Or, why the best way to achieve immortality is to not die.
Maybe.
Hmm, last time I checked the uncertainty in the rate of binary supermassive black hole mergers was about 3 orders of magnitude, and had not converged in 4 decades of modeling...
- Log in to post comments
Well, Steinn, regarding the modeling uncertainties of supermassive BH mergers, we didn't write the linked article, and the actual paper DOES have a range of uncertainties, and even that is an uncertainty assuming the central hypotheses of merger-driven galaxy evolution and textbook dynamical friction are correct. I think a fair statement is ~1 order of magnitude uncertainty in our answer, maybe pushing 2 if you assume we're really ignorant, but you can make it infinite if you want to - just assume a last parsec problem. So, in summary, don't shoot the message writer in this case, shoot the messenger if you must, and preferably neither ;-)
Man, everyone is giving you a hard time about this, eh?
So, Thorne and Braginsky, 1976, as I recall, bracketed the rate estimate to within 3 orders of magnitude. When I reviewed the literature for some LISA meeting, the formal uncertainty had broadened by another order of magnitude.
Ignoring the last parsec, just the systemic errors in the semi-analytic model merger rates are easily 2-3 orders of magnitude, depending on what you define as "supermassive" and what mass ratios you cut on.
Key uncertainties include "initial masses" in proto-galaxies, threshold masses of halos at high z for there to be SMBH at all, and whether mergers drive SMBH growth or not.
Yeah, its been a rough couple of days :-). I think you'll find our paper interesting if you haven't read it, since our model is 100% insensitive to the 3 key uncertainties you've listed. We're talking about pulsar timing arrays, not LISA (nothing "imminent" about LISA, sadly), so we don't care about seed masses, merger trees, hierarchical buildup, etc. All we care about is how many red and dead very massive galaxies have merged since z=1. No merger trees, no DM halos... our uncertainties are 1) how negligible is star formation for growth of very massive guys since z=1 (we assume totally), 2) how well can we relate M_stars to M_BH for same, and 3) how well does Binney and Tremaine describe the BHs' trips down the potential well.
Well, Martin Weinberg and Chung Pei Ma separately took a good look at the Chandrasekhar formalism for dynamical friction and consistent with numerical studies find it roughly good to within a factor of 2.
But the new results on M_BH this week totally trash all correlations and all scenarios for merger driven BH growth.