Refine
Has Fulltext
- yes (276)
Year of publication
Document Type
- Conference Proceeding (118)
- Article (78)
- Postprint (52)
- Review (15)
- Working Paper (6)
- Doctoral Thesis (4)
- Monograph/Edited Volume (3)
Language
- English (276) (remove)
Is part of the Bibliography
- no (276) (remove)
Keywords
- USA (7)
- United States (7)
- moderne jüdische Geschichte (6)
- modern Jewish history (5)
- 20. Jahrhundert (4)
- 20th century (4)
- 19. Jahrhundert (3)
- Diversity (3)
- 19th century (2)
- Fluoreszenz-Resonanz-Energie-Transfer (2)
Institute
- Extern (276) (remove)
Preface
(2010)
The workshops on (constraint) logic programming (WLP) are the annual meeting of the Society of Logic Programming (GLP e.V.) and bring together researchers interested in logic programming, constraint programming, and related areas like databases, artificial intelligence and operations research. In this decade, previous workshops took place in Dresden (2008), Würzburg (2007), Vienna (2006), Ulm (2005), Potsdam (2004), Dresden (2002), Kiel (2001), and Würzburg (2000). Contributions to workshops deal with all theoretical, experimental, and application aspects of constraint programming (CP) and logic programming (LP), including foundations of constraint/ logic programming. Some of the special topics are constraint solving and optimization, extensions of functional logic programming, deductive databases, data mining, nonmonotonic reasoning, , interaction of CP/LP with other formalisms like agents, XML, JAVA, program analysis, program transformation, program verification, meta programming, parallelism and concurrency, answer set programming, implementation and software techniques (e.g., types, modularity, design patterns), applications (e.g., in production, environment, education, internet), constraint/logic programming for semantic web systems and applications, reasoning on the semantic web, data modelling for the web, semistructured data, and web query languages.
In this paper, we present a finite-state approach to constituency and therewith an analysis of coordination phenomena involving so-called non-constituents. We show that non-constituents can be seen as parts of fully-fledged constituents and therefore be coordinated in the same way. We have implemented an algorithm based on finite state automata that generates an LFG grammar assigning valid analyses to non-constituent coordination structures in the German language.
Generalized Two-Level Grammar (GTWOL) provides a new method for compilation of parallel replacement rules into transducers. The current paper identifies the role of generalized lenient composition (GLC) in this method. Thanks to the GLC operation, the compilation method becomes bipartite and easily extendible to capture various application modes. In the light of three notions of obligatoriness, a modification to the compilation method is proposed. We argue that the bipartite design makes implementation of parallel obligatoriness, directionality, length and rank based application modes extremely easy, which is the main result of the paper.
Morphological analyses based on word syntax approaches can encounter difficulties with long distance dependencies. The reason is that in some cases an affix has to have access to the inner structure of the form with which it combines. One solution is the percolation of features from ther inner morphemes to the outer morphemes with some process of feature unification. However, the obstacle of percolation constraints or stipulated features has lead some linguists to argue in favour of other frameworks such as, e.g., realizational morphology or parallel approaches like optimality theory. This paper proposes a linguistic analysis of two long distance dependencies in the morphology of Russian verbs, namely secondary imperfectivization and deverbal nominalization.We show how these processes can be reanalysed as local dependencies. Although finitestate frameworks are not bound by such linguistically motivated considerations, we present an implementation of our analysis as proposed in [1] that does not complicate the grammar or enlarge the network unproportionally.
The emergence of information extraction (IE) oriented pattern engines has been observed during the last decade. Most of them exploit heavily finite-state devices. This paper introduces ExPRESS – a new extraction pattern engine, whose rules are regular expressions over flat feature structures. The underlying pattern language is a blend of two previously introduced IE oriented pattern formalisms, namely, JAPE, used in the widely known GATE system, and the unificationbased XTDL formalism used in SProUT. A brief and technical overview of ExPRESS, its pattern language and the pool of its native linguistic components is given. Furthermore, the implementation of the grammar interpreter is addressed too.
In this work an extension of CSSR algorithm using Maximum Entropy Models is introduced. Preliminary experiments to perform Named Entity Recognition with this new system are presented.
In the last years, statistical machine translation has already demonstrated its usefulness within a wide variety of translation applications. In this line, phrase-based alignment models have become the reference to follow in order to build competitive systems. Finite state models are always an interesting framework because there are well-known efficient algorithms for their representation and manipulation. This document is a contribution to the evolution of finite state models towards a phrase-based approach. The inference of stochastic transducers that are based on bilingual phrases is carefully analysed from a finite state point of view. Indeed, the algorithmic phenomena that have to be taken into account in order to deal with such phrase-based finite state models when in decoding time are also in-depth detailed.
Temporal propositions are mapped to sets of strings that witness (in a precise sense) the propositions over discrete linear Kripke frames. The strings are collected into regular languages to ensure the decidability of entailments given by inclusions between languages. (Various notions of bounded entailment are shown to be expressible as language inclusions.) The languages unwind computations implicit in the logical (and temporal) connectives via a system of finite-state constraints adapted from finite-state morphology. Applications to Hybrid Logic and non-monotonic inertial reasoning are briefly considered.
This paper presents a system for the detection and correction of syntactic errors. It combines a robust morphosyntactic analyser and two groups of finite-state transducers specified using the Xerox Finite State Tool (xfst). One of the groups is used for the description of syntactic error patterns while the second one is used for the correction of the detected errors. The system has been tested on a corpus of real texts, containing both correct and incorrect sentences, with good results.
This paper describes the key aspects of the system SynCoP (Syntactic Constraint Parser) developed at the Berlin-Brandenburgische Akademie der Wissenschaften. The parser allows to combine syntactic tagging and chunking by means of constraint grammar using weighted finite state transducers (WFST). Chunks are interpreted as local dependency structures within syntactic tagging. The linguistic theories are formulated by criteria which are formalized by a semiring; these criteria allow structural preferences and gradual grammaticality. The parser is essentially a cascade of WFSTs. To find the most likely syntactic readings a best-path search is used.
We present an algorithm that computes a function that assigns consecutive integers to trees recognized by a deterministic, acyclic, finite-state, bottom-up tree automaton. Such function is called minimal perfect hashing. It can be used to identify trees recognized by the automaton. Its value may be seen as an index in some other data structures. We also present an algorithm for inverted hashing.
