I've written about the somewhat double-gaussian approach to fitting the BHT3 vernier scans already, so this is just a brief addendum:

I'm using two_gaus_cheat_readHist.C, called that because instead of assuming a double-gaussian beam profile, the code assumes taht the convolution of the beam profiles with each other produces a double gaussian output.  The results of the fit can be related to an approximate single-gaussian, and from there we can extract the beam parameters that we need.  I don't have a good strategy for propagating the errors in this approach.

The new addition, here, is correcting the resulting cross sections by the fraction of the barrel that was live during that run:

 run 10097097 has 4651 working towers.  The raw cross section was 342.08 and the corrected is 353.04nb.  The fit also shows a background rate of 1.7Hz

run 10103044 has 4621 working towers.  The raw cross section was 370.89 and the corrected is 385.26nb.  The fit shows a background rate of 0.75Hz.

To compare, the back-of-the-envelope calculation for run 10100015 is:

run 10100015 has fdet=.969=4651 working towers.

from http://online.star.bnl.gov/RTS/2009/triggerApp.php?screen=trgInfo&run=10100015 we see a trig rate of 116852, and a trig/baserate for bht3 of 1.50e-4, meaning a BHT3 trigger rate of 17Hz.  From our analysis of that run, we saw 228 SBB events before prescaling, out of 837 BHT3 events before prescaling.  This number agrees very well with the 41857 events listed by http://online.star.bnl.gov/RunLogRun9/Summary.php?run=10100015 . 

Removing these SBB events, we have to scale the rate by 1-228/837=.728, meaning the recorded rate of ~17Hz is actually only ~12.75Hz.

The instantaneous luminosity for that run is Rate(ZDC)/xsec(ZDC)=116852/2.3mb

The cross section for bht3 is Rate/instantaneous luminosity=12Hz*2.3mb/116852Hz=251nb,  ... we expect a bit better agreement.  However, if we don't correct for this (eg if background at a comparable level is still included in the ZDC rate number), then the  number is 345nb, which is very reasonable.