next up previous
Next: About this document ... Up: The Limiting Effects of Previous: Discussion and Conclusions

Figure captions


  
Figure: Run of the relative abundances of gas phase (full lines) and crystallized species across a T $_{\rm
eff}=2600$ K model atmosphere typical of the young Pleiades brown dwarfs Teide1 and Calar3. The condensation of perovskite (CaTiO3, dashed line) is the principle cause of TiO depletion in the atmospheres of dwarfs later than about M6. The abundance of the condensate Ca2SiO4 is drawn at log10($\tau$)=-5.0 but is not labeled for sake of clarity.
\begin{figure}
\psfig{file=f1.ps,width=\hsize,angle=0}
\end{figure}


  
Figure: Same as above for a T $_{\rm eff}=1800$ K model atmosphere typical of the reddest known field dwarfs GD165B, Kelu1, and the DENIS objects.
\begin{figure}
\psfig{file=f2.ps,width=\hsize,angle=0}
\end{figure}


  
Figure: The extinction cross-sections per particle of dust grains. The mie formalism is used assuming a power-law ( $\alpha=-3.5$) grain size distribution with diameters from 0.00625 and 0.24 $\mu$m. Monoatomic grains such as Fe, Cu, and Ni contribute scattering at optical wavelengths only, while corundum, magnesium aluminium spinel, calcium titenide, hematite, magnetite, and Ca2Al2SiO7 crystals show strong peaks of absortion, at infrared wavelengths, that could compete with the local water vapor ``continuum'' in hot/young brown dwarfs. Note however that if the grains were elliptical and randomly oriented, the sharp absorption peaks shown here could be washed out.
\begin{figure}
\psfig{file=f3.ps,width=\hsize,angle=0}
\end{figure}


  
Figure: The extinction profiles are compared for grain size distributions with 1, 2, 10 and 100 times the ISM values adopted for this work (full lines from bottom to top respectively, where the two first curves are nearly undistinguishable). The scattering and absorption contributions of the 100 ISM profile are also shown (dotted lines). The conditions are those of the photospheric layers ( $\tau_{1.2{\mu}m}\approx 10^{-4}$, i.e. T $\approx 1300$K) of our standard 1800K AMES-Dusty model atmosphere. The structures seen in the profile at $\lambda > 8.5\mu$m are due to dust absorption (Mg2SiO4 at 10 and 16.5 $\mu$m and MgAl2O4at 13 $\mu$m). Scattering contributions dominate below $0.5\mu$m, and remain modest at longer wavelengths for grain sizes $\leq 10$ times the adopted values. The absoption profile, on the other hand, is only little sensitive to grain sizes.
\begin{figure}
\psfig{file=f4.ps,width=\hsize,angle=0}
\end{figure}


  
Figure: Thermal structures of the AMES-Cond models with T $_{\rm eff}$ ranging from 3000 to 100K by steps of 500K, with two additional models at 700 and 400K, logg=5.0, and solar metallicity. The convection zones are labeled with cross-symbols. The approximate location of the photosphere is indicated with filled circles and triangles marking the $\tau_{\rm 1.2{\mu}m}= 10^{-4}$ and 1.0 optical depths. All models shown stop at $\tau_{\rm 1.2{\mu}m}=
10$. As ${\rm T}_{\rm eff}$ decreases, the photosphere becomes progressively more isothermal.
\begin{figure}
\psfig{file=f5.ps,width=\hsize,angle=90}
\end{figure}


  
Figure: Thermal structures of the fully dusty models with T $_{\rm eff}$ ranging from 2400 to 1600K by steps of 200K, with two additional models at 2500 and 1500K, logg=5.0. None of the curves shown actually overcross. The radiative zones are marked by full lines while the convective region is shown as dotted lines. The location of the photosphere is also indicated, with full circles and triangles marking the $\tau_{\rm 1.2{\mu}m}= 10^{-4}$ and 1.0 optical depths respectively. The strongest optical molecular bands and resonance lines form near $\tau_{\rm 1.2{\mu}m}= 10^{-4}$. All models shown stop at $\tau_{\rm 1.2{\mu}m}=
10$.
\begin{figure}
\psfig{file=f6.ps,width=\hsize,angle=0}
\end{figure}