We introduce and discuss a number of issues that arise in the process of building a finite-state morphological analyzer for Urdu, in particular issues with potential ambiguity and non-concatenative morphology. Our approach allows for an underlyingly similar treatment of both Urdu and Hindi via a cascade of finite-state transducers that transliterates the very different scripts into a common ASCII transcription system. As this transliteration system is based on the XFST tools that the Urdu/Hindi common morphological analyzer is also implemented in, no compatibility problems arise.
Finite state methods for natural language processing often require the construction and the intersection of several automata. In this paper, we investigate the question of determining the best order in which these intersections should be performed. We take as an example lexical disambiguation in polarity grammars. We show that there is no efficient way to minimize the state complexity of these intersections.
Since Harris’ parser in the late 50s, multiword units have been progressively integrated in parsers. Nevertheless, in the most part, they are still restricted to compound words, that are more stable and less numerous. Actually, language is full of semi-fixed expressions that also form basic semantic units: semi-fixed adverbial expressions (e.g. time), collocations. Like compounds, the identification of these structures limits the combinatorial complexity induced by lexical ambiguity. In this paper, we detail an experiment that largely integrates these notions in a finite-state procedure of segmentation into super-chunks, preliminary to a parser.We show that the chunker, developped for French, reaches 92.9% precision and 98.7% recall. Moreover, multiword units realize 36.6% of the attachments within nominal and prepositional phrases.
This paper describes a two-level formalism where feature structures are used in contextual rules. Whereas usual two-level grammars describe rational sets over symbol pairs, this new formalism uses tree structured regular expressions. They allow an explicit and precise definition of the scope of feature structures. A given surface form may be described using several feature structures. Feature unification is expressed in contextual rules using variables, like in a unification grammar. Grammars are compiled in finite state multi-tape transducers.
This article describes a HMM-based word-alignment method that can selectively enforce a contiguity constraint. This method has a direct application in the extraction of a bilingual terminological lexicon from a parallel corpus, but can also be used as a preliminary step for the extraction of phrase pairs in a Phrase-Based Statistical Machine Translation system. Contiguous source words composing terms are aligned to contiguous target language words. The HMM is transformed into a Weighted Finite State Transducer (WFST) and contiguity constraints are enforced by specific multi-tape WFSTs. The proposed method is especially suited when basic linguistic resources (morphological analyzer, part-of-speech taggers and term extractors) are available for the source language only.
Nested complementation plays an important role in expressing counter- i.e. star-free and first-order definable languages and their hierarchies. In addition, methods that compile phonological rules into finite-state networks use double-nested complementation or “double negation”. This paper reviews how the double-nested complementation extends to a relatively new operation, generalized restriction (GR), coined by the author (Yli-Jyrä and Koskenniemi 2004). This operation encapsulates a double-nested complementation and elimination of a concatenation marker, diamond, whose finite occurrences align concatenations in the arguments of the operation. The paper demonstrates that the GR operation has an interesting potential in expressing regular languages, various kinds of grammars, bimorphisms and relations. This motivates a further study of optimized implementation of the operator.
The most recent trend in the studies of LF intervention effects makes crucial reference to focusing effects on the interveners, and this paper critically examines the representative analyses of the focus-based approach. While each analysis has its own merits and shortcomings, I argue that a pragmatic analysis that does not make appeal to syntactic configurations is better equipped to deal with many of the complex and delicate facts surrounding intervention effects.
When we pay close attention to the prosody of Wh-questions in Japanese, we discover many novel and interesting empirical puzzles that would require us to devise a much finer syntactic component of grammar. This paper addresses the issues that pose some problems to such an elaborated grammar, and offers solutions, making an appeal to the information structure and sentence processing involved in the interpretation of interrogative and focus constructions.
This paper discusses how focus changes prosodic structure in Tokyo Japanese. It is generally believed that focus blocks the intonational process of downstep and causes a pitch reset. This paper presents experimental evidence against this traditional view by looking at the prosodic behavior of Wh words, which receive focus lexically in Japanese as in other languages. It is demonstrated, specifically, that the focused Wh element does not block downstep although it receives a much higher pitch than its preceding element. This suggests that presence of lexical focus does not trigger pitch reset in Japanese.
The end of culture?
(2000)
Local Orders, Global Chaos
(1999)
The requirements of modern e-learning techniques change. Aspects such as community interaction, flexibility, pervasive learning and increasing mobility in communication habits become more important. To meet these challenges e-learning platforms must provide support on mobile learning. Most approaches try to adopt centralised and static e-learning mechanisms to mobile devices. However, often technically it is not possible for all kinds of devices to be connected to a central server. Therefore we introduce an application of a mobile e-learning network which operates totally decentralised with the help of an underlying ad hoc network architecture. Furthermore the concept of ad hoc messaging network (AMNET) is used as basis system architecture for our approach to implement a platform for pervasive mobile e-learning.
Monolayers of rod-shaped and disc-shaped liquid crystalline compounds at the air-water interface
(1986)
Calamitic (rod-shaped) and discotic (disc-shaped) thermotropic liquid crystalline (LC) compounds were spread at the air-water interface, and their ability to form monolayers was studied. The calamitic LCs investigated were found to form monolayers which behave analogously to conventional amphiphiles such as fatty acids. The spreading of the discotic LCs produced monolayers as well, but with a behaviour different from classical amphiphiles. The areas occupied per molecule are too small to allow the contact of all hydrophilic groups with the water surface and the packing of all hydrophobic chains. Various molecular arrangements of the discotics at the water surface to fit the spreading data are discussed.
Cinnamic acid moieties were incorporated into amphiphilic compounds containing one and two alkyl chains. These lipid-like compounds with photoreactive units undergo self-organization to form monolayers at the gas-water interface and bilayer structures (vesicles) in aqueous solutions. The photoreaction of the cinnamic acid moiety induced by 254 nm UV light was investigated in the crystalline state, in monolayers, in vesicles and in solution in organic solvents. The single-chain amphiphiles undergo dimerization to yield photoproducts with twice the molecular weight of the corresponding monomers in organized systems. The photoreaction of amphiphiles containing two cinnamic acid groups occurs via two mechanisms: The intramolecular dimerization produces bicycles, with retention of the molecular weight of the corresponding monomer. The intermolecular reaction leads to oligomeric and polymeric photoproducts. In contrast to the single-chain amphiphiles, photodimerization processes of lipoids containing two cinnamic acid moieties also occur in solution in organic solvents.
