Continuum Veiling

The question of what creates non-chromospheric Ha is a difficult one that is far from resolved. Before delving into the matter, let us first complete our description of the emission characteristics of T Tauri stars generally. Lines in the ultraviolet region of the spectrum also attest to elevated activity. This radiation was first extensively studied using the International Ultraviolet

Explorer (IUE) satellite, launched in 1978. The spectrum of classical T Tauri stars from 1100 to 3100 A exhibits, at modest resolution, a smooth continuum with superposed emission lines. Transitions are partially from neutral and singly ionized species, as in the optical. The strongest flux is from the Mg II h and k doublet near 2800 A. However, there are also lines from such ions as C IV and Si IV. These latter transitions require ambient temperatures near 105 K.

Both the roster of ultraviolet emission lines and even their flux ratios are similar to those observed in active dwarfs, including the Sun. This correspondence again implicates a chromosphere and corona in the pre-main-sequence objects. On the other hand, individual flux levels can be vastly greater than solar, three orders of magnitude or even more. Our reasoning concerning Ha would then suggest a non-chromospheric origin for the ultraviolet lines. Complicating this puzzle is the fact that the richness of the spectrum at the shortest wavelengths bears little relation to emission strength in the optical regime. To cite but one example, the extreme T Tauri star RW Aurigae displays few of the far-ultraviolet lines that originate from very high-temperature gas.

We stress that this apparently contradictory behavior applies only to spectral lines. The ultraviolet continuum flux in classical T Tauri stars is always greater than that of a weak-lined or main-sequence star, and to a degree that correlates well with optical emission activity. That is, the excess emission at the shortest observable wavelengths appears to be a smooth extension of that seen in the optical.

What creates the additional continuum radiation? We cannot hope to address this question without first establishing the wavelength dependence of the emission. That is, we must separate out the spectral energy distribution of the continuum excess from that of the underlying photosphere. This separation is currently feasible only in the optical regime, where the spectral resolution is highest, and even there requires some care. Suppose one has the spectrum of a classical T Tauri star and that of a weak-lined object with nearly the same effective temperature but negligible veiling. Then one cannot simply difference the two spectra and obtain the excess. All observed fluxes must first be dereddened to eliminate the effect of intervening dust. Dereddening, in turn, requires comparison of an object's apparent colors with those that would be seen by an observer close to the star. This correction is straightforward for the weak-lined star. In a classical T Tauri star, however, the observed colors are influenced both by reddening and by the presence of the continuum excess.

We may disentangle the two effects through detailed comparison of absorption line strengths in the classical and weak-lined stars. Recall that lines in the former are shallower as a result of the veiling. Furnished with high-resolution spectrograms of both objects, we may find the relative excess emission within small wavelength intervals.

Suppose, as in the top panel of Figure 17.4, that we have portions of a weak-lined and classical spectrum, both with a continuum flux level of F0. Suppose further that the observed depths of a particular absorption line in the weak-lined and classical stars are SF0 and SF1, respectively. If we add a broadband, veiling flux AFX to the weak-lined spectrum, we preserve the depth of the line, but raise the background continuum flux from F0 to F0 + AFX. We then rescale the augmented, weak-lined spectrum until the continuum level is again F0. If we have added the proper veiling, then the new line depth matches 5F\, the one in the classical star:

We define the veiling index, r\, for the classical star as AF\/F0. The reader may verify that, in a wavelength interval where rx = 1, we would have SF0 = 2 SF1. That is, the lines in the classical star would be half as deep as in the weak-lined object with no veiling.

The final step in the procedure is to take a dereddened, low-resolution spectrum of the weak-lined star and multiply it at each wavelength by rx. One thus obtains the properly dereddened AF\ for the continuum excess. Figure 17.4 shows the calculated result for BP Tau. This is the moderately active T Tauri star whose full optical spectrum we displayed earlier. Here, we show the specific flux relative to that in a weak-lined, template star at 5500 A. It is clear that the optical excess represents a significant fraction of the star's radiative output in this wavelength range. Discounting observational noise, the flux distribution appears featureless, and declines gently from 4000 to 6800 A. The simplest model consistent with this result is a geometrically thin, gaseous slab with a temperature near 10,000 K. Since the absorption lines of BP Tau are only partially filled in, i. e., since rx never greatly exceeds unity, the putative slab must cover only a small fraction (a few percent in this case) of the stellar surface.

