Sometimes things come together in a strange way. First, I am having conversations with my office mate, Bret Lehmer, about work that he is involved in -- studying X-ray binary populations using X-ray observations. Second, I am writing a Chandra proposal aiming to study Xrays, due to star formation, in a new sample of supercompact UVLGs. Turns out that these two topics start to intersect. And because the whole topic of SFR, as measured by Xrays, is new to me, I thought I should make a few notes here. Though I expect I do not know much of the details, hopefully I can capture the broad stokes of what is known and what is still an open question in this topic. Here goes:
* The Xrays, related to star formation, arise from 1. diffuse emission from hot gas (>106 from O and B stars heating surrounding gas through winds and also by shock-heating as these massive stars supernovae (dominates Soft Xray band), 2. discrete emission from Xray binaries, as matter is accreted onto compact objects, such as neutron stars and black holes (dominates Hard X-ray band). Both types of emission scale with SFR.
* The second method of emission can be further categorized into two separate types: if the donor star in the binary system is a high mass star (High Mass Xray Binary, HMXB) or a low mass star (Low Mass Xray Binary, LMXB). Since the HMXB are short-lived systems, as massive stars are short-lived, the contribution of the HMXBs describe the recent star formation activity. The LMXBs, meanwhile, trace the past history of star formation, and therefore scale with the Mass of the galaxy. Of course, trying to separate out the contributions of the separate X-ray binary populations is tricky.
* The luminosity functions of the binaries within a specific galaxy can help separate HMXBs and LMXBs. The LF of LMXBs is relatively well characterized by a universal luminosity function and scales with galaxy stellar mass. By fitting models to these luminosity functions of resolve galaxies, we can hope to understand the complicated physics of the accretion properties of the X-ray binaries. These models depend on details such as metallicity, age of stellar populations in the galaxy, donor star mass, compact object mass, etc.
* There is an observed scaling relation between SFR and LX. However it is not clear that this relation holds to high redshift or for galaxies with different SFR/M* (sSFR), given the intrinsic scatter in the relation. While stacking results from Lyman Break Galaxies provide some constraints for both high-z and high sSFR galaxies, the stacked value is an average and also might include contamination from low-luminosity AGN. This is one way that the local LBG-analog population can offer another way to probe this sSFR population.
*The hard X-rays are necessary for studying the binary population, without significant contamination of hot gas -- and a high S/N observation is too expensive with Chandra for our sample of supercompact UVLGs. But... we are exploring some alternative ideas (Joint XMM proposal for Hard X-rays).
That's all for now. Perhaps I will add more as I write the proposal and research more about this topic.
Nice! A few questions:
ReplyDelete1) what exactly are the timescales for L and H MXRB creation? Has anyone looked at hardness vs sSFR (to explore for example if if hard x-rays coming from old HMXB and soft from young stars)?
2) I suppose you could get soft X-ray emission from a number of other processes than SF winds and supernovae (hot halos, turbulent shocks, etc..) but don't know if these are important
3)A key thing I think is to get enough precision to be able to distinguish SFR correlations from simple luminosity-luminosity correlations. That is, you want to see if L(X) vs L(Ha) is tighter than say L(X) vs L(total)
4) How good are the models of L and H MXB production as a function of time and SFR? What is the order of magnitude estimate for X-ray luminosity of SNe+massive star driver winds, per young star mass? Does this depend on the ISM properties? Will you be constraining X-ray binary production channels this way?