LMC

From Tauwiki

[edit] Notes

The old stuff is in Lmc Details of wavelengths, resolutions of various science missions. The paper is at LMC_paper

[edit] Technical Details

[edit] Bands

Wavelengths of the bands we have defined are in FUSE_Bands

[edit] Apertures

The details are from the Observers' Guide with the slit positions from the diagram

[edit] HIRS

The aperture is 1.25 × 20 arcseconds in size with a center to center offset of 105 arcseconds from the LWRS aperture. The edge to edge offset is 80 arcseconds.

[edit] MDRS

The aperture is 4 x 20" in size with a center to center of 210 arcseconds from LWRS and edge to edge offset of 185 arcseconds.

[edit] LWRS

The aperture is 30 x 30" in size.

[edit] Data Files

A listing of the data files is in FUSE_files.

[edit] Observations

[edit] Observation Log

There are a total of 603 FUSE observations within 5° of the LMC. The first step is to reject all point source observations within the LWRS. This is done using the Object Class from the FUSE archive (class <= 72; 77 - 88). This leaves a total of 172 observations.

Fig 1 : LMC R-band image from Bothun and Thumpson (1988) showing both the UIT fields and FUSE observations. Red circles are UIT fields and plus signs are the FUSE observation. Diameter of blue and black circles proportional to the FUSE flux values for that observation and scale ratio of blue to black is 1:5.

Of these, we have found 91 diffuse observations. These observations have been selected by checking the apertures to ensure that no sources are within and, more reliably, by checking the FWHM of the spectrum on the detector. Point sources have a FWHM of less than 20 pixels while diffuse, aperture filling sources have a FWHM of about 30 pixels.

As detailed below, we leave out more observations which results in a total of 81 diffuse observations.

[edit] Funny Data

1. D9042501: Side 2 is much brighter than side 1 but is not consistent with a diffuse source. Leave this one out.
2. D9042002: Looks ok at first glance but if I leave it out, the correlation increases to 0.985 and the linear fit becomes much better. It is a very crowded area and it is possible that there are two sources in the slit. On the other hand, the image certainly looks consistent with a diffuse radiation field. There is signal in all the three apertures. This is only a problem in Side 2. It might be better to look at the spectrum directly. No, it is much higher even in the spectrum. In this case, it's better to just leave it out. Note that this is a SNR so is not useful for us anyway.
3. X0050201: 1A2 is way off which really increases the χ². The others are ok. This is clearly because of a very bright OVI feature. There is a published paper which identifies it as emission from the N49 supernova remnant (Blair et al 2000).
4. B0100601: The exposure time is very short so this can be left out.
5. S5055502: Very low signal. Can be left out.
6. D0981701: Exposure time of 56 seconds. Leave out.
7. F0740102: 1B is high for some unknown reason. This is in both the spectra and the areas. Leave this one out.
8. G0530202: This is clearly a point source. Something this bright should also be visible in the MDRS but it is not. Leave it out.

In principle, we are printing out 1 sigma values but the χ² values are clearly high by a factor of 9, implying that we should increase the error by a factor of 3 to 5. The real effect is because the χ² is dominated by a small number of points. Most are well within the errors.

If we look at some of the specific values for the χ²:

First the points for which the 2A2-1B1 are high:

1. D9042301: This has a uniformly high χ² for all bands. If I look at the spectrum, 1B is lower than 2A. There is nothing overtly wrong with the spectrum except for this mismatch.
2. P2130101: This is an interesting spectrum. The 2A/1B data look good but the 1A spectrum has a lot of structure. These spectra are probably ok.
3. D0300201000: Data looks perfectly ok.
4. S4055401000: Problem with 1B and 2A. Leave it out.
5. D0880103000: This is fine except that there was a problem in background subtraction. Possibly the MDRS source was wider than normal.
6. D9043301000: Big difference between 1B and 2A. Leave it out.

Now the points where 1A1 is much different from 1B1:
These are all SN1987A so are probably real.

