Antares/KM3NeT group meeting - reconstruction

Europe/Amsterdam
Zoom

Zoom

Description

Join Zoom Meeting
https://nikhef.zoom.us/j/95893671004?pwd=aGI0cnllZnd3NnZ2RFBZZzg0Z3MyUT09

Meeting ID: 958 9367 1004
Passcode: 745504

 

Brían

On small PDF values in combined fit: these are a bit suspicious, no?

Bouke: Depends where you are evaluating.

Maarten: Could be just the background PDF (if background ~ 10 kHZ, then indeed PDF ~1e-6).

 

Valentin

On energy resolution for track reconstruction:
We have events where the neutrino energy is grossly overestimated.

Brían: Are these atmospheric muons or neutrinos?

Valentin: Neutrinos.

There are even horizontal events with reconstructed energies of 100 GeV for the ORCA-4 detector, which is suspsicious considering the small diameter of the detector.

Maarten: The energy estimation makes use of a five-point iterative search for the maximum likelihood energy. Maybe this five-point character, causes a division of the neutrino events into very high and low energies.
I'm tempted to dramatically increase the timewindow for the hit correlations. It could be that we just have extremely few hits. Currently, the time-window is ~+/- 25 ns.

Action point: Rerun energy estimation with greater time-window. Play with the input parameters and check whether the performance can be improved.

Bouke: Would it also help to increase the number of points in the energy search?

Maarten: Of course we can move the start point upwards or downwards in JEnergy. But again, this is just an input parameter. We should expand the time-window and then tune the input parameters.

 

There are minutes attached to this event. Show them.