10 - H1 - Desy

4.2 Event reconstruction 53
4.2 Event reconstruction
Electronic signals collected directly during the data taking are subject to algorithms that
builds more physical objects. H1 track finding strategy (based on the algorithm described
in [106]) and calorimetry clustering is in details explained in [67]. For an overview of the
primary vertex reconstruction see [107]. On the following pages the scattered electron
identification and jet algorithm is presented.
4.2.1 Scattered electron identification
Two scattered electron finding algorithms are used in this analysis, with one looking for
electron in LAr calorimeter, the second in SPACAL.
The algorithm of scattered electron finding in the LAr calorimeter is based on the QESCAT
finder which main characteristics are described in [108]. The LAr clusters composed of at
least three cells, with energy higher than 5 GeV and transverse momentum higher than
3 GeV are selected. Additional cuts on cluster quality criteria are introduced such as
electromagnetic energy fraction, cluster compactness, radius and isolation. Finally, all
the scattered electron candidates are required to have an associated track with a distance
of closest approach below 12 cm.
In SPACAL all the clusters with electromagnetic energy above 5 GeV and radius below
4 cm are selected. The BPC and BST tracks are used to validate the electron candidate
position, but are not explicitely required.
From the selection of all LAr and SPACAL electron candidates the one with the highest
transverse momentum is selected as the scattered electron and used to veto DIS events
as explained in section 5.1.
4.2.2 Jet algorithm
Due to the confinement effect, partons from hard interaction can not be observed directly.
Instead, during the hadronisation process many colourless secondary particles are created
and leave their signal in the detector. Those final state particles can be grouped back into
jets containing the information about the initial partons produced directly in the collision.
A good jet algorithm should construct objects which well map the parton configuration
and be collinear and infrared safe. Infrared safety means that adding a soft particle into
the final state should not change the jet configuration and is a requirement needed to
minimise the influence of an electronic noise. Collinear safety means that two parallel
particles are exactly equivalent to one particle with the sum of the momenta of the pair.
This requirement accounts for resolution effects of the detector. In this analysis the jet
k T algorithm [109] is used which is both collinear and infrared safe. The algorithm works
as follows:
1. The initial object list is created from all the particle candidates being all the tracks