  
Figure: Thermal structures of models with T $_{\rm eff}
= 2800$K and 1800K, logg=5.0, and solar metallicity for three types of models: (1) the standard NextGen models treated in gas phase only (dotted line); (2) the AMES-Dusty models assuming a full distribution of the dust (full line), and; (3) the AMES-Cond models including dust in the CE but ignoring their opacities (dashed line). All models are converged. Note that the NextGen models use a different source of water vapor opacity (see text).
\begin{figure}
\psfig{file=f7.ps,width=\hsize,angle=0}
\end{figure}


  
Figure: Spectral sequence of brown dwarfs to EGP model atmospheres in the total settling (AMES-Cond) approximation. From top to bottom: T $_{\rm eff}= 2500, 1900, 1300, 700, 400$K, and 200K. The gravity is fixed to $\log g= 5.0$. These models (AMES-Cond) assume complete settling of the grains (i.e. neglects all dust opacity). The spectral resolution has been reduced from 2Å to 30Å by boxcar smoothing in order to make comparison of the spectra easier. We observe that CH4 bands already develop at 2000K in these extremely transparent and cool AMES-Cond atmospheres. They gradually replace water vapor bands while H2O condenses to ice below 300K.
\begin{figure}
\psfig{file=f8.ps,width=\hsize,angle=90}
\end{figure}


  
Figure: To isolate atomic features we compare a ${\rm T}_{\rm eff}= 1000$K AMES-Cond model (full line) with a spectrum obtained by neglecting all molecular lines (dotted lines). The pseudo-continuum is essentially formed by the van der Waals wings of the Na I D and K I resonance doublets at $\lambda$5891,5897Å and $\lambda$7687,7701Å. Weaker lines of Rb I ($\lambda$7802 and 7949Å), Cs I ($\lambda$8523 and 8946Å), and K I ($\lambda$11693,11776 and $\lambda$12436,12525Å) are also seen. The background flux (obtained by neglecting both atomic and molecular lines, not shown) lies outside the plot! The spectral resolution is reduced to 30Å for this illustration.
\begin{figure}
\psfig{file=f9.ps,width=\hsize,angle=0}
\end{figure}


  
Figure: Same as Figure 8 in the optical to near-red spectral range. Here the spectra have been arbitrarily scaled to facilitate the comparison. From top to bottom: T $_{\rm eff}= 2500, 2400, 2300, 2000, 1700$, and 1500K. The spectral resolution is 2Å.
\begin{figure}
\psfig{file=f10.ps,width=\hsize,angle=90}
\end{figure}


  
Figure: Same as Figure 10 for models from 1500K to 100K at 100K steps.
\begin{figure}
\psfig{file=f11.ps,width=\hsize,angle=90}
\end{figure}


  
Figure: Same as Figure 10 where we zoom in on the CH3D band system at 4.55 $\mu$m. From top to bottom: T $_{\rm eff}= 2000, 1700, 1500, 1300, 1000, 800, 600$, and 400K. The CH3D band appears at 1000K at this gravity and for this dust treatment limit. Note that ${\rm T}_{\rm eff}=
2000$K dwarfs are dusty and that, though they don't appear here, CO bands at 4.67 $\mu$m easily are detectable in these hotter atmospheres.
\begin{figure}
\psfig{file=f12.ps,width=\hsize,angle=90}
\end{figure}


  
Figure: Same as Figure 10 where we zoom in on the NH3 band system at 11.012 $\mu$m. The ammonia band system appears at 1000K, along with several other molecular lines essentially due to methane. The spectral resolution is 5Å.
\begin{figure}
\psfig{file=f13.ps,width=\hsize,angle=90}
\end{figure}


  
Figure: Same as Figure 8 in the full dusty (AMES-Dusty) limiting case. From top to bottom: T $_{\rm
eff}= 2500, 2000, 1800, 1600$K, and 1500K. The gravity is fixed at $\log g= 5.0$. Here the strong heating effects of dust opacities prevent the formation of methane bands, and dissociate H2O while producing a hotter water vapor opacity profile, much weaker and transparent to radiation. From 1700K, the grain opacity profiles rapidly dominate the UV to red spectral region, smoothing out the emergent flux into a continuum where only the core of the strongest atomic resonance lines (Na I D and K I doublets) are seen.
\begin{figure}
\psfig{file=f14.ps,width=\hsize,angle=90}
\end{figure}


