From Tauwiki

[edit] Stellar Data

[edit] Spica

WORK IN PROGRESS

[edit] Coordinates

RA: 201.298 (13 25 11.579) Dec: -11.161 (−11° 09′ 40.759) GL: 316.112 GB: 50.845

[edit] Properties

Spectral Type: B1III B-V: -0.24; v: 1.04

distance: 80 pc (Perryman et al. 1997).

flux at 1520 Å is 10000 ph cm-2 s-1 Å-1 at the Earth.

E(B-V) = 0.02 (Morales et al. ApJ 530, 430)

Log(NHI) = 18.83 - 19.00 (Fruscione et al.)

[edit] Observations

[edit] GALEX

[edit] Query

select mpstype as survey,
filename,tilename,tilenum,
avaspra,avaspdec /* center of field of view */
from photoextract
where nexptime > 10 /* some threshold in seconds*/
and avaspra > 190
and avaspra < 210
and avaspdec > -25
and avaspdec < 0
order by nexptime desc

[edit] Individual Observations

There are a total of 532 observations fitting the above criterion. The distribution of these observations with angle from Spica is below:

489 of these observations are AIS with an exposure time on the order of 100 seconds but more interesting are those observations within about 6 degrees of Spica. Of these 4 are GI observations with the nearest being at a distance of 5.29 degrees and an exposure time of 3216 s.

[edit] Swift

There were a total of 299 Swift observations within 5 degrees of Spica. These files have been removed from the list of observations near Spica, most of them for being observations of large and ugly comets, leaving a total of 5 or so observations. Each Swift image consists of a number of independent exposures, each in a different extension. I added them all up, found the point sources and then again added removing all the point sources. I now find that there is too much scatter in the individual bins so that the sigma is meaningless. The angular plots are also difficult to understand because the Swift fields only sample a very small area. So my final conclusion is that Swift data are just too short exposure to be useful, especially as the closest Swift observation is only 2.47 degrees away. The observations are at Swift_Spica_observations.

[edit] IUE

There are three IUE spectra:

swp49994 is the furthest from Spica and doesn't have much flux. The mean between 1400 and 1800 A is 4e4 with σ = 3.7e4.

swp50185 and swp50184 both follow the same shape. The ratio of energy to Spica is 3 - 4x10-5 for 50184 and 3 - 4x10-6 for 50185. The expected scatter from Witt et al. (ApJ 1982 261, 492) is 2x10-5 for 50184 which is close enough to the observed value that we cannot reject it.

[edit] Results

The flux falls off with angle from Spica in the GALEX data. This agrees with the all-sky data.I have plotted the average flux in a ring around the star in Fig. 1.

Fig. 1: Flux as angle from Spica. The average of the individual observations is overplotted.

The next thing to try is to see what the 2-D distribution looks like:

Fig. 2: GALEX average flux for each observation. There is one GALEX number for each 1.25 degree field with the coordinate at the center of the field. The data points are shown. Spica is in the center of the field.

Here is another view in which I've separated into 0.5 degree bins.

Fig. 2: GALEX average flux in 0.5 degree bins.

[edit] Achernar

[edit] Coordinates

RA: 01:37:42.85   (24.43)
DEC: -57:14:12.3  (-57.24)
GL:                 290.8412
GB: -58.7920
B3 (V=0.5)

[edit] Stellar Properties

d = 44 pc
N(HI) ! 19 (York and Rogerson 1976; Bruhweiler et al. 1984;
Centurion et al. 1986)

[edit] Query

select mpstype as survey,
filename,tilename,tilenum,
avaspra,avaspdec /* center of field of view */
from photoextract
where nexptime > 10 /* some threshold in seconds*/
and avaspra > 10
and avaspra < 40
and avaspdec > -70
and avaspdec < -45
order by nexptime desc

[edit] Data

All the GALEX data are AIS. The nearest Swift observation is 3.71 degrees away.

[edit] Results

Fig. 1: Flux as angle from Achernar. The average of the individual observations is overplotted.

[edit] Sirius

[edit] Coordinates

RA: 06:45:08.9173   (101.287)
DEC: -16:42:58.017  (-16.716)
GL: 227.2303
GB: -8.8903
A1 (V=-1.46)
d = 2.64 pc

[edit] Query

select mpstype as survey,
filename,tilename,tilenum,
avaspra,avaspdec /* center of field of view */
from photoextract
where nexptime > 10 /* some threshold in seconds*/
and avaspra > 90
and avaspra < 110
and avaspdec > -25
and avaspdec < -0
order by nexptime desc

[edit] Results

No sign of scattering from star.