The spectral efficiency of blackness induction was measured in three normal trichromatic observers and in one deuteranomalous observer. The psychophysical task was to adjust the radiance of a monochromatic 60–120′ annulus until a 45′ central broadband field just turned black and its contour became indiscriminable from a dark surrounding gap that separated it from the annulus. The reciprocal of the radiance required to induce blackness with annulus wavelengths between 420 and 680 nm was used to define a spectral-efficiency function for the blackness component of the achromatic process. For each observer, the shape of this blackness-sensitivity function agreed with the spectral-efficiency function based on heterochromatic flicker photometry when measured with the same 60–120′ annulus. Both of these functions matched the Commission Internationale de l'Eclairage Vλ function except at short wavelengths. Ancillary measurements showed that the latter difference in sensitivity can be ascribed to nonuniformities of preretinal absorption, since the annular field excluded the central 60′ of the fovea. Thus our evidence indicates that, at least to a good first approximation, induced blackness is inversely related to the spectral-luminosity function. These findings are consistent with a model that separates the achromatic and the chromatic pathways.
The development of phonetic codes in memory of 141 pairs of normal and disabled readers from 7.8 to 16.8 years of age was tested with a task adapted from L. S. Mark, D. Shankweiler, I. Y. Liberman, and C. A. Fowler (Memory & Cognition, 1977, 5, 623–629) that measured false-positive errors in recognition memory for foil words which rhymed with words in the memory list versus foil words that did not rhyme. Our younger subjects replicated Mark et al., showing a larger difference between rhyming and nonrhyming false-positive errors for the normal readers. The older disabled readers' phonetic effect was comparable to that of the younger normal readers, suggesting a developmental lag in their use of phonetic coding in memory. Surprisingly, the normal readers' phonetic effect declined with age in the recognition task, but they maintained a significant advantage across age in the auditory WISC-R digit span recall test, and a test of phonological nonword decoding. The normals' decline with age in rhyming confusion may be due to an increase in the precision of their phonetic codes.
Linguistic and psycholinguistic accounts based on the study of English may prove unreliable as guides to sentence processing in even closely related languages. The present study illustrates this claim in a test of sentence interpretation by German-, Italian-, and English-speaking adults. Subjects were presented with simple transitive sentences in which contrasts of (1) word order, (2) agreement, (3) animacy, and (4) stress were systematically varied. For each sentence, subjects were asked to state which of the two nouns was the actor. The results indicated that Americans relied overwhelming on word order, using a first-noun strategy in NVN and a second-noun strategy in VNN and NNV sentences. Germans relied on both agreement and animacy. Italians showed extreme reliance on agreement cues. In both German and Italian, stress played a role in terms of complex interactions with word order and agreement. The findings were interpreted in terms of the “competition model” of Bates and MacWhinney (in H. Winitz (Ed.), Annals of the New York Academy of Sciences Conference on Native and Foreign Language Acquisition. New York: New York Academy of Sciences, 1982) in which cue validity is considered to be the primary determinant of cue strength. According to this model, cues are said to be high in validity when they are also high in applicability and reliability.
The optical density of human macular pigment was measured for 50 observers ranging in age from 10 to 90 years. The psychophysical method required adjusting the radiance of a 1°, monochromatic light (400–550 nm) to minimize flicker (15 Hz) when presented in counterphase with a 460 nm standard. This test stimulus was presented superimposed on a broad-band, short-wave background. Macular pigment density was determined by comparing sensitivity under these conditions for the fovea, where macular pigment is maximal, and 5° temporally. This difference spectrum, measured for 12 observers, matched Wyszecki and Stiles's standard density spectrum for macular pigment. To study variation in macular pigment density for a larger group of observers, measurements were made at only selected spectral points (460, 500 and 550 nm). The mean optical density at 460 nm for the complete sample of 50 subjects was 0.39. Substantial individual differences in density were found (ca. 0.10–0.80), but this variation was not systematically related to age.
In order to investigate the temporal characteristics of cognitive processing, we apply multivariate phase synchronization analysis to event-related potentials. The experimental design combines a semantic incongruity in a sentence context with a physical mismatch (color change). In the ERP average, these result in an N400 component and a P300-like positivity, respectively. The synchronization analysis shows an effect of global desynchronization in the theta band around 288ms after stimulus presentation for the semantic incongruity, while the physical mismatch elicits an increase of global synchronization in the alpha band around 204ms. Both of these effects clearly precede those in the ERP average. Moreover, the delay between synchronization effect and ERP component correlates with the complexity of the cognitive processes.
Fast analysis of different species of molecules in soils is investigated by capillary electrophoresis (CE). Several CE techniques for the analysis of inorganic ions and carbohydrates have been tested. With regard to the intents of pedologists and the usually large number of soil analyses a bundle of CE systems is proposed, capable of effecting time-saving soil analyses. Adapted electrolyte systems recently published and new separation systems are described. Examples of the application of these methods to two different soil samples are presented.
The haemolymph of the adult Colorado potato beetle, Lepinotarsa decemlineata Say, contains a high molecular weight (MW > 200,000) JH-III specific binding protein. The Kd value of the protein for racemic JH-III is 1.3 ± 0.2 × 10−7 M. It has a lower affinity for racemic JH-I and it does not bind JH-III-diol or JH-III-acid. The binding protein does discriminate between the enantiomers of synthetic, racemic JH-III as was determined by stereochemical anaysis of the bound and the free JH-III. Incubation of racemic JH-III with crude haemolymph results in preferential formation of (10S)-JH-III-acid, the unnatural configuration. The JH-esterase present in L. decemlineata haemolymph is not enantioselective. It is concluded that the most important function of the binding protein is that of a specific carrier, protecting the natural hormone against degradation by esterases. The carrier does not protect JH-I as efficiently as the lower homologue.
