Present: Thomas Junk, Marc Paterno, Gianluca Petrillo, Ruth Pordes, Brian Rebel, Saba Sehsish, Erica Snider
Remote: Jim Kowalkowski
The topic of the meeting was the support of LArSoft for ProtoDUNE needs.
A summary of the conclusions follows.
Given that the main purpose of ProtoDUNE is to inform the community in the choice between single and dual phase argon TPC (LArTPC2P), a comparison between the simulation of the two with shared physical events can provide valuable information.
If this comparison is performed on output of QSCAN on one side and LArSoft on the other (option A), the contribution to the differences from the different detector and from the different software need to be disentangled.
This may well be still the case even comparing LArSoft on both sides (option B, provided that LArSoft is able to support LArTPC2P), if the algorithms require adaption to extend the support to LArTPC2P -- and if no adaption is made, that could be argued to be a cause of the differences.
One leading way to perform this comparison fulfilling option (A) is to generate an event and perform a liquid-argon interaction simulation, and feed the two systems, LArSoft and QSCAN, with it. For example, the event might be generated with a particle gun (or GENIE if simulating neutrinos) and processed through GEANT4 to produce ionization electrons and scintillation photons, and those br fed into LArSoft and QSCAN for a comparison. For other comparisons, the shared part may be limited to the first step, the event generation.
Another level of comparison can be achieved sharing suitable high level reconstructed objects, like a set of tracks and showers. Given that both systems rely on similar abstractions, what is left is to determine which of these abstractions are suitable for exchange, and whether the content of the data objects can be mapped between the two.
About the direct support of LArTPC2P by LarSoft: some major areas have been identified where either work is known to be needed, or an assessment of the amount of such work is needed, for LArSoft to support a dual-phase argon TPC (LArTPC2P) at a level of any other LAr TPC detector:
-
the code supporting a geometry with wires parallel to z axis (that is used as reference in many places)
-
strategies of reconstruction algorithms, that have to cope with two 90 degree views
There are minutes attached to this event.
Show them.