Other classical T Tauri stars are amenable to such modeling, although the parameters of the slab vary widely. It is not difficult to find a more physical basis for this result. As we shall see in § 17.2, there is ample evidence for infalling gas surrounding many T Tauri stars. There are also, curiously enough, signs of outflow, often within the same object. The hot gas crudely represented as homogeneous slabs must actually arise in shocks located above or on the star's surface. The sparse areal coverage of a typical slab reflects a corresponding sparsity of the shock loci, where either outflowing gas impacts external material, or infalling streams collide with the surface layers.

17.1.4 Infrared Excess

The continuum excess produced by T Tauri stars spans a much broader range in wavelength than just the ultraviolet and optical. Figure 17.5 displays the full spectral energy distributions for the same three stars that were shown in Figure 17.1. The weak-lined star V830 Tau peaks in emission near 1 pm and differs little from a main-sequence star of the same spectral type. In contrast, the broadband spectrum of BP Tau has a far shallower decline toward the infrared. Finally, DR Tau has so much continuum emission at all wavelengths that the peak of its spectral energy distribution is shifted redward by a large amount.

We have encountered the infrared excess before, of course. According to the classification scheme introduced in Chapter 4, weak-lined stars such as V830 Tau fall into Class III, since they exhibit no excess in the near-infrared. The moderately active BP Tau is a Class II source. So is the extreme (or "continuum") star DR Tau, since A F\ still declines between 2.2 and 10 pm. A true Class I object, we recall, has rising flux in this interval and a spectral energy distribution that may peak at wavelengths as large as 100 pm. We argued in Chapter 11 that these sources are also likely to be pre-main-sequence stars, but ones that are buried within especially massive and dense gaseous envelopes.

Veiling Optical Infrared
Figure 17.4 Top panel: Determining the veiling flux in a classical T Tauri star. Bottom panel: Spectral energy distribution of veiling emission in the star BP Tau. The specific flux is normalized to F◦, the value in a weak-lined, template star at 5500 A. The curve is a linear fit to the data.

Returning to Figure 17.5, the star BP Tau is the only one for which we may reliably obtain an infrared excess. Weak-lined stars have essentially no excess, while the photospheric emission cannot be accurately gauged for DR Tau. Such extreme objects comprise only about 10 percent of all classical T Tauri stars. Thus, there exists a large number of documented infrared excesses. As we shall discuss in § 17.3, the most plausible source for this emission is a circumstellar disk. Much of the "excess" radiation simply represents incident stellar photons reprocessed by dust

Wavelength log X (|xm)

Figure 17.5 Broadband spectral energy distributions for the three stars in Figure 17.1. The spectra for DR Tau and V830 Tau have been displaced up and down, respectively, by A log F\ = 0.5. Note that the fluxes for BP Tau and V830 Tau have been dereddened for interstellar extinction.

Wavelength log X (|xm)

Figure 17.5 Broadband spectral energy distributions for the three stars in Figure 17.1. The spectra for DR Tau and V830 Tau have been displaced up and down, respectively, by A log F\ = 0.5. Note that the fluxes for BP Tau and V830 Tau have been dereddened for interstellar extinction.

into the infrared. In some cases, the disk itself may be supplying additional energy. If so, the underlying physics is still unclear.

Having an infrared excess changes a star's intrinsic colors, as measured in the standard bandpasses. We discussed in Chapter 4 how near-infrared color-color plots are useful for picking out the classical T Tauri stars within populous groups. Figure 4.2 shows, in particular, that the observed H — K values are too large relative to J — H for the objects of interest, even after accounting for interstellar reddening. Of course, a reliable estimate for AV allows one to dered-den any single color index. Comparison with a main-sequence color then gives a quantitative measure of the amount of excess.

A dereddened color, such as (H — K)◦ or (K — L)a, is thus useful and far more convenient than a direct assessment of the total, additional luminosity at longer wavelengths. Figure 17.6 compares one such index, (K — L)a, with another, ostensibly unrelated one, the equivalent width of Ha. Here, the sample population consists of stars in Taurus-Auriga. Despite the large scatter, there is evidently a relation between the two numbers. Weak-lined stars, i. e., those with WHa < 10 A, have (K — L)a ranging from 0.1 to about 0.3 mag. This is also the expected range for main-sequence stars of the appropriate spectral types. Classical T Tauri stars with higher Ha fluxes show correspondingly stronger infrared excesses. This trend illustrates the fact that the permitted line emission and the continuum infrared radiation both depend on the presence of circumstellar matter.

Figure 17.6 The Ha equivalent width of T Tauri stars plotted against their dereddened, K — L color index. All stars are located in the Taurus-Auriga association.

Figure 17.6 The Ha equivalent width of T Tauri stars plotted against their dereddened, K — L color index. All stars are located in the Taurus-Auriga association.

Was this article helpful?

0 0

Post a comment