The site of confluence of the artery and the portal vein in the liver still appears to be controversial. Anatomical studies suggested a presinusoidal or an intrasinusoidal confluence in the first, second or even final third of the sinusoids. The objective of this investigation was to study the problem with functional biochemical techniques. Rat livers were perfused through the hepatic artery and simultaneously either in the orthograde direction from the portal vein to the hepatic vein or in the retrograde direction from the hepatic vein to the portal vein. Arterial how was linearly dependent on arterial pressure between 70 cm H2O and 120 cm H2O at a constant portal or hepatovenous pressure of 18 cm H2O. An arterial pressure of 100 cm H2O was required for the maintenance of a homogeneous orthograde perfusion of the whole parenchyma and of a physiologic ratio of arterial to portal how of about 1:3. Glucagon was infused either through the artery or the portal vein and hepatic vein, respectively, to a submaximally effective ''calculated'' sinusoidal concentration after mixing of 0.1 nmol/L. During orthograde perfusions, arterial and portal glucagon caused the same increases in glucose output. Yet during retrograde perfusions, hepatovenous glucagon elicited metabolic alterations equal to those in orthograde perfusions, whereas arterial glucagon effected changes strongly reduced to between 10% and 50%. Arterially infused trypan blue was distributed homogeneously in the parenchyma during orthograde perfusions, whereas it reached clearly smaller areas of parenchyma during retrograde perfusions. Finally, arterially applied acridine orange was taken up by all periportal hepatocytes in the proximal half of the acinus during orthograde perfusions but only by a much smaller portion of periportal cells in the proximal third of the acinus during retrograde perfusions. These findings suggest that in rat liver, the hepatic artery and the portal vein mix before and within the first third of the sinusoids, rather than in the middle or even last third.
Small livestock is an important resource for rural human populations in dry climates. How strongly will climate change affect the capacity of the rangeland? We used hierarchical modelling to scale quantitatively the growth of shrubs and annual plants, the main food of sheep and goats, to the landscape extent in the eastern Mediterranean region. Without grazing, productivity increased in a sigmoid way with mean annual precipitation. Grazing reduced productivity more strongly the drier the landscape. At a point just under the stocking capacity of the vegetation, productivity declined precipitously with more intense grazing due to a lack of seed production of annuals. We repeated simulations with precipitation patterns projected by two contrasting IPCC scenarios. Compared to results based on historic patterns, productivity and stocking capacity did not differ in most cases. Thus, grazing intensity remains the stronger impact on landscape productivity in this dry region even in the future.
A multitype Dawson-Watanabe process is conditioned, in subcritical and critical cases, on non-extinction in the remote future. On every finite time interval, its distribution is absolutely continuous with respect to the law of the unconditioned process. A martingale problem characterization is also given. Several results on the long time behavior of the conditioned mass process - the conditioned multitype Feller branching diffusion - are then proved. The general case is first considered, where the mutation matrix which models the interaction between the types, is irreducible. Several two-type models with decomposable mutation matrices are analyzed too .
Team diversity
(2007)
Team diversity refers to the differences between team members on any attribute that may lead each single member of the group to perceive any other member of the group as being different from the self of this particular member. These attributes and perceptions refer to all dimensions people can differ on, such as age, gender, ethnicity, religious and functional background, personality, skills, abilities, beliefs, and attitudes.
The authors used the frameworks of reciprocal determinism and occupational socialization to study the effects of work characteristics (consisting of control and complexity of work) on personal initiative (PI)--mediated by control orientation (a 2nd-order factor consisting of control aspiration, perceived opportunity for control, and self-efficacy) and the reciprocal effects of PI on changes in work characteristics. They applied structural equation modeling to a longitudinal study with 4 measurement waves (N = 268) in a transitional economy: East Germany. Results confirm the model plus 1 additional, nonhypothesized effect. Work characteristics had a synchronous effect on PI via control orientation (full mediation). There were also effects of control orientation and of PI on later changes in work characteristics: As predicted, PI functioned as partial mediator, changing work characteristics in the long term (reciprocal effect); unexpectedly, there was a 2nd reciprocal effect of an additional lagged partial mediation of control orientation on later work characteristics.
Observational evidence exists that winds of massive stars are clumped. Many massive star systems are known as non-thermal particle production sites, as indicated by their synchrotron emission in the radio band. As a consequence they are also considered as candidate sites for non-thermal high-energy photon production up to gamma-ray energies. The present work considers the effects of wind clumpiness expected on the emitting relativistic particle spectrum in colliding wind systems, built up from the pool of thermal wind particles through diffusive particle acceleration, and taking into account inverse Compton and synchrotron losses. In comparison to a homogeneous wind, a clumpy wind causes flux variations of the emitting particle spectrum when the clump enters the wind collision region. It is found that the spectral features associated with this variability moves temporally from low to high energy bands with the time shift between any two spectral bands being dependent on clump size, filling factor, and the energy-dependence of particle energy gains and losses.
The most massive stars are those with the shortest but most active life. One group of massive stars, the Luminous Blue Variables (LBVs), of which only a few objects are known, are in particular of interest concerning the stability of stars. They have a high mass loss rate and are close to being instable. This is even more likely as rotation becomes an important factor in stellar evolution of these stars. Through massive stellar winds and sometimes giant eruptions, LBV nebulae are formed. Various aspects in the evolution in the LBV phase lead, beside the large scale morphological and kinematical differences, to a diversity of small structures like clumps, rims, and outflows in these nebulae.
Gamma-rays can be produced by the interaction of a relativistic jet and the matter of the stellar wind in the subclass of massive X-ray binaries known as “microquasars”. The relativistic jet is ejected from the surroundings of the compact object and interacts with cold protons from the stellar wind, producing pions that then quickly decay into gamma-rays. Since the resulting gamma-ray emissivity depends on the target density, the detection of rapid variability in microquasars with GLAST and the new generation of Cherenkov imaging arrays could be used to probe the clumped structure of the stellar wind. In particular, we show here that the relative fluctuation in gamma rays may scale with the square root of the ratio of porosity length to binary separation, $\sqrt{h/a}$, implying for example a ca. 10% variation in gamma ray emission for a quite moderate porosity, h/a ∼ 0.01.
The optical spectrum of Eta Carinae (η Car) is prominent in H I, He i and Fe ii wind lines, all of which vary both in absorption and emission with phase. The phase dependance is a consequence of the interaction between the two objects in the η Car binary (η Car A & B). The binary system is enshrouded by ejecta from previous mass ejection events and consequently, η Car B is not directly observable. We have traced the He i lines over η Car’s spectroscopic period, using HST/STIS data obtained with medium spectral, but high angular, resolving power, and created a radial velocity curve for the system. The He I lines are formed in the core of the system, and appear to be a composite of multiple features formed in spatially separated regions. The sources of their irregular line profiles are still not fully understood, but can be attributed to emission/absorption near the wind-wind interface and/or a direct consequence of the η Car A’s, massive, clumpy wind. This paper will discuss the spectral variability, the narrow emission structure of the He i lines and how clumpiness of the winds may impede the construction of the reliable radial velocity curve, necessary for characterizations of especially η Car B.
