Table of Contents
This chapter describes specific elements, data and their attributes of the Whole Detector (ATLAS) (7.5 Whole Detector (ATLAS)) and its subsystems Inner Detector (InDet) (7.1 Inner Detector (InDet)), Calorimeters (Calo) (7.2 Calorimeters (Calo)) and Muon Detector (Muon) (7.3 Muon Detector (MuonDet)).
Section Analysis Object Data (AOD) (7.4 Analysis Object Data (AOD)) describes data and their attributes from AOD collections.
Section General Information (7.7 General Information) contains information about generally used attributes.
Elements, attributes and control are grouped into:
SpacePoint (7.1.1 Silicon 3D SpacePoints (SpacePoint)) : silicon space points
SpacePoint hits may be cleaned with the Silicon hit filter (7.1.13 Hit Filter Setup (Hit Filter)).
PixelRDO (7.1.2 Pixel Raw Data Object (PixelRDO)) : Pixel Raw Data Object
TrigSiSpacePoint (7.1.3 Online Silicon 3D Space Points (TrigSiSpacePoint)) : Online Silicon 3D Space Points
SiHit (7.1.4 Geant Silicon Hits (SiHit)) : Geant Silicon Hits
SiCluster (7.1.5 Silicon Cluster (SiCluster)) : Silicon Cluster
SiClusterRDO (7.1.6 Silicon Cluster Raw Data Object (SiClusterRDO)) : Silicon Cluster Raw Data Object
TRT_DriftCircle (7.1.7 Transition Radiation Tracker (TRT_DriftCircle)) : TRT straws
STr (7.1.8 Simulated Tracks (STr)) : simulated tracks
Segment (7.1.9 Reconstructed InDet Segment Collections (Segment)) : Reconstructed InDet Segment Collections
Track (7.1.10 Reconstructed Track Collections (Track)) : Reconstructed Track Collections
SimVertex (7.1.11 Simulated Vertices (SimVertex)) : simulated vertices
RecVertex (7.1.12 Reconstructed Vertices (RecVertex)) : reconstructed vertices
Hit Filter (7.1.13 Hit Filter Setup (Hit Filter)) : Hit Filter Setup
Space points from silicon pixel (7.1.1.1 Silicon Pixels) clusters and calculated from the low angle stereo silicon strips (7.1.1.2 Silicon Strips) are called S3D hits.
The following element and attribute items are available and can be controlled:
Color Function
The color of each hit is defined by the color function selected from the following options:
Constant (default)
in which case the color defined by item Constant is used for all hits
Simulated Track
the colors of unconnected and shared hits can be selected by items Unconnected and Shared
Reconstructed Track
the colors of unconnected and shared hits can be selected
SubDetector
(Barrel, Endcap+, Endcap-),
Group
as determined by the filter (7.1.13 Hit Filter Setup (Hit Filter)), the color of ungrouped hits can be selected
Layer.
Constant
The color of all hits, used for the Constant color function.
Unconnected
The color used for unconnected hits.
Shared
The color used for shared hits.
Ungrouped
Symbol Size
The size in pixels of the symbol used to represent hits.
Symbol
The type of symbol used to represent hits.
Frame
Frame Width
Noise Size
Noise
See Graphics Attributes of Hits (7.7.2 Graphics Attributes of Hits) for a description of graphics parameters related to hits.
Silicon pixels give directly 3D positions. From each cluster of pixels a space point is calculated. The pixel dimensions are 50µm×400µm. The barrel of the pixel detector has three layers at 4, 11 and 15 cm, respectively, which provide position measurements with a resolution of 12 µm in ρ-φ and 66 µm in Z. The endcap has four layers at ±35, ±67, ±95 and ±115 cm, respectively, which provide position measurements with a resolution of 12 µm in ρ-φ and 77 µm in ρ.
Silicon clusters are formed from groups of neighbouring silicon strips. In the barrel these silicon strips are either parallel to the beam axis (φ-strips) or inclined by either +0.04 radians = 2.3° (u-strips) or by -0.04 radians (v-strips). In the endcaps the φ-strips are perpendicular to the cylinder axis. Each layer contains φ-strips and either u-strips or v-strips (alternating between neighbouring layers) to allow for low angle stereo. If both such strips fire, when a charged particle traverses the layer, a space point can be calculated. If more than one particle traverse the same layer in close proximity ghost space points are produced.
