head 1.2; access; symbols; locks http:1.2; strict; comment @# @; expand @b@; 1.2 date 2005.01.19.16.57.34; author PaoloPadovani; state Exp; branches; next 1.1; 1.1 date 2004.12.22.12.48.51; author PaoloPadovani; state Exp; branches; next ; desc @none @ 1.2 log @none @ text @The contribution from low and intermediate mass stars to the ISM by Pedro Garcia-Lario (ESAC), Florian Kerber (ST-ECF), and Anita Richards (Jodrell Bank) ---------------------------------------------------------------- The knowledge of the chemical evolution of stars of low- and intermediate-mass (from 1-8 solar masses) is crucial to determine the overall contribution of nuclear-processed material from these stars to the ISM and, subsequently, to develop models which can predict the chemical enrichment of galaxies in general. Although all stars are born O-rich (C/O ratio less than unity) in the local Universe, several dredge-up processes take place during the evolution of low-mass and intermediate-mass stars which may eventually turn their atmospheres into C-rich (C/O ratio larger than unity) environments. This switch between O-rich and C-rich chemistry has an enormous impact on the nature of the gas and dust returned to the ISM and usually takes place during the latest stages of the evolution of these stars in the so-called 'Asymptotic Giant Branch' phase or 'AGB'. The processes involved are known to be (according to dredge-up and hot bottom burning theoretical models) strongly dependent on the main-sequence mass of the star and on its metallicity. Clues to the C-O transition can be found in a rare class of objects displaying lines indicative of both O- and C-dominated chemistries. One possibility is that these posess a fossil O-rich shell (e.g. with dust showing silicate features, or water masers) and a newly C-enriched CS (central star). The study of stars on the AGB is hampered from the observational point of view for two main reasons: firstly, this is a short-lived phase (10^5-10^6 years), so that there are only a limited number of stars evolving in this phase at any given moment. The transition from 0- to C-rich can happen at any time during this evolutionary stage (in some cases only at the very last moment after a so-called 'late thermal-pulse' which may be triggering the formation of WC-type CSPNe. Secondly, AGB evolution is dominated by heavy mass loss (driven by radiation pressure on grains), which results in the formation of thick circumstellar shells of gas and dust. Mass loss from the more massive stars is so huge (10^-4 solar masses/year) that the shells become optically thick at visible wavelengths (sometimes even beyond). The CS loses most of its original stellar atmosphere, and appears as a heavily obscured source in the optical range being detectable only at infrared (or radio) wavelengths. Such observations also reveal a wealth of molecular and even solid species in the stellar ejecta. It only reappears (usually after a few ~10^3 years), first in the near infrared and then in the optical, when the mass loss has ceased. By this time, the envelope has expanded sufficiently to become diluted in the ISM and thus becomes optically thinner. Planetary Nebulae (PNe) result from this heavy mass loss by AGB stars. They are understood in terms of the two-wind model by Kwok et al. (1978, ApJ, 219, L125) as the product of the mass-loss history on the AGB and the CS evolution. After leaving the AGB the CS becomes hotter. A fast collimated wind is often observed in proto-PNe. Upon reaching ca. 30000 K the CS is able to ionize the hydrogen in its vicinty and a PN becomes observable. During the PN phase the temperature of the CS can reach in excess of 100000 K. The resulting hard UV radiation creates a fully ionized PN plasma which is rich in the recombination lines of H and He and strong forbidden lines from heavier elements (C, N, O, S, Ne etc.). This makes it possible to derive the chemical abundances in the matter ejected on the AGB. The hot CS gives off a fast thin wind which interacts with the much slower but denser AGB ejecta shaping the complex morphologies of PNe. Over recent years, it has been recognized that the interaction of PNe with the surrounding ISM becomes a very important process in the late stages of PN evolution. Once considered a curiosity only seen in a few odd examples, like A35 recent survey work has shown that it is commonly found in old PNe. These old PNe represent the final observable stages during which the matter, which was processed inside the star and expelled during the AGB phase, is actually returned to the ISM. When a PN ages the density drops because of the expansion of the nebular shell; at some point the ISM pressure upstream will become comparable to the pressure in the PN shell and interaction with the ISM is initiated. The gas in the shell will be compressed which results in an asymmetric brightness distribution, while the shape remains largely spherical. Later in the evolution the density decreases further and finally the expansion of the PN is significantly slowed upstream, first leading to a deformation of the nebula and later to its disruption. During this process the CS, which is not affected by the slowing, will move out of the geometrical center of the PN, subsequently leaving its nebula behind, as observed in Sh2-174. The final result of this evolution will be a white dwarf stripped off its nebula, which mixes with the ISM. In a recent