ANÁLISIS DESCRIPTIVO DE LA VARIABLE INDEPENDIENTE REFERIDO A LA GESTIÓN DEL CONOCIMIENTO.

Figure 4.7: Image of the best-fit model to the IRAC data only for IRAS 04368+2557, shown in the same bands (IRAC 3.6, 4.5, and 8.0µm) and with the same scaling as the image in Figure 4.5.

As mentioned previously, although the models include inner holes, these holes are completely evacuated, rather than having low levels of dust. A small but non-zero amount of dust inside an inner hole may be enough to provide the IRAC fluxes observed, while decreasing the amount of far-infrared flux. The parameters of the best-fitting models for the IRAC data only, and for the full SED, are listed in Table 4.7. As well as demonstrating SED fitting of resolved sources, this is an illustration of how detailed modelling of an individual source can be used to improve the models in future.

Since this work was carried out, Tobin et al. (2008) have suggested that both the SED and the IRAC images can be well reproduced by a model in which a ‘double’ bipolar cavity is present, with a large scale bipolar cavity corresponding to that seen in the IRAC images, and a smaller-scale cavity with a smaller opening angle in the inner hundred AU. The smaller opening angle of the ‘inner’ cavity intercepts more radiation, scattering more stellar light at mid-infrared wavelengths, and therefore increasing the amount of mid-infrared emission relative to the longer wavelengths, and producing a central peak of emission as seen in the IRAC images. Near-infrared HST or ground-based adaptive optics observations could verify whether this is indeed the case.

4.4

Summary and Conclusions

A method to fit observed YSO SEDs using a pre-computed set of models was presented. Al- though the values of 14 physical parameters were varied in the models from Chapter 3, the

in six apertures, then fitting the full SED including the multi-aperture IRAC data.

aim of this method is not to determine all of these parameters from fitting observed SEDs, but to determine which parameters can be constrained, if any, and if so, what range of val- ues provide acceptable fits. Data from UBVRI to sub-mm wavelengths were compiled for 30 YSOs in the Taurus-Auriga star-forming region. The evolutionary stages, stellar temperatures, disk masses, and disk accretion rates derived from fitting the model SEDs to the data are in good agreement with independently determined values (for example from spectroscopy). In cases where the best fitting model is not in agreement with the literature value, therangeof parameter values of the model SEDs that provide a good fit is in general still consistent with this value. An analysis was also presented to show how adding fluxes at various wavelengths helps constrain different parameters. For example it was found, as in Chapter 3, that data in the range 20−100µm in addition to shorter wavelength data are very useful in constraining parameters such as the envelope accretion rate, and thus the evolutionary stage of a source. An online fitting tool making use of this method is available to the community[4]. In future, this method will be extended to fit spectra (e.g. SpitzerIRS data) and polarisation measure- ments simultaneously with the broadband SED, in order to place further constraints on the geometry and the chemistry of the circumstellar environment of YSOs.

5

Red Sources in the Galactic Mid-Plane

A highly reliable flux-limited census of 18,949 point sources in the Galactic mid-plane that have intrinsically red mid-infrared colours is presented. These sources were selected from the SpitzerGLIMPSE I and II surveys of 274 deg2 of the Galactic mid-plane, and consist mostly of high- and intermediate-mass YSOs and AGB stars. The selection criteria were carefully chosen to minimise the effects of position-dependent sensitivity, saturation, and confusion. The dis- tribution of sources on the sky and their location in IRAC and MIPS 24µm colour-magnitude and colour-colour space are presented. Using this large sample, it is shown that YSOs and AGB stars can be mostly separated by simple colour-magnitude selection criteria into approx- imately 50−70 % of YSOs and 30−50 % of AGB stars. Planetary nebulae and background galaxies together represent at most 2−3 % of all the red sources. 1,004 red sources in the GLIMPSE II region, mostly AGB stars with high mass-loss rates, show significant (≥0.3 mag) variability at 4.5 and/or 8.0µm. With over 11,000 likely YSOs and over 7,000 likely AGB stars, this is to date the largest uniform census of AGB stars and high- and intermediate mass YSOs in the Milky-Way Galaxy.

they also show the distributed star formation between these regions. Therefore, in addition to focusing on specific regions, a whole continuum of star formation environments can now be studied. When seen in this light, these observations have the potential to revolutionise our view of Galactic star formation.

As previously described in §2.3, these surveys are not the first of their kind at mid- and far-infrared wavelengths. Previous major surveys covering the Galactic plane include the IRAS all-sky survey in 1983 at 12, 25, 60, and 100µm, the ISO Galactic plane survey (Omont et al., 2003) at 7 and 15µm, and the MSX survey of the Galactic plane (Price et al., 2001) at 8.28, 12.13, 14.65, and 21.3µm. However, the combination of coverage, sensitivity and resolution of theSpitzerobservations is unprecedented: the FWHM of the PSF is 2′′at 8µm, and 6′′at 24µm, compared to the detector resolution of 18.3′′for MSX 8.28 and 21.3µm and a FWHM of approximately 3-5′ _{for IRAS 12 and 25µm. The point source sensitivity at 8µm is 100}

and 1,000 times better than MSX 8.28µm and IRAS 12µm respectively, and the sensitivity at 24µm is also approximately 100 and 1,000 times more sensitive than MSX 21.3µm and IRAS 25µm respectively. At 7µm, the sensitivity and resolution of the ISOGAL observations (9 mJy and 6′′ _{respectively) approach those of theSpitzerGLIMPSE survey, but the coverage of the}

ISOGAL survey is only 6 % of that of the GLIMPSE and MIPSGAL surveys (16 deg2_{for ISOGAL} versus 274 deg2 for GLIMPSE).

These three previous surveys have been used to search for YSOs, which are brighter and redder at infrared wavelengths than field stars due to thermal emission from circumstellar dust. Wood & Churchwell (1989) selected 1,717 candidate embedded massive stars with UCHII regions from the IRAS data, 1,646 of which are associated with the Galactic plane; Felli

5.2. Observations