Fig. 1: Flux as angle from Sirius. The average of the individual observations is overplotted.

The HI is 17.60 (Hebrard et al. 1999) corresonding to ebv of 0.000069. This is so low that we will not get any scattering from the star.

[edit] Mirzam (beta CMa)

[edit] Coordinates

RA: 6 22 41.99 (95.67)
Dec: -17 57 21.304 (-17.96)
GL: 226.06
GB: -14.268

[edit] Properties

distance = 150 pc
V = 1.98
B1II

[edit] Query

select mpstype as survey,
filename,tilename,tilenum,
avaspra,avaspdec /* center of field of view */
from photoextract
where nexptime > 10 /* some threshold in seconds*/
and avaspra > 80
and avaspra < 110
and avaspdec > -30
and avaspdec < -0
order by nexptime desc

[edit] Results

Fig. 1: Flux as angle from Mirzam. The average of the individual observations is overplotted.

Rather noisy data.

[edit] Eps CMa

[edit] Coordinates

RA: 6 58 37.54 (104.66)
Dec: -28 58 19.5 (-28.97)
GL: 239.8311
GB: -11.3289
B2I
V=1.5
d = 132

[edit] Query

select mpstype as survey,
filename,tilename,tilenum,
avaspra,avaspdec /* center of field of view */
from photoextract
where nexptime > 10 /* some threshold in seconds*/
and avaspra > 90
and avaspra < 120
and avaspdec > -40
and avaspdec < 0
order by nexptime desc

[edit] Data

All data are AIS.

[edit] Results

Fig. 1: Flux as angle from Eps CMa. The average of the individual observations is overplotted.

[edit] Gamma Peg

[edit] Coordinates

RA: 00 13 14.15 (3.31)
Dec: 15 11 00.94 (15.184)
GL: 109.4338
GB: -46.6845
B2 (V=2.83)

[edit] Properties

d= 103 pc

[edit] Query

select mpstype as survey,
filename,tilename,tilenum,
avaspra,avaspdec /* center of field of view */
from photoextract
where nexptime > 10 /* some threshold in seconds*/
and avaspra > 0
and avaspra < 15
and avaspdec > 0
and avaspdec < 30
order by nexptime desc

select mpstype as survey,
filename,tilename,tilenum,
avaspra,avaspdec /* center of field of view */
from photoextract
where nexptime > 10 /* some threshold in seconds*/
and avaspra > 350
and avaspra < 360
and avaspdec > 0
and avaspdec < 30
order by nexptime desc

[edit] Data

The observations nearest the star are longer observations and so have background subtraction issues. The offsets for the 9 FUV observations are 500 900 500 - 1400 500 1000 900 900. For NUV they are 800 1000 800 3000 1300 800 1100 1400 1000. There is something odd about the 4th image. It shows a bright spot which persists throughout the observation with a generally high background throughout the image. It will be best to leave it out altogether. The file which we now should leave out is MISDR2_28640_0752. This may even be scattered light from gamma peg as it appears to point to the star. For the sake of consistency, I will just leave out all the observations.

Fig. 1: Wikisky image of the GALEX fields around Algenib. The bright region is visble in only one image and doesn't extend into adjacent tiles.

[edit] Results

Fig. 1: Flux as angle from Gamma Peg. The average of the individual observations is overplotted.

[edit] Other Bright Stars

There are no GALEX fields near the other bright stars:

bet cen: nearest is 8.6 degrees
alpha cru: 8.6
beta cru: 9.86

[edit] Stellar Properties

Star: Spectral type; V; Expected Counts
Spica: B1III; 1.04: 1.6e8
Sirius: A1V; -1.46: 1.4e7
Mirzam: B1II; 1.98: 1.6e7
Gamma Peg: B2IV; 2.83: 8e6
Eps CMa: B2I; 1.5: 2.7e7
Achernar: B3V; 0.5: 7e7

[edit] Expected Scattering

The instrumentally scattered fraction will be a function of the counts in the central star. We are actually using photon units so we simply need to divide the counts of a star by the solid angle of a pixel (1.5" x 1.5" = 5.29e-11 sr).

if we use Spica at 30,000 photon cm-2 s-1 A-1, we get 5.67e14 photon units. 5000 photon units is 8.8e-12 of the star which could easily be instrumental scattering. Better calculate more carefully; this doesn't sound right.