survey Kerber et al. have found that in about 50% of interacting PNe the CS is no longer located in the geometric center of the PN. This is a direct manifestation of the galactic orbital motion of the CS after the interaction process has decoupled the motion of the nebula and its CS. Recent studies (we have done some work here and I can add references if needed) have demonstrated that the relative motion between the CS (plus PN) and the ISM is determining the details of the interaction process. This is a direct manifestation of the Galactic orbital motion of the CS: - low-eccentricity, low inclination orbits (typical for thin disk population) result in low relative velocities and mild forms of interaction, the PN basically fossilizes and slowly diffuses into the ISM. - high-eccentricity, high inclination orbits (typical for thick disk or halo population) results in high relative velocities and violent forms of interaction in which the PN shell is broken apart and significant amounts of matter is stripped off the PN along its path. The interstellar magnetic field may also play a very important role in the PN shell break up and the mixing of the matter into the ISM. The interaction of the stellar and interstellar fields can be studied using radio lines from magnetisable molecules in (and before) the formation of young PNe, whilst IR polarimetry reveals the magnetic field acting on dust grains from this stage onwards. A more quantitative understanding of these processes will provide valuable insight into the chemical evolution of galaxies. In this context it is clear that the study of the late stages of stellar evolution of low- and intermediate-mass stars can strongly benefit from the VO ability: - to perform multi-wavelength analysis of individual sources in order to interpret the rapid changes that these stars experience in such a short period of time, in both continuum and spectral features, especially in the transition phase from the end of the AGB until a new PN is formed. - to obtain data from different instruments which can cope with objects which are suddenly changing from optically bright to heavily obscured infrared sources and back again. - to discover new sources in this short transition phase hidden in existing catalogues and archives in order to improve the reliability of diagnostics for the analysis of data on candidate transition sources. At present the literature only contains a a very limited sample of sources which are known to be in this elusive evolutionary stage. - to combine proper motion information with radial velocity data in order to derive the kinematic properties, orbits and population membership of these stars. VO Tools that will be needed: `-----------------------------' i) for multi-wavelength analysis of individual sources; This can be done by putting together photometric and spectroscopic data taken with various ground-based and space facilities to determine the mass loss history and the overall chemical properties of the circumstellar gas and dust ejected by the central star and injected in the ISM as a function of the stellar mass/luminosity (e.g. using VOSpec as starting point). ii) for the discovery of new candidate sources in this AGB to PN transition phase; This can be done by determining the characteristic location of transition sources in multi-colour-colour diagrams and/or colour-magnitude diagrams. This would increase the number of sources known to be evolving along this short-lived transition phase in our Galaxy and in other galaxies to study the impact of different metallicity environments on the observed chemistry and on the efficiency of dust formation (e.g. using existing filtering facilities in Aladin and in VOPlot combined with cross-match plugins as the starting point for future developments which are now being extended to cross-match capabilities against SIMBAD to characterize the colours of known sources and to identify new sources not yet identified in the literature in the diagrams). iii) for studying the rapid SED changes that these stars experience; This can be done through the systematic characterization of the SEDs associated to stars known to be evolving in the AGB to PN transition phase. For this purpose, automated classification systems (based on neural networks, knowledge-based systems, or similar) should be developed under the VO environment. These systems must be able to learn by themselves and improve their classification scheme as a consequence of the experience achieved during the classification process (this will require the development of specialized web services with these goals). Some partners in VOTech have an interest in developing such systems. As an alternative, web services can certainly be developed which should be able to perform systematic comparisons of a large number of SEDs against a given model (or a set of models) and determine the main physical and chemical properties of the central star and of the circumstellar shell as a result of this comparison. (e.g. given a SED, determine the best model fit, or given a model, search for sources that fit a given evolutionary stage predicted by that model,...) iv) for studying the physical and chemical evolution of the star and the chemistry of the mass-loss; In some cases the nature of the CS has changed since human records began. Data mining and access to the AAVSO data base will reveal long-term photometric trends, whilst spectra show