Subject: Re: clumping From: Rich Townsend Date: Tue, 26 Jun 2007 17:20:17 -0400 To: David Cohen Hi David -- Sorry I didn't get back to you quicker -- I've been working today to get my Cambridge review talk done. I would certainly be keen to publish a paper along these lines -- and I'm very grateful for the work you've already done setting up the framework and intro for the paper. (Not often I get a freebie like this). My major constraint at the moment is time --- fleshing out the paper shouldn't take too long, but I've got a lot of other competing tasks on my plate. With this issue in mind, I've just pulled out from the HELAS II conference on helioseismology and asteroseismology. This was due to take place in Gottingen, Germany, toward the end of August. Conny Aerts had invited me to give a review talk; I originally agreed, but over the past few weeks I became increasingly concerned about the cost (both in financial terms, and time-wise). Given that I'm already going to the pulsation conference in Cambridge, I figured my time and money was better spent elsewhere. So, my schedule is a little clearer now. I've just submitted the RFHD paper, and so I hope to be able to devote a significant fraction of my time to the porosity stuff (in parallel with some other short papers on magnetic field topics). But rather than going completely solo on this, I'd appreciate it if you would join me as a co-author. I think in addition to contributing to the science, your name on the paper would be a definite plus -- without the concommitant 'overshadowing' effect that might come, for instance, from co-authoring with Stan. And in any case, let's be frank -- it's your idea, and so even if I end up doing most of the calculations, it's only fair that you are an author. What do you think? cheers, Rich David Cohen wrote: > Rich, > > I mentioned to you in Potsdam that I thought it might be a good idea to write up some succinct notes on the different types of clumping and their effects on diagnostics, both as a way to explore physical parameters (e.g. how small can clumps get in a porous wind and still be optically thick, given reasonable assumptions about mass-loss rates and atomic opacities) and as a way of getting your clumping visualizations out there. And of course, to clarify in people's minds the differences between microclumping and macroclumping. > > In a sense, this paper would be in a similar spirit to Stan and Nathan's paper - not work that's entirely new, but synthesizing information that's already out there but which other people haven't fully appreciated. > > I've taken the liberty of writing an outline, and I wanted to get your opinion about whether this is a good idea before I share it with Stan and others. I see you as, perhaps, being the sole author of this, if you think it's a good idea. > > Looking forward to hearing your thoughts on this. > > David > > > > > --- > The Distinct Types of Wind Clumping in O Stars and Their Effects on Diagnostics and Mass-Loss Rate Determinations > > Intro: > > It has long been known that O star winds are inhomogeneous ("clumped"). New attention has been focused on this topic in recent years as quantitative modeling and data from different spectral regions show that density-squared emission diagnostics have overestimated the mass-loss rates of O stars by factors of at least a few. As researchers' attention has been turned to this problem, questions about the physical characteristics of the wind inhomogeneities have become more pressing - both for their impact on the interpretation of data and the consequences for mass-loss rate revisions, and also for understanding the physical mechanisms that give rise to the wind inhomogeneities. > > There are important distinctions to be drawn regarding at least three different categories of wind inhomogeneity: (1) Microclumping, in which clumps are small in scale and optically thin, that only affect density-squared diagnostics; (2) Line Macroclumping, in which clumps are large in scale and optically thick in resonance lines, that affects resonance absorption line diagnostics; and (3) Continuum Macroclumping (or "porosity"), in which clumps are optically thick in the continuum and the wind has a large inter-clump spacing (really a large porosity length), that affects continuum attenuation diagnostics like X-ray emission line profiles. Note that the second and third types of clumping will affect density-squared diagnostics as well, and that the second and third cases may sometimes both apply and sometimes not. Finally, we point out that line macroclumping can be influenced not just by spatial inhomogeneities in the wind but also by the velocity gradient in the wind. > > In this paper, we explore the several distinct types of wind clumping abd their effects on diagnostic interpretation and interactions with each other, We also investigate plausible constraints that they may allow us to put on clump densities, scale, and geometric and velocity-space distribution in the wind. Additionally, we produce visualizations of several different types of clump distributions, and show under what conditions different distributions can have the same effect on a given type of diagnostic. > > > Types of clumping: > > 1) Microclumping - just briefly go over the formalism, establishing the filling factor as the key parameter, pointing out that taking the interclump medium to be void isn't too bad an approximation (and that relaxing this assumption can be parameterized, and in any case, only serves to lessen the effect), and emphasizing the independence of the effects of microclumping on clump size scale. > > Keep in mind that x-ray continuum opacity can be effected by recombination rates, especially for He III -> He II, which are density-squared. > > 2) Line Macroclumping - elucidate the basic OFH2007 idea using more sensible formalism, explore the required density and size scales (but also velocity gradients) given realistic opacities/f-values - reference the upcoming vorocity paper. Mention Derck's doublet ratio thing too? > > *I was just thinking - when bigger clumps are considered (Line Macroclumping), emission lines like H-alpha (and free-free in principle, too) may also become optically thick, and how even density squared diagnostics are affected could become complicated. Maybe not so relevant for H-alpha, which is a recombination line; though if Lyman alpha is optically thick, then the population of the n=2 level could be affected... I don't know if this has been addressed in the H-alpha-Mdot literature. > > 3) Continuum Macroclumping - concisely restate the porosity-length formalism. Using a realistic X-ray opacity and assumptions about clump densities, see how small clumps can be and still plausibly be optically thick. > > Visualizations - something akin to the 3 X 3 movie, showing how structure with the same h is equivalent for case (3) and maybe for case (2) under certain circumstances; but that f is what matters for case (1). Also, explore (and visualize) different assumptions about radial clump distribution, lateral clump distribution (and when the one-parameter porosity length treatment for x-ray line profiles may break down), and the radial dependence on clump size. > > I assume Rich wants to publish some of MC etc. analysis separately, but maybe this could be a venue for discussing some of it (how clumps and porosity alter the mfp of photons, perhaps). > > > Discussion: > > Relate the observational constraints on f and h to actual clump properties. > > Summarize reasonable limits on clump properties under the assumptions that lines are optically thick and that the continuum is optically thick. > > Maybe discuss both lateral and radial clumping scales; and relate these to observational constraints from WR stars and Luc and Stan's patch model analysis in combination with both 1- and 2-D LDI simulations. > > Emphasize/recap theoretical "evidence" that clumping occurs on a small scale (e.g. remind people about Dessart and Owocki).