The spatially-resolved winds of the massive binary, Eta Carinae, extend an arcsecond on the sky, well beyond the 10 to 20 milliarcsecond binary orbital dimension. Stellar wind line profiles, observed at very different angular resolutions of VLTI/AMBER, HST/STIS and VLT/UVES, provide spatial information on the extended wind interaction structure as it changes with orbital phase. These same wind lines, observable in the starlight scattered off the foreground lobe of the dusty Homunculus, provide time-variant line profiles viewed from significantly different angles. Comparisons of direct and scattered wind profiles observed in the same epoch and at different orbital phases provide insight on the extended wind structure and promise the potential for three-dimensional imaging of the outer wind structures. Massive, long-lasting clumps, including the nebularWeigelt blobs, originated during the two historical ejection events. Wind interactions with these clumps are quite noticeable in spatially-resolved spectroscopy. As the 2009.0 minimum approaches, analysis of existing spectra and 3-D modeling are providing bases for key observations to gain further understanding of this complex massive binary.
The H.E.S.S. collaboration recently reported the discovery of VHE γ-ray emission coincident with the young stellar cluster Westerlund 2. This system is known to host a population of hot, massive stars, and, most particularly, the WR binary WR 20a. Particle acceleration to TeV energies in Westerlund 2 can be accomplished in several alternative scenarios, therefore we only discuss energetic constraints based on the total available kinetic energy in the system, the actual mass loss rates of respective cluster members, and implied gamma-ray production from processes such as inverse Compton scattering or neutral pion decay. From the inferred gammaray luminosity of the order of 1035erg/s, implications for the efficiency of converting available kinetic energy into non-thermal radiation associated with stellar winds in the Westerlund 2 cluster are discussed under consideration of either the presence or absence of wind clumping.
We model the line profile variability (lpv) in spectra of clumped stellar atmospheres using the Stochastic Clump Model (SCM) of the winds of early-type stars. In this model the formation of dense inhomogeneities (clumps) in the line driven winds is considered as being a stochastic process. It is supposed that the emission due to clumps mainly contributes to the intensities of emission lines in the stellar spectra. It is shown that in the framework of the SCM it is possible to reproduce both the mean line profiles and a common pattern of the lpv.
We study the time variability of emission lines in three WNE stars : WR 2 (WN2), WR 3 (WN3ha) and WR152 (WN3). While WR 2 shows no variability above the noise level, the other stars do show variation, which are like other WR stars in WR 152 but very fast in WR 3. From these motions, we deduce a value of β ∼1 for WR 3 that is like that seen in O stars and β ∼2–3 for WR 152, that is intermediate between other WR stars and WR 3.
Luminous Blue Variables show strong changes in their stellar wind on time scales of typically years to decades when they expand and contract radially at approximately constant luminosity. Micro-variability on shorter time scales and amplitudes can be observed superimposed to the larger scale radial changes. I will show long-term time series of high resolution spectra which we have collected in the past 20 years for many of the well known LBVs together with a few time series of weekly sampling (HR Car, R40, R71, R110, R127, S Dor) covering a time windows of up to a few months. Wind variability is seen on short and intermediate time scales with the line profiles changing from P Cygni to inverse P Cygni and double peeked profiles sometimes for the same star and spectral line. On longer time scales the ionisation levels for all chemical elements change drastically due to the strong change of the temperature on the stellar surface. While on the long term the characteristic radial changes may have impact on the over all mass loss rates, the variabilities and asymmetries on short and intermediate time scales may cause false estimates of the mass loss rates when confronting models with the observed line profiles
Hα observations of Rigel obtained on 184 nights during the past ten years with the 1-m telescope and ´echelle spectrograph of Ritter Observatory are surveyed. The line profiles were classified in terms of morphology. About 1/4 of them are of P Cygni type, about 15% inverse P Cygni, about 25% double-peaked, about 1/3 pure absorption, and a few are single emission lines. Transformation of the profile from one type to another typically takes a few days. Although the line stays in absorption for extended intervals, only one high-velocity absorption event of the intensity reported by Kaufer et al. (1996a) was observed, in late 2006. Late in this event, Hα absorption occurred farther to the red than the red wing of a plausible photospheric absorption component, an indication of infalling material. In general, as the absorption events come to an end, the emission typically returns with an inverse P Cygni profile. The Hα profile class shows no obvious correlation with the radial velocity of C II λ6578, a photospheric absorption line.
By quantitatively fitting simple emission line profile models that include both atomic opacity and porosity to the Chandra X-ray spectrum of ζ Pup, we are able to explore the trade-offs between reduced mass-loss rates and wind porosity. We find that reducing the mass-loss rate of ζ Pup by roughly a factor of four, to 1.5 × 10−6 M⊙ yr−1, enables simple non-porous wind models to provide good fits to the data. If, on the other hand, we take the literature mass-loss rate of 6×10−6 M⊙ yr−1, then to produce X-ray line profiles that fit the data, extreme porosity lengths – of h∞ ≈ 3 R∗ – are required. Moreover, these porous models do not provide better fits to the data than the non-porous, low optical depth models. Additionally, such huge porosity lengths do not seem realistic in light of 2-D numerical simulations of the wind instability.
We review the effects of clumping on the profiles of resonance doublets. By allowing the ratio of the doublet oscillator strenghts to be a free parameter, we demonstrate that doublet profiles contain more information than is normally utilized. In clumped (or porous) winds, this ratio can lies between unity and the ratio of the f-values, and can change as a function of velocity and time, depending on the fraction of the stellar disk that is covered by material moving at a particular velocity at a given moment. Using these insights, we present the results of SEI modeling of a sample of B supergiants, ζ Pup and a time series for a star whose terminal velocity is low enough to make the components of its Si VIλλ1400 independent. These results are interpreted within the framewrok of the Oskinova et al. (2007) model, and demonstrate how the doublet profiles can be used to extract infromation about wind structure.
We present XMM-Newton Reflection Grating Spectrometer observations of pairs of X-ray emission line profiles from the O star ζ Pup that originate from the same He-like ion. The two profiles in each pair have different shapes and cannot both be consistently fit by models assuming the same wind parameters. We show that the differences in profile shape can be accounted for in a model including the effects of resonance scattering, which affects the resonance line in the pair but not the intercombination line. This implies that resonance scattering is also important in single resonance lines, where its effect is difficult to distinguish from a low effective continuum optical depth in the wind. Thus, resonance scattering may help reconcile X-ray line profile shapes with literature mass-loss rates.