transitions from e.g. M to F-type. To date, there has been no attempt at a comprehensive correlation of observed pulsational and chemical changes with the rate and nature of mass loss, although models exist. Line spectroscopy and imaging shows changes in the chemistry of the shell as material moves away from the star, such as the history of the O-C transition. Automatic identification of diagnostic lines (in all wavelength regimes) could be provided by a specialised web service. v) for high-resolution studies of the morphology and kinematics of CSE and PNe This requires optical/IR/radio data (depending on evolutionary stage) with high spatial or spectral resolution or both. Stellar astrometry will reveal evolutionary variability, proper motions and binarity; VOs provide access to data over a long time baseline and need to develop access to tools for astrometric solutions. Space-based or interferometric imaging reveals the nebular morphology and (with good astrometry) the relative motion of the CS with respect to its shell. Surveying and classifying the appearance of a large number of PNe (resloved in the radio or optical) and the properties of their central stars (spectra in the uv/optical/IR) will contribute to settling the heated debated about whether all asymmetric PNe are binary and whether an off-centre CS position is due to a companion or to its relative motion with respect to the nebula as outlined above. The expansion velocity is deduced from multi-epoch imaging. The kinematics of the shell are measured using 1-, 2- or 3-dimensional spectral data (which also yields the distance if the expansion velocity is known). VO tools currently under development can be extended to handle long-slit spectroscopy, data cubes etc. and to handle velocity axes, complemented by increased publication of radio spectral archives. vi) for using polarimetry to measure magnetic fields. Existing tools can display spectra and images in the various Stokes or other polarization parameters but the IVOA needs to establish standards to characterise polarization data before we can provide access to analysis tools. We also need a means of displaying and analysing polarization angle vectors. In liason with experts in this field, VOs will eventually provide access to specialised services for comparisons with models of the inferred magnetic fields. As well as their role in the dispersal of PNe, magnetic fields of stellar origin influence the shape of the circumstellar shell. Combined with the studies in v), systematic comparison of the properties of objects known to be binary or to lack a significant companion, will enable us to use the nebular morphology and polarization characteristics of obscured systems to deduce whether or not the CS has a companion, and relate this to the chemistry of material returned to the ISM. All six points above can be generalized to other scientific goals other than the analysis of the AGB to PN transition phase. For example: i) the on-the-fly generation of SEDs for multi-wavelength analysis can be applied to any kind of astronomical sources ii) the search for astronomical sources of a given class based on colour-colour diagrams and/or colour-magnitude diagrams can be extended as well to any type of astronomical source showing characteristic colour properties. iii) the automated classification of SEDs can be used for spectral type classification of normal or peculiar stars, for the derivation of chemical abundances in stars or nebulae, comparisons with photoionization models can be applied to PNe or HII regions, with circumstellar disk models in the case of YSOs, with galaxy models trying to determine whether the observed SEDs are compatible with the unification scheme, etc. iv) tools for the manipulation of heterogenous spectral data (for example, comparing radio and IR line systems on the same velocity scale) will assist the interpretation of present and near-future observations nearby galaxies on scales small enough to distinguish the influence of interactions, star-forming regions of varying ferocity, etc., on chemical gradients. v) and vi) contain important tools for studying star formation, such as the kinematics of protostellar discs, in the transition from accretion to planet formation, and the role of magnetic support in cloud collapse. Star forming regions are also heavily obscured and have a complex 3D structure. VO methods can be used to investigate fully a 'learning set' of well-understood objects with multi-wavelength multi-resolution observations and establish diagnostics for the age, mass etc. of other protostars for which only partial information is available. Why the VO approach is unique ? `-------------------------------' The VO approach will be a new, powerful tool for: i) multi-wavelength analysis of individual sources: For this purpose the VO should be able to put together in a plot data (photometry and/or spectroscopy) taken from various archives/catalogues in order to have a quick look to the overall SED. The tool should be intelligent enough to cope with: - different physical units - different beams/apertures - especially for extended sources - astrometric information and its uncertainties - again needed for extended sources - different dates when data were taken (for variability analysis; the combination of multi-wavelength data taken at different epochs can in principle be used to derive light curves at different wavelengths) - display