DMNu Group Meeting Updates

This page collects my updates for the DMNu group meeting, indexed by date (scroll down -> older).

5th October 2016

Neck Model

• When ShellFit calculates its 'integral' of PE/photon (at a point on TPB) over TPB to get PE, it randomly samples from points over the TPB
• Says the area sampled is approx. = area it could possibly sample from / the number of samples.
  • e.g. 50 samples from a sphere, patch solid angle = (4*pi)/50.
  • Using these areas, its integral over TPB is a solid angle weighted mean over points
• It samples more points close to the PMT it's thinking about than far away, because PE per emitted photon from the TPB rises faster near the PMT (which makes sense, each PMT sees most light from TPB near the PMT!).
  • So there's small patches and big patches, and the cutoff = 2 * inter-PMT angle

• Just because a point doesn't fall in the neck doesn't mean part of the patch doesn't.
  • So removing points that fall in the neck we still integrate over too much TPB (this checks out if you compare omega_neck/(4*pi) to this method)
• If I:
  • + in neck loss light from a detailed solid angle calculation of the neck
  • - using a coarse calculation with too small a neck solid angle
  • -> I net overpredict number of PE near neck in top PMTs. To compensate the event is reconstructed away from the neck.
• Quick job to try out integrating over a proper neck area and subtracting the contribution.

Timing

-- NavinSeeburn - 05 Oct 2016