We summarize Chandra observations of the emission line profiles from 17 OB stars. The lines tend to be broad and unshifted. The forbidden/intercombination line ratios arising from Helium-like ions provide radial distance information for the X-ray emission sources, while the H-like to He-like line ratios provide X-ray temperatures, and thus also source temperature versus radius distributions. OB stars usually show power law differential emission measure distributions versus temperature. In models of bow shocks, we find a power law differential emission measure, a wide range of ion stages, and the bow shock flow around the clumps provides transverse velocities comparable to HWHM values. We find that the bow shock results for the line profile properties, consistent with the observations of X-ray line emission for a broad range of OB star properties.
We present one-dimensional, time-dependent models of the clumps generated by the linedeshadowing instability. In order to follow the clumps out to distances of more than 1000 R∗, we use an efficient moving-box technique. We show that, within the approximations, the wind can remain clumped well into the formation region of the radio continuum.
INTEGRAL tripled the number of super-giant high-mass X-ray binaries (sgHMXB) known in the Galaxy by revealing absorbed and fast transient (SFXT) systems. Quantitative constraints on the wind clumping of massive stars can be obtained from the study of the hard X-ray variability of SFXT. A large fraction of the hard X-ray emission is emitted in the form of flares with a typical duration of 3 ksec, frequency of 7 days and luminosity of $10^{36}$ erg/s. Such flares are most probably emitted by the interaction of a compact object orbiting at $\sim10~R_*$ with wind clumps ($10^{22 ... 23}$ g) representing a large fraction of the stellar mass-loss rate. The density ratio between the clumps and the inter-clump medium is $10^{2 ... 4}$. The parameters of the clumps and of the inter-clump medium, derived from the SFXT flaring behavior, are in good agreement with macro-clumping scenario and line-driven instability simulations. SFXT are likely to have larger orbital radius than classical sgHMXB.
Magnetic fields influence the dynamics of hot-star winds and create large scale structure. Based on numerical magnetohydrodynamic (MHD) simulations, we model the wind of θ¹ Ori C, and then use the SEI method to compute synthetic line profiles for a range of viewing angles as function of rotational phase. The resulting dynamic spectrum for a moderately strong line shows a distinct modulation, but with a phase that seems at odds with available observations.
Discussion : X-rays
(2007)
Dynamical simulation of the “velocity-porosity” reduction in observed strength of stellar wind lines
(2007)
I use dynamical simulations of the line-driven instability to examine the potential role of the resulting flow structure in reducing the observed strength of wind absorption lines. Instead of the porosity length formalism used to model effects on continuum absorption, I suggest reductions in line strength can be better characterized in terms of a velocity clumping factor that is insensitive to spatial scales. Examples of dynamic spectra computed directly from instability simulations do exhibit a net reduction in absorption, but only at a modest 10-20% level that is well short of the ca. factor 10 required by recent analyses of PV lines.
The James Webb Space Telescope (JWST) is a large, infrared-optimized space telescope scheduled for launch in 2013. JWST will find the first stars and galaxies that formed in the early universe, connecting the Big Bang to our own Milky Way galaxy. JWST will peer through dusty clouds to see stars forming planetary systems, connecting the MilkyWay to our own Solar System. JWST’s instruments are designed to work primarily in the infrared range of 1 - 28 μm, with some capability in the visible range. JWST will have a large mirror, 6.5 m in diameter, and will be diffraction-limited at 2 μm (0.1 arcsec resolution). JWST will be placed in an L2 orbit about 1.5 million km from the Earth. The instruments will provide imaging, coronography, and multi-object and integral-field spectroscopy across the 1 - 28 μm wavelength range. The breakthrough capabilities of JWST will enable new studies of massive star winds from the Milky Way to the early universe.
General Discussion
(2007)
We study the influence of clumping on the predicted wind structure of O-type stars. For this purpose we artificially include clumping into our stationary wind models. When the clumps are assumed to be optically thin, the radiative line force increases compared to corresponding unclumped models, with a similar effect on either the mass-loss rate or the terminal velocity (depending on the onset of clumping). Optically thick clumps, alternatively, might be able to decrease the radiative force.
We present the results of Monte Carlo mass-loss predictions for massive stars covering a wide range of stellar parameters. We critically test our predictions against a range of observed massloss rates – in light of the recent discussions on wind clumping. We also present a model to compute the clumping-induced polarimetric variability of hot stars and we compare this with observations of Luminous Blue Variables, for which polarimetric variability is larger than for O and Wolf-Rayet stars. Luminous Blue Variables comprise an ideal testbed for studies of wind clumping and wind geometry, as well as for wind strength calculations, and we propose they may be direct supernova progenitors.
Many hot stars exhibit stochastic polarimetric variability, thought to arise from clumping low in the wind. Here we investigate the wind properties required to reproduce this variability using analytic models, with particular emphasis on Luminous Blue Variables. We find that the winds must be highly structured, consisting of a large number of optically-thin clumps; while we find that the overall level of polarization should scale with mass-loss rate – consistent with observations of LBVs. The models also predict variability on very short timescales, which is supported by the results of a recent polarimetric monitoring campaign.
Overwhelming observational and theoretical evidence suggests that the winds of massive stars are highly clumped. We briefly discuss the influence of clumping on model diagnostics and the difficulties of allowing for the influence of clumping on model spectra. Because of its simplicity, and because of computational ease, most spectroscopic analyses incorporate clumping using the volume filling factor. The biases introduced by this approach are uncertain. To investigate alternative clumping models, and to help determine the validity of parameters derived using the volume filling factor method, we discuss results derived using an alternative model in which we assume that the wind is composed of optically thick shells.
We report FUSE observations in 2005–2006 of three O-type, double-lined spectroscopic binaries in the Magellanic Clouds. The systems have very short periods (1.4–2.25 d), represent rare, young evolutionary stages of massive stars and binaries, and provide a unique glimpse at some of the most massive systems that form in dense clusters of massive stars. Improved orbit parameters, including revised masses, for LH54-425 are derived from new ctio spectroscopy. The systems are: LH54-425 in the LMC (O3V + O5V, P=2.25d, 62+37M⊙), J053441-693139 in the LMC (O2-3If+O6V, P=1.4 d, 41+27M⊙), and Hodge 53-47 in the SMC (O6V + O4-5IIIf, P=2.2 d, 24+14M⊙, where the O4 star appears to be less massive than the O6 star). Their short periods indicates that wind interaction and mass transfer are likely important factors in their evolution. The spectra provide quantitative and systematic studies of phase-dependent stellar wind properties, wind collision effects in O+O binaries at lower metallicities, improved radial velocity curves, and FUV spectro-photometric changes as a function of orbital phase.