1. Z0080103000: Fine.
2. Z0080102000: Fine.
3. P2130101000: Fine
4. Z0080101000: Fine.

[edit] List of Excluded Data

1. B0100601
2. D9042002
3. D9042301
4. D9042501
5. D9043301
6. D0981701
7. F0740102
8. S4055401
9. S5055502

[edit] Spectra of FUSE Observations

Here are the spectra of each of the FUSE Observations.link

Spectra of a region has been plotted altogether here.

1. spectra of N11
2. spectra of 30_Doradus & SN1987
3. spectra of Shapley constellation
4. spectra of LMC_Bar
5. spectra of LMC_West
6. spectra of Others

[edit] UIT With FUSE Observation

[edit] Correlations

[edit] FUSE - FUSE correlations

There are excellent correlations between the different bands and, in particular, those bands with the same wavelength. Although the chi square values are high, these are due to a few discrepant points, such as SN1987A. The slope and offset have been calculated using a least squares procedure.

FUSE-FUSE Correlation taking 81 points.

[edit] FUSE UIT Correlations

Of the 81 FUSE targets, 33 were observed by the UIT B1 filter (1521 A) and 29 with the B5 filter (1615 A). The UIT 1.13 " pixels from the calibrated, geometrically corrected images (conversion of UIT fluxes to photon units) were binned over a 27 x 27 pixel box to get the observed UIT flux and were compared with the FUSE 1B1 band.

[edit] Discussion of FUSE-UIT correlation plots

Fig-2:FUSE-UIT correlation for B5 filter

Fig-3:FUSE-UIT correlation for B5 filter removing the three discrepant points.

B5 filter:

Fig (2) is the correlation plot of UIT-B5 band flux with FUSE 1B1 flux. Here, we have considered all the 29 observations of FUSE that has been observed by UIT- B5 filter and found a correlation of 0.92. Using a minimum χ² method, we got a slope of 0.57 with an offset of -62028.2. The three points in SN1987a (Z0080101000, Z0080102000, Z0080103000) give a much higher ratio. If we remove them (Fig. 3), the slope is 0.32 with an offset of -18337.4. The correlation coefficient is not much changed at 0.93.

Fig-4:FUSE-UIT correlation for B1 filter

Fig-5: FUSE-UIT correlation for B1 filter removing 5 points (3 in SN1987 and 2 in N11)

B1 filter:

There is also a good correlation (Fig. 4)between the UIT-B1 band and FUSE 1B1 flux with r = 0.77 for all 33 observations. The corresponding slope is 0.47 and the offset is -35212.8 photon units. 5 of the observations affect the slope heavily (3 SN 1987A and 2 N11: B0100101000 and B0100201000). If we remove those points (Fig. 5), the correlation coefficient rises to 0.88 with a slope of 0.35 and an offset of -16625.2.

[edit] Combined Filters

Correlation plots of UIT combined filter flux with all FUSE bands.

Fig-6:FUSE-UIT correlation taking all the points together (1B1-Band)

Fig-7:FUSE-UIT correlation taking all the points together except four discrepant points (1B1-Band)

The two bands (B1 and B5) are only 80 Å apart and so can, for our purposes here, be treated as a single band. The correlation coefficient is 0.78 but rises to 0.92 after 4 points (SN1987A and one of the N11 points) are removed. If we force an offset of 0, the slope is 0.26; otherwise, with an offset of -10939.6 photon units, the slope is 0.30. The FUSE/UIT ratio for the diffuse FUSE pointings is consistent with a constant value of 0.26 to 0.30 throughout the entire LMC.

[edit] Diffuse Fraction

The ratio of the diffuse FUSE flux in each of the FUSE bandsto the diffuse UIT flux is constant over the entire data set, barring 4 points (B0100201000, Z0080101000, Z0080102000, Z0080103000). We have used the ratio at zero offset to calculate the fraction of the total light that escapes in the FUV in the LMC (Link_to_tables_containing_the_FUSE_diffuse_fraction_for_each_individual_bands) and the fraction of diffuse light is summarized below (table 2). We have merged the two UIT bands which are only about 80 Å apart. There is a clear relationship although the value is different for each.

Table 2: Calculation of fraction of FUV diffuse emissions through different bands of FUSE.

Table 3: Total integrated diffuse flux coming from different regions in FUSE bands (in units of erg/cm2/s/A).

