Jake Calcutt Sep. 12, 2019 Jake Calcutt 2 In addition to the total - - PowerPoint PPT Presentation

Jake Calcutt

• Sep. 12, 2019

Jake Calcutt 2

In addition to the total (elastic + inelastic) cross section, exclusive (subsets of inelastic) pion-Ar cross sections provide useful information Have the added benefit of unique signals Exclusive channels:

Jake Calcutt 3 π+ X π+ X π0

Absorption Charge Exchange

X = Nucleons

Jake Calcutt 4

Last talk (July 11) I showed some studies on building up to an event selection for this selection Came across some issues with reconstruction and how I was defining some things between reco and truth

• Was looking at events/tracks too broadly -- i.e. at the track level
• But: need to consider things at the slice/vertex level
• Refining this is necessary to determine selection performance

Jake Calcutt 5

Jake Calcutt 6

Right: Example of an event with an elastic scatter splitting a track Previously: Looked at the process ending the track (i.e. ‘pi+Inelastic’) and the true particles leaving that interaction

• But within the Thin Slice Method,

we only care about what happens at the end of the red track

• Becomes a question of “How do I

define the event/elastic vertex?”

Jake Calcutt 7

Just look at the list of processes from the true particle?

• Pandora will miss low-angle elastic scatters, which is fine for my

definition (to the point of a ‘small’ energy-loss systematic)

• Need to dig deeper into what’s happening along the true trajectory

and determine what’s “causing” pandora to end the track

Elastic Inelastic Reco Track

Jake Calcutt 8

Geant steps the particle through the Argon (MCTrajectory points)

• Some points have a process (‘hadElastic’, ‘pi+Inelastic’, etc.)

associated Between each step, a set of ionization deposits occur (IDE)

Jake Calcutt 9

Ionization drifts towards wires, creating signals on the wires

• (reconstructed as Hits)

Pandora collects Hits to form reconstructed Tracks/Showers Note: picture above is heavily simplified

Jake Calcutt 10

For a Thin-Slice analysis: split up the track/hits by wire-spacing

Jake Calcutt 11

Try to categorize the reco vertex by matching the Hits within the Slice to the IDEs that created them

Jake Calcutt 12

Once the hits from the vertex slice have been backtracked to their IDEs, try to determine their distance (ΔR) to any interaction points

• If any ΔR < (some value -- 5 cm for the following studies): consider

the reco vertex as matched to a true interaction

○ Refine this? -- i.e. ΔR of center of IDEs, average ΔR, etc...

• Use this to categorize what is happening at the reconstructed vertex

○ I.e. Elastic scatter vs. Inelastic scatter, something else? Collection of IDEs that created the hits True interaction points

Jake Calcutt 13

Vertex ‘matched’ to elastic scatter Inelastic scatter (can further classify this by daughters) ‘Unmatched’ (ΔR to any interaction is too large) Matched to both inelastic and elastic Some other process (i.e. decay)

Jake Calcutt 14

Of course, this is not the whole story and much more can happen

• I.e. the reco track extends past the inelastic vertex due to hits from

daughters For now: just want to test event selection using daughter information, deal with subtleties like this at another point

Jake Calcutt 15

Also realized that I had to refine how to include daughter track info Idea: put some conditions on the daughter track positions relative to the vertex position

• Should help with reco

issues such as cosmics tagged as daughers

Hits from the beam particle reconstructed as daughter Interaction point

Jake Calcutt 16

Looking at the distance between vertex point and start* of daughter track (dR -- note this is different than the vertex-defining ΔR) Legend notes:

• Nuc, Gamma, Pi, Proton, Mu are

true daughters

○ (i.e. pi → pi )

• GDaughter: Grand Daughter

○ (i.e. pi → pi → pi)

• GGDaughter

○ Great Grand Daughter ○ (i.e. pi → pi → pi → pi) * Start/end of daughter -- whichever is closer to the vertex Inelastic Vertex

Jake Calcutt 17

Looking at the distance between vertex point and start* of daughter track

* Start/end of daughter -- whichever is closer to the vertex Inelastic Vertex

Cosmics are far away

Jake Calcutt 18

Looking at the distance between vertex point and start* of daughter track

* Start/end of daughter -- whichever is closer to the vertex Inelastic Vertex

Pi0 (from primary interaction) decay gammas tagged as tracks

Jake Calcutt 19

Looking at the distance between vertex point and start* of daughter track

* Start/end of daughter -- whichever is closer to the vertex Inelastic Vertex

Mostly primary interaction daughters Some grand daughters from prompt reinteractions

• Mix of muons, protons,

pions, nuclear fragments, etc.

Jake Calcutt 20

Looking at the distance between vertex point and start* of daughter track

* Start/end of daughter -- whichever is closer to the vertex Elastic Vertex

Need to understand why true daughters are associated close to vertex for elastic scatters Could be inaccuracies due to how I’ve defined the vertex (i.e. matching IDEs within 5cm of interactions)

Jake Calcutt 21

‘Unmatched’ (no vertex-IDEs within 5cm of an interaction point) is a bit more complicated Track could end in the middle of the true beam particle (itself reco’d as daughter) Or maybe some place on a downstream particle (true daughters and grand daughters reco’d)

Unmatched

Jake Calcutt 22

Refining how I treat reconstruction and truth should help with defining an event selection

• More granular/in-depth info is always better
• Need to ‘tune’ the IDE/interaction matching -- suggestions welcome
• Also need to look into shower daughter information
• Quantify how the event selection performs

Thanks for listening

Jake Calcutt