Four hidden dwarfs: Distance, # Stars, Radius, CMD

Dwarf 1

44 kpc

3854 stars

0.905949 circular radius

Dwarf 2

150 kpc

233 stars

0.112978 circular radius

Dwarf 3

225 kpc

157 stars

0.0687904 circular radius

Dwarf 4

300 kpc

120 stars

0.060705 circular radius

spatially, these are:

3, 4
1, 2

LSST spatial, r=22.0 limit

This plot includes only those stars where(r le 22.0), simulating SDSS depth. Here two of the four dwarfs are still clearly visible.

LSST Dwarf Distance Moduli

objType=18 (dwarfs only) CMD

Isochrone shifted +18.217 (44 kpc):

Isochrone shifted +20.880 (150 kpc):

Isochrone shifted +21.761 (225 kpc):

Isochrone shifted +22.386 (300 kpc):

LSST Galaxy CMD

Color-magnitude diagrams for galaxies in LSST data (tisStar=0)

All galaxies:

Galaxies around dwarf1 (top left of spatial plot, 0.4x0.4 square centered around it) 

Galaxies around dwarf2 (top right of spatial plot, 0.4x0.4 square centered around it)

Galaxies around dwarf3 (bottom right of spatial plot, 0.4x0.4 square centered around it)

Galaxies around dwarf4 [largest] (bottom left of spatial plot, 0.4x0.4 square centered around it)

Getting plot_smooth running

Spatial plots using plot_smooth:

ComaB:

Draco:

Leo IV:

LSST "Hidden" Dwarfs: Luminosity Functions

 These are the luminosity functions of the four dwarf galaxies hidden in the sample LSST data. On each graph, the white plot is the same – the sample dwarf galaxy used to generate the data. The yellow plot is the luminosity function of one of the hidden dwarfs, starting from the top-left and continuing clockwise as seen in this spatial plot:

Comparing Sample Dwarf Galaxies

Left plots are "mario", right plots are "dwarfGalaxy"


g-r vs g CMD:

r vs # of stars:

r vs # of stars, log10 y-axis (note different x-axis scales):

LSST sample data - Accuracy Plots

r_mag vs (r_mag – t_r_mag) [i.e., accuracy of LSST]

All data:

Stars only:

Galaxies only:

Automated searching with varied distances, part 2

Still working with Leo IV. 

First, a familiar pair of graphs. Left graph is distance modulus being tested versus size of stellar population it selects. Right graph is distance modulus versus the calculated "metric" for that distance. The "metric" sums 3*(value of highest density histogram bin) + 2*(value of 2nd highest) + 1*(value of 3rd highest).

Using this data, I then try to calculated an estimated distance modulus for Leo IV.

The "adjusted metric" at the bottom is the estimated distance I come up with as a final answer: 21.00 (actual value = 20.98). This is very accurate for Leo IV, but quite possibly will not work well for other dwarfs.

How this is calculated:

First the distance which gives the highest weighted metric (i.e, greatest y-value on the right graph) is the first part (= 20.50 for Leo IV).

Second, the distance which has a local maxima of stellar population size (local max y-value on left graph) is found (= 21.50 for Leo IV).

These two are then averaged to find the final estimated distance (=21.00 for Leo IV). 

More Automated Distance-Finding for Leo IV

Left plot:

Distance modulus versus # stars selected. The yellow line at 750 stars represents the cutoff point; distances which select fewer stars are not included on the right plot (i.e, y-axis values set to zero). This minimum value is chosen to reduce the impact of very small populations on the final distance estimated. It is not a great solution, but in this case it does improve the results.

Right plot: 

Distance modulus versus # max-density (color=255) bins. The distance with the maximum y-value on this plot is returned as the estimated distance.

In this case, this algorithm returns 21.5 as the estimated distance modulus. Actual distance modulus: 20.98 (according to NED). 

Ideally the algorithm should return 21.0 as the estimated distance. This requires a more effective way of quantitatively valuing the different distances. 

(Early) Automation of finding distance - step through Leo IV

The left plot shows the relationship between distance modulus and number of stars selected by the isochrone for Leo IV. The distance moduli are stepped through one at a time, from 16.5 to 24.0 by increments of 0.5. The right plot shows distance modulus (same steps) versus the value of the densest bin in the histogram.

Leo IV: Recreating discovery paper plots

I have been working on recreating plots from the Leo IV (and other dwarf) discovery paper (BELOKUROV ET AL. 2007). Here are plots from that paper: 

[spatial of all, smoothed spatial, CMD g-i vs i of inner circle selection]

Pulling data from SDSS covering these same ra and dec ranges, I have made the following plots. First, the simple spatial plotting of all the included stars:

Spatial plot of density: 

Note that Leo IV is at the center, where the bin is white. 

Now for the third graph, the color-magnitude diagram. I have done two separate plots for this, and then a third one which overlays them. First, the g-i,i of all the stars in the dataset:

And then a similar graph but only plotting points within a radius 0.1 from the center of the spatial plot (i.e,  hopefully the Leo IV stars):

Finally, the two overlaid (with the selected stars plotted now as triangles):

Draco population selection and density plots

Moving through LG dwarfs, returning to Draco – I had already tried to do stellar population selection with Draco data without much luck.

New SDSS pull with no stars fainter than r=22.0:

Spatial density plot of all stars in the SDSS pull:

...and of only the stellar population selected (note some shifting of isochrone was done by hand even after reddening and distance corrections):

Leo II density plots

Continuing with Leo II, with a cleaner SDSS data pull for the same square degree of sky (notably, nothing fainter than r=22.0). This changes the color-magnitude diagram for Leo II:

This greatly reduces the number of stars included in the selected stellar population. However, as seen below, it still gives a useful population.

Main progress today code-wise is spatially plotting the population as 2d histogram based on density, rather than simply plotting ra/dec points. Plotting the newly pulled SDSS data around Leo II using this method, prior to doing population selection, gives the following graph:

Leo II is already clearly visible in this graph, seen as a ring like in the previous non-density spatial plots. Reducing the data included to just the selected stellar population around the isochrone cleans up the plot, leaving Leo II distinctly visible. Notably, the center of the ring is left empty, indicating a good selection.