A different picture arises if we normalize the points to the 1B1 fraction (Fig. 9). The ratio follows clear trends from FUSE to UIT with different regions following different slopes. One point in N11 is much higher in the UIT flux while the high SN1987A data have a correspondingly lower UIT flux.

Fig 8: Diffuse fractions are from the table 2.

Fig 9:Total integrated diffuse flux from table 3 (in units of erg/cm2/s/A).

Fig 9:Total stellar flux (in units of erg/cm2/s/A).

[edit] Diffuse Fraction Error

DF = Diffuse FUSE flux

DU = Diffuse UIT flux

a = slope = FUSE/UIT

SF = Stellar flux

df = FUSE diffuse fraction

DF = a * DU

er_DF = (er_a/a + er_DU/DU) * DF

where er_a is calculated from the least square fitting of FUSE and UIT, er_DU is 10% (Parker 1998).

df = DF/(DF + SF)

df_max = (DF + er_DF)/[(DF - er_DF) + (SF - er_SF)]

df_min = (DF - er_DF)/[(DF + er_DF) + (SF + er_SF)]

er_df_max = df_max - df

er_df_min = df - df_min

[edit] FUSE - IR correlations

We find good correlation between FUSE and IRAS except for those locations from SN1987A and some locations in N11 region. For these locations the IR values are almost constant but the FUSE values are varying significantly. This could be due to the difference in the proximity of these locations from the contributors.

[edit] FUSE - NHI correlation

The Galactic neutral hydrogen over the entire sky is from LABsurvey.(Hartmann and Burton 1997)

N(HI) for LMC is from Kim et al. (1998).Taken from their webapge.

Correlation between atomic hydrogen column density (NHI) and FUSE intensity. NHI is from Kim et al. (1998).r=0.18, m=7.65e-18, c=18310

Correlation between atomic hydrogen column density (NHI) removing 3 descrepant observations and FUSE intensity. NHI is from Kim et al. (1998).r=0.40, m=8.16e-18, c=6530

Correlation between Galactic atomic hydrogen column density (NHI) and FUSE intensity. NHI is from Hartmann and Burton (1997). r=0.24, m=2.37e-17, c=-16438

Correlation between Galactic atomic hydrogen column density (NHI) removing 3 descrepant observations and FUSE intensity. NHI is from Hartmann and Burton (1997).r=0.28, m=1.3e-17, c=-287067

We have used high resolution N(HI) observations of LMC from Kim et al (1998) to measure the neutral hydrogen column density and Galactic NHI for our observations from 21 cm observations of Leiden/Dwingeloo Survey (LDS: Hartmann and Burton 1997) of the sky. There is well consensus that the diffuse ultraviolet emission correlates Galactic neutral hydrogen column density but in LMC case is different.

[edit] 30 Dor

This is the Tarantula Nebula which is an HII region. The energy for this nebula comes from a star cluster comprised of O stars. We have two FUSE pointings in the 30 Dor complex as well as several others near SN1987a.

[edit] Cole et al 1999 WISP Paper

1. Starlight from the N11 complex and the LH 15 association are the strongest contributors to the scattered light component of LMC's diffuse galactic light.

1. Diffuse UV back ground in uncrowded areas, rises from a minimum of (5.63.1) x 10^-8 to (9.31.1) x 10^-8 ergs s^-1 cm^-2 sr^-1 A^-1 in regions near the bright associations. In Photon units this is from 60571190 to 100681190 Ph s^-1 cm^-2 sr^-1 A^-1 at 2150 A.

1. From there mean polarisation map and UV surface brightness around OB associations, they concluded that scattered light contributes at least 20% to the UV surface brightness outside the associations and possible more that 50%.

1. Scattered fraction is higher where stellar densities are lower.

[edit] Maucherat et al A&A 86 299 1980 Paper

1. Taking D2B Aura Satellite observation Maucherat et al have given the surface brightness maps of the LMC after removing the contribution of the bright stars.They found intense variation in UV emissions.

1. They have given the surface brightness variation over the three wave length band (3100A, 2200A, 1690A) which is given below.

[edit] LMC References