For barrel layers the φ-ρ position of a space point is taken from the position of the φ strip. Z is calculated through low angle stereo. For endcap the φ-Z position is taken from the φ-strip and ρ is calculated through low angle stereo. Each barrel space point is measured with a precision of 16µm in ρ-φ and 580 µm in Z, respectively. Two tracks can be distinguished if separated by more than ≈200 µm.
The distance between φ strips and the u- or v- strips in the same module is 400 µm. The calculated value of Z for the barrel and ρ for the endcap depends of the direction of the track passing through the module. As this is normally unknown, most space point calculations assume particles of infinite momentum coming from X, Y, Z=0.
There are four layers in the barrel between 35, and 115 cm, and nine layers in the endcap between ±82 and ±276 cm.
Graphical Representation
The clusters are represented graphically as a line segment joining their two endpoints.
The color of each hit is defined by the color function which may be either
constant, in which case the constant color is used for all hits,
or it may vary with each hit being given the color of its associated
simulated track. The colors of unconnected and shared hits can be selected.
reconstructed track. The colors of unconnected and shared hits can be selected.
subdetector. (Barrel,Endcap+,Endcap-)
layer.
orientation. (φ-strips, u-strips or v-strips)
More information about colors in Atlantis is given in 7.7.1 About the Use of Colors.
The TRT barrel has straws parallel to the beam axis in the barrel and straws perpendicular to the beam axis in the endcap. The data from each TRT straw is best represented as a circle, however, in practice the radius of each straw (2mm) is so small compared to the length of a track segment (55cm) that when looking at such a track segment the circle can be adequately approximated as a straight line with a length given by twice the drift radius. The barrel straws are represented as lines perpendicular to a radial line in the φ/ρ (5.5 φ/ρ Projection) and Y/X (5.1 Y/X Projection) projections. The endcap straws are represented as lines perpendicular to the z-axis in the φ/Z (5.6 φ/Z Projection) projection. The drift radius is measured with a precision of around 170µm.
Transition radiation hits
Each straw provides an indication of whether a hit passes a second higher threshold associated with transition radiation. This transition radiation is associated with the passage of electrons and such hits may be used in electron identification. These hits can be viewed on their own by selecting the high threshold in the Cuts (6.4 Cuts).
On the Y/X projection (5.1 Y/X Projection) if the data contains drift signs and if zoomed in far enough the line showing the TRT data will change to an arrow to represent the drift.
Graphical Representation
The color of each hit is defined by the color function which may be either
constant, in which case the constant color is used for all hits,
or it may vary with each hit being given the color of its associated
simulated track, the colors of unconnected and shared hits can be selected,
reconstructed track, the colors of unconnected and shared hits can be selected,
subdetector (Endcap-,Barrel-,Barrel+,Endcap+).
More information about colors in Atlantis is given in 7.7.1 About the Use of Colors.
Simulated particles may be either charged STr or neutral SNP, and may be defined by:
the 3D position of the vertex, from which the particle originated,
the 3D momentum vector at the vertex position,
the particle type.
For charged particles, these parameters may be used to derive helix parameters (perigee parameters), extended by a further value, which gives the vertex position on the helix.
The paths travelled by simulated particles are defined in 3D and can therefore be seen in all projections. These paths do not reflect multiple scattering. Furthermore, for charged particles the helix approximation used is only valid inside the homogeneous solenoidal magnetic field and assuming negligible energy loss.
The color of each simulated particle is defined by the color function which may be either
constant, in which case the constant color is used for all particles,
or it may vary with each particle being given the color of its
index: in the range 0 to numSTr-1 or 0 to numSNP-1,
particle type:
cyan - charged hadrons,
yellow - electrons,
green - muons,
magenta - neutral hadrons,
red - photons,
orange - neutrinos.
The radius (radius Tr) and length (Z Tr) of the cylinder to which the particles are drawn can be changed.
Track images are easier recognised if tracks get a black frame with a small frame width, especially in the V-plot if calorimeters are displayed as well.