We present preliminary results of a tailored atmosphere analysis of six Galactic WC stars using UV, optical, and mid-infrared Spitzer IRS data. With these data, we are able to sample regions from 10 to 10³ stellar radii, thus to determine wind clumping in different parts of the wind. Ultimately, derived wind parameters will be used to accuratelymeasure neon abundances, and to so test predicted nuclear-reaction rates.
Mass accretion onto compact objects through accretion disks is a common phenomenon in the universe. It is seen in all energy domains from active galactic nuclei through cataclysmic variables (CVs) to young stellar objects. Because CVs are fairly easy to observe, they provide an ideal opportunity to study accretion disks in great detail and thus help us to understand accretion also in other energy ranges. Mass accretion in these objects is often accompanied by mass outflow from the disks. This accretion disk wind, at least in CVs, is thought to be radiatively driven, similar to O star winds. WOMPAT, a 3-D Monte Carlo radiative transfer code for accretion disk winds of CVs is presented.
We apply the 3-dimensional radiative transport codeWind3D to 3D hydrodynamic models of Corotating Interaction Regions to fit the detailed variability of Discrete Absorption Components observed in Si iv UV resonance lines of HD 64760 (B0.5 Ib). We discuss important effects of the hydrodynamic input parameters on these large-scale equatorial wind structures that determine the detailed morphology of the DACs computed with 3D transfer. The best fit model reveals that the CIR in HD 64760 is produced by a source at the base of the wind that lags behind the stellar surface rotation. The non-corotating coherent wind structure is an extended density wave produced by a local increase of only 0.6% in the smooth symmetric wind mass-loss rate.
Clumping in Galactic WN stars : a comparison of mass loss rates from UV/optical & radio diagnostics
(2007)
The mass loss rates and other parameters for a large sample of Galactic WN stars have been revised by Hamann et al. (2006), using the most up-to date Potsdam Wolf-Rayet (PoWR) model atmospheres. For a sub-sample of these stars exist measurements of their radio free-free emission. After harmonizing the adopted distance and terminal wind velocities, we compare the mass loss rates obtained from the two diagnostics. The differences are discussed as a possible consequence of different clumping contrast in the line-forming and radio-emitting regions.
Recent studies of massive O-type stars present clear evidences of inhomogeneous and clumped winds. O-type (H-rich) central stars of planetary nebulae (CSPNs) are in some ways the low mass–low luminosity analogous of those massive stars. In this contribution, we present preliminary results of our on-going multi-wavelength (FUV, UV and optical) study of the winds of Galactic CSPNs. Particular emphasis will be given to the clumping factors derived by means of optical lines (Hα and Heii 4686) and “classic” FUV (and UV) lines.
We exploit time-series $FUSE$ spectroscopy to {\it uniquely} probe spatial structure and clumping in the fast wind of the central star of the H-rich planetary nebula NGC~6543 (HD~164963). Episodic and recurrent optical depth enhancements are discovered in the P{\sc v} absorption troughs, with some evidence for a $\sim$ 0.17-day modulation time-scale. The characteristics of these features are essentially identical to the discrete absorption components' (DACs) commonly seen in the UV lines of massive OB stars, suggesting the temporal structures seen in NGC~6543 likely have a physical origin that is similar to that operating in massive, luminous stars. The mechanism for forming coherent perturbations in the outflows is therefore apparently operating equally in the radiation-pressure-driven winds of widely differing momenta ($\mdot$$v_\infty$$R_\star^{0.5}$) and flow times, as represented by OB stars and CSPN.
This paper outlines a newly-developed method to include the effects of time variability in the radiative transfer code CMFGEN. It is shown that the flow timescale is often large compared to the variability timescale of LBVs. Thus, time-dependent effects significantly change the velocity law and density structure of the wind, affecting the derivation of the mass-loss rate, volume filling factor, wind terminal velocity, and luminosity. The results of this work are directly applicable to all active LBVs in the Galaxy and in the LMC, such as AG Car, HR Car, S Dor and R 127, and could result in a revision of stellar and wind parameters. The massloss rate evolution of AG Car during the last 20 years is presented, highlighting the need for time-dependent models to correctly interpret the evolution of LBVs.
We discuss the results of time-resolved spectroscopy of three presumably single Population I Wolf-Rayet stars in the Small Magellanic Cloud, where the ambient metallicity is $\sim 1/5 Z_\odot$. We were able to detect and follow numerous small-scale wind-embedded inhomogeneities in all observed stars. The general properties of the moving features, such as their velocity dispersions, emissivities and average accelerations, closely match the corresponding characteristics of small-scale inhomogeneities in the winds of Galactic Wolf-Rayet stars.
The influence of the wind to the total continuum of OB supergiants is discussed. For wind velocity distributions with β > 1.0, the wind can have strong influence to the total continuum emission, even at optical wavelengths. Comparing the continuum emission of clumped and unclumped winds, especially for stars with high β values, delivers flux differences of up to 30% with maximum in the near-IR. Continuum observations at these wavelengths are therefore an ideal tool to discriminate between clumped and unclumped winds of OB supergiants.
Massive stars usually form groups such as OB associations. Their fast stellar winds sweep up collectively the surrounding insterstellar medium (ISM) to generate superbubbles. Observations suggest that superbubble evolution on the surrounding ISM can be very irregular. Numerical simulations considering these conditions could help to understand the evolution of these superbubbles and to clarify the dynamics of these objects as well as the difference between observed X-ray luminosities and the predicted ones by the standard model (Weaver et al. 1977).