on a velocity axis if suitable metadata are provided - polarized spectra - different photometric systems (ideally a web service should exist providing a database of filter profiles, central wavelengths, zero points, colour corrections,...) - different quality of the input data (data quality information should be always attached to any data plotted in a SED to facilitate reliability analysis and evaluate the need of retrieving original files from the contributing archive/ catalogue if needed) The required information must be found either attached to any input data ingested in the VO as metadata information or must be retrievable from specialized web services providing this information (e.g. the filter database information) automatically during the ingestion process. VO tools such as SpecView and VOSpec can already perform some of these tasks, such as the construction of SEDs from multiple input tables in heterogenous formats by VOSpec. These tools can perform basic analysis e.g. BB or line fitting. However, VOs should encourage data and model providers to supply user-friendly interfaces to more specialised tools for, e.g., multiple BB fitting, comparison with existing models, automated classification of SEDs under a given scheme, line identification linked to chemical databases. The job of the VO is to develop standards which will allow data to be transferred between such services and provide the interface for easy access. ii) for the discovery of new candidate sources in this AGB to PN transition phase VO should provide: - Filtering facilities applicable: * before loading the catalogue from Vizier or any other data server so that only a subset of a given catalogue/data set is ingested for further manipulation with the VO (according to more or less complex user selected criteria). * at any moment during the VO session: as a real-time decision of the user. - Manipulation of columns * especially interesting is the creation of new columns from existing ones (already exists in current VO prototype) - Cross-matching tools: - between multiple catalogues containing a very large number of sources (at least 10^5-10^6 sources each or even more in the future) all them previously ingested as VOtables in the VO session - between one catalogue (or a subset of a catalogue which may have resulted from the application of filter(s) within the VO session) and an external catalogue in Vizier when the latter is too big to be ingested as a VOtable during a VO session ('query by list' option which has just been implemented in the latest version of the VO prototype). - between one catalogue (or a sub-catalogue) and the SIMBAD database (this is very important to distinguish between sources which are well known in the literature and those which are not yet identified in SIMBAD for discovery purposes). The result of this cross-match could be one or more new columns added to the original VOtable incorporating for instance the classification code assigned to that source in SIMBAD (best match) and the number of references in the bibliography (also recently implemented in the AVO prototype although still the output is an independent VOtable which still needs to be cross-matched against the original list) - the cross-matching tools should be able to deal with the astrometric uncertainties of the sources associated to individual entries in the catalogues to be matched and to return an estimate of the reliability of each cross-identification (potentially incorporating statistical methods using parameters such as sample completeness). - Easy comparison of spectra and images (superposition of slit direction, aperture size etc.) to identify transition sources where molecular emission is associated with the distinctive morphology of a fossil shell. iii) for studying the rapid SED changes that these stars experience VO role here should be to provide a link to specialized web services as explained above. Both automated classification of SEDs and comparison with models should be considered under this area of development. Eventually, it may be possible to schedule robotic telescope observations via VOs (e.g. the ESTAR project) for objects which are passing through the sort of rapid changes seen in Sakurai's Object or IRC+10420. iv) for study of PNe - comparison between observed and model spectra of the CS - comparison between observed and model spectra of the PN - combination of proper motion information with radial velocity data in order to derive the kinematic properties (this will benefit tremendously from GAIA ...) - morphological classification using pattern recognition or other automated image analysis of extended objects. The role of VOs is not to write tools, but to provide access to tools as well as data. To this end, we establish protocols and standards in consultation with the astronomical community. This project will need strengthening of these standards, in particular - Metadata standards for describing polarization - Implementation of the metadata velocity standards so that spectral tools can be used to investigate multi-wavelength molecular transitions. In the longer term, the VO may provide interfaces to chemical modelling databases for line identification. In return, VOs need providers of - data - tools - models to liaise with VOs to enable access to data in suitable formats and to their resources via interfaces which also communicate with all other relevant resources. @ 1.1 log @none @ text @d3 2 a4 1 by Pedro Garcia-Lario (ESAC) d7 1 a7 1 The knowledge of the chemical evolution of stars of low- and d9 135 a143 59 overall contribution of nuclear-processed material from these stars to the ISM and, subsequently to develop models which can predict the chemical enrichment of galaxies in general. Although originally all stars are born O-rich (C/O ratio less than unity) in our Universe, several dredge-up processes take place during the evolution of low-mass and intermediate-mass stars which may eventually turn their atmospheres into C-rich (C/O ratio larger than unity) environments. This switch between O-rich and C-rich chemistry has an enormous impact on the nature of the gas and dust returned to the ISM and usually takes place during the latest stages of the evolution of these stars in the so-called 'Asymptotic Giant Branch' phase or 'AGB'. The processes involved are known to be (according to dredge-up and hot bottom burning theoretical models) strongly dependent on the main-sequence mass of the star and of the metallicity. The study of stars in the AGB phase, however, is hampered from the observational point of view, mainly because of two different reasons: first, this is a short-lived phase (10^5-10^6 years) when the stars evolve very fast in the H-R diagram and, thus, the number of stars evolving in this phase at a given moment is limited. Second, this is a phase in which the evolution is dominated by heavy mass loss (driven by radiation pressure on the dust grains), which results in the loss of most of the original stellar atmosphere and the formation of thick circumstellar shells of gas and dust in a process which leads to the formation of planetary nebulae. In the most extreme cases (the more massive stars) the mass loss is so huge (10^-4 solar masses/year) that the shells become optically thick at visible wavelengths (sometimes even beyond) and the central star does not show an optical counterpart for some time. Meanwhile the star is only detectable at infrared (or radio) wavelengths. Only at a later stage (usually after a few ~10^3 years) the central star reappears again first in the near infrared and then in the optical only when the mass loss has completely ceased and the envelope is diluted in the ISM as a consequence of its expansion (thus decreasing its optical thickness). In this context it is clear that the study of the late stages of stellar evolution of low- and intermediate-mass stars can strongly benefit from the VO ability: - to perform multi-wavelength analysis of individual sources in order to interpret the rapid changes that these stars experience in such a short lapse of time - both in the continuum and in the spectral features - and be able to cope with objects which are suddenly turning from optically bright to heavily obscured infrared sources) - to discover new sources in this short transition phase hidden in currently existing catalogues (to improve the validity of the results obtained from the analysis of the still very limited sample of transition sources available in the literature which are known to be in this elusive evolutionary stage). In addition to this, I would expect the VO to be able to perform in the future other interesting tasks like: automated classification of SEDs, comparison with available models, etc. (this section can be extended with the description of existing models; may be Florian Kerber can help me here,..) d145 1 a145 1 VO Tools that will be needed: d148 1 a148 1 i) for multi-wavelength analysis of individual sources d150 135 a284 62 This can be done by putting together photometric and spectroscopic data taken with various ground-based and space facilities to determine the mass loss history and the overall chemical properties of the circumstellar gas and dust ejected by the central star and injected in the ISM as a function of the stellar mass/luminosity (e.g. using VOSpec as starting point) ii) for the discovery of new candidate sources in this AGB to PN transition phase This can be done by determining the characteristic location of transition sources in multi-colour-colour diagrams and/or colour-magnitude diagrams. This would increase the number of sources known to be evolving along this short-lived transition phase in our Galaxy and in other galaxies. to study the impact of different metallicity environments on the observed chemistry and on the efficiency of dust formation (e.g. using existing filtering facilities in Aladin and in VOPlot combined with cross-match plugins as the starting point for future developments - should be extended to cross-match against SIMBAD to characterize the colours of known sources and to identify new sources not yet identified in the literature in the diagrams). iii) for studying the rapid SED changes that these stars experience This can be done through the systematic characterization of the SEDs associated to stars known to be evolving in the AGB to PN transition phase. For this purpose, automated classification systems (based on neural networks, knowledge-based systems, or similar) should be developed under the VO environment which must be able to learn by themselves and improve their classification scheme as a consequence of the experience achieved during the classification process (this will require the development of specialized web services with these goals). As an alternative, web services can certainly be developed which should be able to