A 10th magnitude B star gives 10,000 counts/s/pixel in FUV so a 0th magnitude star will give 10^8 counts in FUV. 5000 photon units converts to 2.44e-3 counts/s so again we get on the order of 2.4x-11 counts/s. Now I have to integrate around the ring at a distance of 2 degrees from the star = 7200/1.5 = 30,000 pixels so the scattering in a ring at distance of 2 degrees is 7.36e-7.

A rough estimate of the lunar scattering at 20 degrees is 8e-10 per pixel which is more than I see here. More investigation is called for.

So now, I've done the calculation correctly and the plot shows the scattered light divided by the stellar flux. The best limits come from Achernar while (notably) Spica and Gamma Peg are much brighter. This is not quite correct. What I have done is divided diffuse flux by the number of counts from the star. I still have to do it properly and divide the diffuse counts by the stellar photon counts.

Fig. 1: The ratio of the scattered light to the stellar flux for each of the stars.

[edit] Modeling

The IR ratio from Draine and Li (2007) shows that the 60/100 micron ratio at a distance from Spica of:

Fig. 1: The dependence on angle of the IR ratio.

The observed ratio of 0.75 puts the dust at about 3 pc from the star.

Achernar is a much cooler star (B3) and the observed ratio of 0.4 puts the dust at 2 - 4 pc away.

There is no IR enhancement around Mirzam which place the dust much further away, greater than 10 pc.

There is an IR enhancement near Adhara to a peak of near 0.3 putting the dust at about 8 pc from the star. However no UV enhancement is seen.

It's not possible to check for Gamma Peg.

The offset comes from the background far from the star and is 500 - 600 photon units. I'll assume a range of 450 - 650 units.

[edit] N(HI) to each star

[edit] Dependence on distance and phase function

The closer to the star the scattering cloud is, the faster the dropoff in the scattered light. The observed scatter suggests that we should be 60 - 70 pc from the star.

Fig. 1: The dependence on distance of the scattering

The observed scattering clearly indicates a high value for g.

Fig. 1: The dependence on the phase function of the scattering.

[edit] Model Runs

I have run many different models for different values of g and distance. Because they are single scattering, applying other albedo, stellar flux and cross-sectional values will be simply a matter of multiplication.

Parameters used:
a = 0.1
stellar flux = 10000
cross-section = 0.001378507 (the value in the FUV at 1521 A)
NUV cross-section at 2320 A is 0.00146

These models assume that the entire Schlegel column is in front of Spica. Measured values of N(H) are about 10¹⁹ cm^-2 to the star which is about half the Schlegel value.

The best fit parameters are:

These are all reduced chisq so if I assume an error of 300 units, I will get a reduced chisq of 2.07. If I use Lampton, Margon and Bowyer, I get reduced chisq of 0.839, 1.1525, 1.945 for 50%, 67% and 90% confidence levels with 4 parameters. Assuming an error of 300 units and 354 points, this converts into chisq values of 186579, 186898, 187704 for FUV and 155156, 155475, and 156281 for NUV. This is really a small value corresponding to a difference of about 0.02 in the chisq.

Fig. FUV a-g. Confidence contours for a and g.

Fig. NUV a-g. Confidence contours for a and g. There is not much difference between the different levels.

In both cases, there is good agreement between the observed and the model

Fig. The best fit plot to Spica

[edit] Chisq results

The actual chisq is quite high and implies error bars of about 300 - 400 units, which is certainly much larger than expected

[edit] Distance/Angle/g Relationships

[edit] New GALEX Observations

What will the count rate be? The worst case is from SPEAR at 3E4 photon units at 0.5° from the star. The model predicts lower but, no matter, I get that a diffuse background of 4000 photon units comes to 60,000 counts/s in GALEX - pretty near the upper limit.

[edit] Programs and Techniques

1. process_all_files.pro: write out a single text file for all of the GALEX data.
2. angle_plot.pro: Make image around star. Angular plot around star for UV and IR. Scale IR to UV image.

Spica Paper Figures