In the V-plot simulated neutral particles SNP are drawn as a symbol with a given symbol size. In the other projections simulated neutral particles are drawn as lines.
More information about colors in Atlantis is given in 7.7.1 About the Use of Colors.
Reconstructed tracks are described by helix parameters (perigee parameters). The reconstructed tracks are defined in 3D and can therefore be seen in all projections. The helix approximation used is only valid inside the homogeneous solenoidal magnetic field and assuming negligible energy loss and no multiple scattering.
The color of each reconstructed track is defined by the color function which may be either
constant, in which case the constant color is used for all tracks,
or it may vary with each track being given the color of its
index: in the range 0 to numRTr-1,
associated simulated track
associated E/Gamma reconstucted object
The radius (radius Tr) and length (Z Tr) of the cylinder to which the tracks are drawn can be changed.
Track images are easier recognised if tracks get a black frame with a small frame width, especially in the V-plot if calorimeters are displayed as well.
More information about colors in Atlantis is given in 7.7.1 About the Use of Colors.
Simulated vertices from the Monte Carlo truth. Currently only the start vertices of simulated particles are available and there should be at least one simulated particle originating from each simulated vertex.
The constant color of the simulated vertices may be changed.
More information about colors in Atlantis is given in 7.7.1 About the Use of Colors.
Reconstructed vertices may be formed interactively inside Atlantis from Lists (3.3.4 Lists Window) of reconstructed tracks. These reconstructed vertices are best represented by their error ellipsoids. These representations are currently available only in the Y/X (5.1 Y/X Projection), X'/Z (5.7 X'/Z Projection) and Y'/Z (5.8 Y'/Z Projection) projections. The error ellipses shown represent Num Sigma standard deviations which may be modified.
The constant color of these reconstructed vertices may be changed.
More information about colors in Atlantis is given in 7.7.1 About the Use of Colors.
In the case of normal simulated events and of the real events to be expected, it is difficult if not impossible to recognise tracks, etc., due to ghost hits, delta electrons, hits from the pile up and noise.
A filter is available to clean the events. In the case of fairly clean events, the V-plot may still help out. For the V-plot as for the filter, the Z-position of the primary vertex of the "triggering event" is required, so that η can be calculated.
Single filter
The filter is based on the simple requirement, that a 3D hit passes the filter if at least NH (NH=NumHits) hits on different layers lie in the same direction, as seen from the primary vertex.
In a φη histogram with cell size Δφ×Δη, the hits of a given helix with sufficiently high P t lie in one cell or in its direct neighbours due to rounding or measuring errors.
In a first loop over all hits the histogram is filled, however counting layers and not just hits. This is done using bit patterns. In a second loop over all hits each hit is accepted or rejected depending on the sum of different layers in its histogram bin and its neighbours.
The second loop can be combined with a clustering algorithm, so that hits falling into isolated clusters get a group number, which varies with different clusters. The subsets of S3D hits belonging to the different groups passing the filter may selected using the group parameter.
The strength and Pt acceptance of the filter is defined by Δη and Δφ:
The value of Δη=2×η Range/Numη reflects measuring η precision and should be as small as possible without loosing hits.
The value of Δφ=360°/Numφ defines the Pt range in which tracks are accepted and the strength of the filter. A small Δφ (strong filter) rejects much noise but also low Pt tracks.
In the same way one can filter tracks in a given Pt range through a φ'η histogram, where φ' is defined as φ'=φ-c×s×ρ, where c depends on the magnetic field. Hits of tracks with Pt=1/s have the same φ', and pass the filter. Depending on Δφ also hits of tracks with close Pt do pass. The sign of Pt is equal to the particle charge.
Filter loop
In order to have a strong filter with also low Pt tracks passing, the filter is executed N times in a loop where s is varied between -1/PtMIN and +1/PtMIN with N=1+2×NumSteps. Hits, which pass at least one of the single filters, are accepted. An accepted hit may have different group numbers from different filters. Different clusters, which contain the same hit, get the same new group number. Due to measuring errors, some tracks above PtMIN may get lost, so that PtMIN should be sufficiently low.