perform systematic comparisons of a large number of SEDs against a given model (or a set of models) and determine the main physical and chemical properties of the central star and of the circumstellar shell as a result of this comparison. (e.g. given a SED, determine the best model fit, or given a model, search for sources that fit a given evolutionary stage predicted by that model,...) All three points above can be generalized to other scientific goals other than the analysis of the AGB to PN transition phase. For example: i) the on-the-fly generation of SEDs for multi-wavelength analysis can be applied to any kind of astronomical sources ii) the search for astronomical sources of a given class based on colour-colour diagrams and/or colour-magnitude diagrams can be extended as well to any type of astronomical source showing characteristic colour properties. iii) the automated classification of SEDs can be used for spectral type classification of normal or peculiar stars, for the derivation of chemical abundances in stars or nebulae, comparisons with photoionization models can be applied to PNe or HII regions, with circumstellar disk models in the case of YSOs, with galaxy models trying to determine whether the observed SEDs are compatible with the unification scheme, etc. d286 1 a286 2 Why the VO approach is unique ? d289 1 a289 3 The VO approach will be a new, powerful tool for: i) multi-wavelength analysis of individual sources: d291 1 a291 89 For this purpose the VO should be able to put together in a plot data (photometry and/or spectroscopy) taken from various archives/catalogues in order to have a quick look to the overall SED (much like VOSpec plans to do in the future?) The tool should be intelligent enough to cope with: - different units - different beams/apertures - especially for extended sources - astrometric information and its uncertainties - again needed for extended sources - different dates when data were taken (for variability analysis; the combination of multi-wavelength data taken at different epochs can in principle be used to derive light curves at different wavelengths although this is not first priority at this stage of the project, I guess) - different photometric systems (ideally a web service should exist providing a database of filter profiles, central wavelengths, zero points, colour corrections,...) - different quality of the input data (data quality information should be always attached to any data plotted in a SED to facilitate reliability analysis and evaluate the need of retrieving original files from the contributing archive/ catalogue if needed) All this information must be found either attached to any input data ingested in the VO as metadata information or must be retrievable from specialized web services providing this information (e.g. the filter database information) automatically during the ingestion process. (Note that complex SED analysis tools like: multiple BB fitting, comparison with existing models, automated classification of SEDs under a given scheme, etc. in my opinion should be provided by dedicated web services, and not by the VO plotting facility itself which should just be able to put together the data! That's more than enough!) ii) for the discovery of new candidate sources in this AGB to PN transition phase VO should provide: - Filtering facilities applicable: * before loading the catalogue from Vizier or any other data server so that only a subset of a given catalogue/data set is ingested for further manipulation with the VO (according to more or less complex user selected criteria). * at any moment during the VO session: as a real-time decision of the user. - Manipulation of columns * especially interesting is the creation of new columns from existing ones (already exists in current VO prototype) - Cross-matching tools: - between multiple catalogues containing a very large number of sources (at least 10^5-10^6 or even more in the future) all them previously ingested as VOtables in the VO session - between one catalogue (or a subset of a catalogue which may have resulted from the application of filter(s) within the VO session) and an external catalogue in Vizier when the latter is too big to be ingested as a VOtable during a VO session. - between one catalogue (or a sub-catalogue) and the SIMBAD database (this is very important to distinguish between sources which are well known in the literature and those which are not yet identified in SIMBAD for discovery purposes). The result of this cross-match could be one or more new columns added to the original VOtable incorporating for instance the classification code assigned to that source in SIMBAD (best match) and the number of references in the bibliography. - the cross-matching tools should be able to deal with the astrometric uncertainties of the sources associated to individual entries in the catalogues to be matched and be able to determine by itself whether the match is reliable or not. iii) for studying the rapid SED changes that these stars experience VO role here should be to provide a link to specialized web services as explained above. Both automated classification of SEDs as well as comparison with models should be considered under this area of development. NOTE: I have concentrated my discussion on photometric and spectroscopic data. I have ignored what should be needed to deal with images. May be other people can address this aspect. d293 138 @