as suggested by Nitya: Combined separate queries +/- 2 degrees in RA, and +/- 1 degree in Dec around the following pointings:
    RA      DEC
1) 226.5, -2.5
2) 228.36, -0.66
3) 230.22, 0.75
4) 232.08, 2
5) 233.94, 3.2
6) 239.52, 5.9
7) 241.38, 6.2

SDSS: CMD and Velocity cuts (no members found)

Again, the second (Fe/H, z) SDSS dataset (/home/ch558/palomar5/sdss_pal5_fehz.csv).

CMD cut selects 1880 of the 16982 members. None of those then pass the velocity cut (-70 < * < -35 km/s). Only 3 of the 16982 stars pass the velocity cut, none of which also pass the CMD cut.

SDSS Pal5 dataset 2: radial velocities and [Fe/H]

datafile: /home/ch558/palomar5/sdss_pal5_fehz.csv
(columns: eFeH, eZ, eZerr, z, z_err, ra, dec, g, r)


Spatial Coverage:
(white is this datafile, red X's are selected members of Hydra)

[no matches between this SDSS dataset and our Hydra members)

Proper Motion with error bars

Proper Motion cut

Of 123 possible Hydra members, SDSS has matches for 93. 79 of those pass this PM cut (pm le 10), reducing the number of Hydra members of 109 (those 79 plus the 30 members SDSS does not match).

SDSS proper motions for Palomar 5

First, spatial plot and proper motion (pm_ra and pm_dec added in quadrature) histogram for all stars from SDSS:

Next, using spherematch to find overlap with the hydra members:

 

Palomar 5 Project Summary

Files for this project can be found in /home/ch558/palomar5/

Membership Selection:

=========================

KECK

cmd: dist le 0.1 to isochrone

(isochrone dist modulus = 16.808, see selection.pro for extinction, etc)

velocity cuts: -70 < * < -40 km/s

HYDRA

cmd: dist le 0.1 to isochrone

(isochone dist modulus = 16.808, see hydraselection.pro for extinction, etc)

velocity cuts: -70 < * < -35 km/s

magnesium: on a plot of (mg1 + mg2) vs (mgb), members are < 2 from (0.15,1.0)

ODEN

velo: -80 < * < -30

(done in allvelo.pro and allcenter.pro) 

Overlapping data:

=========================

overlap.pro gives overlaps between keck and hydra data

doubles.pro finds doubles in the hydra data

oden.pro looks for overlap between odenchirken's data and keck+hydra

Relevant code:

=========================

allvelo.pro

Reads in keck, hydra, and oden data and combines it (requires that the alldata_Pal5.fits.gz and allhydra_Pal5.fits.gz files have usable x.member columns (that = 1 for members). Plots l*cos(b) versus velocity, creates bins. Plots the bins and the dispersion bins, and then does a linear fit to the velocity bins. Prints slope of fit line (the velocity gradient per degree) and the average dispersion.

allcenter.pro

Same as allvelo.pro, but only looks at the center of pal5 (limits made and can be changed in the binning section of the code... change max_covered's first definition to control start point of binning and then change n_bins and the binning increments to adjust end point)

overlap.pro

Finds overlap between hydra and keck

doubles.pro

Finds doubles in the hydra data

oden.pro

Looks for overlap between odenkirchen's data and keck+hydra

selection.pro

Does selection for keck data, including lots of plots showing the cuts. Requires the besancon model (mw-model.resu) and m92.iso isochrone)

hydraselection.pro

Same as selection.pro, but for hydra data. 


Note: selection.pro is for keck data, hydraselection.pro for hydra data. Both have commented-out sections at the end which write the "member" column back to the fits file

Data files:
=========================

alldata_Pal5.fits.gz

allhydra_Pal5.fits.gz

(both files have member=1 for members selected by [hydra]selection.pro)

oden_center.dat

oden_tail.dat

(membership cuts only done in allvelo.pro, allcenter.pro, "just in time")


hydradoubles.dat
(two-column file listing indicies for pairs in hydra data after a quality cut)

Misc Notes

=========================

Hydra systemic error calculated to be +/- 9.8, this has already been 

added into the fits file errors

What's Next

=========================

• The doubled stars (doubles.pro, overlap.pro, oden.pro) should be folded

into each other, rather than being counted multiple times.

• In allvelo.pro, try fitting the velocity gradient line to all of the

member stars, instead of binning them (likely will just give far too much 

weight to the center of Pal5, but easily tried)

• Membership probabilites instead of hard binary cuts. Right now the members are likely, but not secure.

Palomar 5 center velocity gradient

Velocity gradient: 0.314982 km/sec per degree
This value is very sensitive to changes in bin size, bin number, and what is chosen as the edge of the center.

Average dispersion: 1.506 km/sec

Palomar 5 Velocity Gradient

All members, velocity and dispersion. Linear fit done with robust_linefit.










Velocity gradient: 0.4236 km/sec per degree
Odenkirchen's velocity gradient: 1.0 ± 0.4 km/sec per degree


Our average dispersion: 2.8 km/sec
Odenkirchen's dispersion: 4.7 km/sec OR 2.2 km/sec (when removing two outliers)