The dusty envelopes of asymptotic giant branch stars with ultraviolet excesses

Abstract

Aims. In a first study, we characterised the properties of the gas component in the circumstellar envelopes surrounding a sample of 29 AGB stars with UV excesses. Now we intend to complement this information with an analysis of the dust component and compare the estimated parameters with those previously inferred from larger samples of AGB stars. Methods. We modelled the spectral energy distributions of the sample using dust radiative transfer models. In some cases, we complemented the analysis with Herschel/PACS radial surface brightness profiles. Results. We derived mass-loss rates and gas-to-dust ratios, which are in the typical ranges for AGB stars. We found that the stellar and mass-loss parameters follow similar trends than those presented in the literature. There is an anticorrelation between the gas-to-dust ratio and the UV emission, although it is weaker than its correlations with pulsation and mass-loss. We also estimated the dust attenuation produced by the dust at UV wavelengths and describe its effects on the intrinsic UV emission. Conclusions. Stellar and mass-loss parameters of UV emitting AGB stars follow similar trends as found for larger samples of AGB stars. High-angular resolution observations are required to explore the dust forming regions and identify the presence of stellar companions. Circumstellar dust attenuation might play a dominant role in the observed UV emission, and needs to be accounted to estimate the intrinsic UV emission.

Figures (17)

• Figure 1: SEDs and best-fit DUSTY models for the dusty AGB stars. Solid blue line: input spectra, solid red line: total DUSTY model, dotted cyan line: Attenuated input spectra component, dashed green line: dust scattered component, dot-dashed orange line: dust thermal component. Photometric fluxes are indicated as filled circles (in a few cases the variability-induced uncertainties are larger than the average value), photometric data not included in the analysis as empty circles, and upper limits as empty triangles. Grey solid lines are GAIA, IRAS, and ISO spectra, shadowed areas indicate 10% uncertainties.
• Figure 2: Same as Fig. \ref{['SED_fitting']} for "naked" sources.
• Figure 3: Histograms showing the distributions of the best-fit solution SED modelling parameters. From left to right: Stellar effective temperature, dust inner radius temperature, density power-law index, dust shell radial optical depth at 550 nm. White bins represent "naked" O-rich uvAGBs, grey bins represent O-rich uvAGBs with dust component and striped bins represent C-rich uvAGBs.
• Figure 4: HRD of our the sample. The size of the markers is proportional to the dust shell masses as shown in the bottom-right legend. The markers colour represents the variability type. Red: semi-regulars (SRs), green: irregulars (LB), and blue: Miras (M). The shape of the markers represents the chemistry of the AGB stars. Circles: O-rich ("naked" as asterisks), crosses: C-rich.
• Figure 5: Comparison of the best-fit solution SED modelling parameters (from top to bottom: $T_{*}$, $T_{\rm inn}$, $n$ and $\tau_{\rm 550}$) with $P$ ( left) and $R^{\rm corr}_{\rm FUV/NUV}$ ( right). The colour and shape of the markers represent the stellar variability and chemistry type as in Fig. \ref{['fig:HR_diagram']}. The horizontal line in the third row represents $n$=2. Error bars of $R^{\rm corr}_{\rm FUV/NUV}$ include multi-epoch variability of sources detected in both GALEX bands, arrows indicate that one observation was an upper limit.