  
Figure: Two ${\rm T}_{\rm eff}= 2500$K AMES-Cond models of different surface gravity are compared: (1) $\log g= 5.5$ (full line), (2) $\log g= 2.5$ (dotted line). The spectral resolution has been reduced from 2Å to 10Å by boxcar smoothing in order to make comparison of the spectra easier.
\begin{figure}
\psfig{file=f15.ps,width=\hsize,angle=90}
\end{figure}


  
Figure: Same as Figure 15 for ${\rm T}_{\rm eff}= 500$K AMES-Cond models, and (1) $\log g= 4.0$ (full line), (2) $\log g= 2.5$ (dotted line). Here the spectral resolution has been reduced to 30Å by boxcar smoothing. In the inset we show a zoom of the optical to red spectral regime, were we distinguish water vapor bands at 0.93, 0.95 and 1.12 $\mu$m, Cs I resonance transitions at 0.86 and 0.89 $\mu$m, the K I resonance doublet at 0.77, 0.79 $\mu$m, and the cores of a few other lines such as the Na I D doublet bluewards of 0.75 $\mu$m. While molecular bands are moderately affected by the gravity change, the optical background opacity due to the wings of the Na I D and K I doublets is reduced by nearly a factor of 10 in the low gravity model. The latter model is more typical of low mass brown dwarfs and jovian planets.
\begin{figure}
\psfig{file=f16.ps,width=\hsize,angle=90}
\end{figure}


  
Figure: Two ${\rm T}_{\rm eff}=
2000$K AMES-Dusty models of different surface gravity are compared: (1) $\log g= 6.0$ (full line), (2) $\log g= 3.5$ (dotted line). The spectral resolution has been reduced to 10Å by boxcar smoothing.
\begin{figure}
\psfig{file=f17.ps,width=\hsize,angle=90}
\end{figure}


  
Figure: Same as Figure 17 for ${\rm T}_{\rm eff}= 1500$K AMES-Dusty models. While gravity effects are quasi nonexistent redwards of 2.5 $\mu$m, they are quite large in the optical to red spectral region, mainly as a result of enhanced efficiency of dust grain formation at the higher pressures and densities of high gravity atmospheres. A $\log g$-value of 5.5 is typical of most field brown dwarfs discovered since 1996.
\begin{figure}
\psfig{file=f18.ps,width=\hsize,angle=90}
\end{figure}


  
Figure: Two ${\rm T}_{\rm eff}=
2000$K models with $\log g= 5.5$are compared to illustrate the difference between our two limiting cases: (1) AMES-Dusty with full dust opacity (full line), and (2) AMES-Cond with full gravitational settling (no dust opacity, dotted line).
\begin{figure}
\psfig{file=f19.ps,width=\hsize,angle=90}
\end{figure}


  
Figure: Same as Figure 19 for ${\rm T}_{\rm eff}= 1500$K models with $\log g= 5.0$. A 1500K blackbody (dashed line) is overplotted for comparison.
\begin{figure}
\psfig{file=f20.ps,width=\hsize,angle=0}
\end{figure}


  
Figure: The AMES-Cond model for ${\rm T}_{\rm eff}= 1000$K and $\log g= 5.0$ (full line) is compared to the corresponding model (dotted line) used in their analysis of the Gl229B spectrum. model (dotted line) used in their analysis of the Gl229B spectrum.
\begin{figure}
\psfig{file=f21.ps,width=\hsize,angle=90}
\end{figure}


  
Figure: The 10 Gyr NextGen isochrones (dotted), and the $\log g= 5.0$ locii of the Cond (short dashed) and Dusty (full) models are compared to the photometric observations of field stars and brown dwarfs, and to Pleiades objects including the brown dwarfs PPl15, Teide1 and Calar3 (star and filled circle symbols). The field T dwarfs Gliese 229B and SDSS1624 are also shown. Unresolved binarity is reflected in this diagram by a red excess in J-K. Note that the Cond and Dusty models have been shifted in J-K by +0.15 in order to eliminate water opacity source effects in this comparison. The Cond models are computed (i) with the line wings coverage value of 5000Å (long-dashed line), and (ii) with a maximum coverage of 15000Å (short-dashed line) to illustrate the impact of this parameter on the Cond models.
\begin{figure}
\psfig{file=f22.ps,width=\hsize,angle=0}
\end{figure}


next up previous
Next: About this document ... Up: The Limiting Effects of Previous: Discussion and Conclusions
Peter Hauschildt
2001-05-23