We present the latest results on the observational dependence of the mass-loss rate in stellar winds of O and early-B stars on the metal content of their atmospheres, and compare these with predictions. Absolute empirical rates for the mass loss of stars brighter than 10$^{5.2} L_{\odot}$, based on H$\alpha$ and ultraviolet (UV) wind lines, are found to be about a factor of two higher than predictions. If this difference is attributed to inhomogeneities in the wind this would imply that luminous O and early-B stars have clumping factors in their H$\alpha$ and UV line forming regime of about a factor of 3--5. The investigated stars cover a metallicity range $Z$ from 0.2 to 1 $Z_{\odot}$. We find a hint towards smaller clumping factors for lower $Z$. The derived clumping factors, however, presuppose that clumping does not impact the predictions of the mass-loss rate. We discuss this assumption and explain how we intend to investigate its validity in more detail.
We report on new mass-loss rate estimates for O stars in six massive binaries using the amplitude of orbital-phase dependent, linear-polarimetric variability caused by electron scattering off free electrons in the winds. Our estimated mass-loss rates for luminous O stars are independent of clumping. They suggest similar clumping corrections as for WR stars and do not support the recently proposed reduction in mass-loss rates of O stars by one or two orders of magnitude.
Clumping in O-star winds
(2007)
We have analyzed the spectra of seven Galactic O4 supergiants, with the NLTE wind code CMFGEN. For all stars, we have found that clumped wind models match well lines from different species spanning a wavelength range from FUV to optical, and remain consistent with Hα data. We have achieved an excellent match of the P V λλ1118, 1128 resonance doublet and N IV λ1718, as well as He II λ4686 suggesting that our physical description of clumping is adequate. We find very small volume filling factors and that clumping starts deep in the wind, near the sonic point. The most crucial consequence of our analysis is that the mass loss rates of O stars need to be revised downward significantly, by a factor of 3 and more compared to those obtained from smooth-wind models.
I discuss observational evidence – independent of the direct spectral diagnostics of stellar winds themselves – suggesting that mass-loss rates for O stars need to be revised downward by roughly a factor of three or more, in line with recent observed mass-loss rates for clumped winds. These independent constraints include the large observed mass-loss rates in LBV eruptions, the large masses of evolved massive stars like LBVs and WNH stars, WR stars in lower metallicity environments, observed rotation rates of massive stars at different metallicity, supernovae that seem to defy expectations of high mass-loss rates in stellar evolution, and other clues. I pay particular attention to the role of feedback that would result from higher mass-loss rates, driving the star to the Eddington limit too soon, and therefore making higher rates appear highly implausible. Some of these arguments by themselves may have more than one interpretation, but together they paint a consistent picture that steady line-driven winds of O-type stars have lower mass-loss rates and are significantly clumped.
The P v λλ1118, 1128 resonance doublet is an extraordinarily useful diagnostic of O-star winds, because it bypasses the traditional problems associated with determining mass-loss rates from UV resonance lines. We discuss critically the assumptions and uncertainties involved with using P v to diagnose mass-loss rates, and conclude that the large discrepancies between massloss rates determined from P v and the rates determined from “density squared” emission processes pose a significant challenge to the “standard model” of hot-star winds. The disparate measurements can be reconciled if the winds of O-type stars are strongly clumped on small spatial scales, which in turn implies that mass-loss rates based on Hα or radio emission are too large by up to an order of magnitude.
Significant seasonal variation in size at settlement has been observed in newly settled larvae of Dreissena polymorpha in Lake Constance. Diet quality, which varies temporally and spatially in freshwater habitats, has been suggested as a significant factor influencing life history and development of freshwater invertebrates. Accordingly, experiments were conducted with field-collected larvae to test the hypothesis that diet quality can determine planktonic larval growth rates, size at settlement and subsequent post-metamorphic growth rates. Larvae were fed one of two diets or starved. One diet was composed of cyanobacterial cells which are deficient in polyunsaturated fatty acids (PUFAs), and the other was a mixed diet rich in PUFAs. Freshly metamorphosed animals from the starvation treatment had a carbon content per individual 70% lower than that of larvae fed the mixed diet. This apparent exhaustion of larval internal reserves resulted in a 50% reduction of the postmetamorphic growth rates. Growth was also reduced in animals previously fed the cyanobacterial diet. Hence, low food quantity or low food quality during the larval stage of D. polymorpha lead to irreversible effects for postmetamorphic animals, and is related to inferior competitive abilities.
In the old days (pre ∼1990) hot stellar winds were assumed to be smooth, which made life fairly easy and bothered no one. Then after suspicious behaviour had been revealed, e.g. stochastic temporal variability in broadband polarimetry of single hot stars, it took the emerging CCD technology developed in the preceding decades (∼1970-80’s) to reveal that these winds were far from smooth. It was mainly high-S/N, time-dependent spectroscopy of strong optical recombination emission lines in WR, and also a few OB and other stars with strong hot winds, that indicated all hot stellar winds likely to be pervaded by thousands of multiscale (compressible supersonic turbulent?) structures, whose driver is probably some kind of radiative instability. Quantitative estimates of clumping-independent mass-loss rates came from various fronts, mainly dependent directly on density (e.g. electron-scattering wings of emission lines, UV spectroscopy of weak resonance lines, and binary-star properties including orbital-period changes, electron-scattering, and X-ray fluxes from colliding winds) rather than the more common, easier-to-obtain but clumping-dependent density-squared diagnostics (e.g. free-free emission in the IR/radio and recombination lines, of which the favourite has always been Hα). Many big questions still remain, such as: What do the clumps really look like? Do clumping properties change as one recedes from the mother star? Is clumping universal? Does the relative clumping correction depend on $\dot{M}$ itself?
Mass loss is a very important aspect of the life of massive stars. After briefly reviewing its importance, we discuss the impact of the recently proposed downward revision of mass loss rates due to clumping (difficulty to form Wolf-Rayet stars and production of critically rotating stars). Although a small reduction might be allowed, large reduction factors around ten are disfavoured. We then discuss the possibility of significant mass loss at very low metallicity due to stars reaching break-up velocities and especially due to the metal enrichment of the surface of the star via rotational and convective mixing. This significant mass loss may help the first very massive stars avoid the fate of pair-creation supernova, the chemical signature of which is not observed in extremely metal poor stars. The chemical composition of the very low metallicity winds is very similar to that of the most metal poor star known to date, HE1327-2326 and offer an interesting explanation for the origin of the metals in this star. We also discuss the importance of mass loss in the context of long and soft gamma-ray bursts and pair-creation supernovae. Finally, we would like to stress that mass loss in cooler parts of the HR-diagram (luminous blue variable and yellow and red supergiant stages) are much more uncertain than in the hot part. More work needs to be done in these areas to better constrain the evolution of the most massive stars.