Photon Reconstruction

Aus HERMESwiki
Zur Navigation springen Zur Suche springen

Page maintainer: Eduard

Review.png Important note: This page has not yet been reviewed by Caro and Gunar! Use the information below with caution.

Other pages that have not been reviewed yet can be found in the category FORREVIEW.

Analysis quicklist

In the analyses using calorimeter the precise determination of the particle (normally photon) energy and position are crucial. Since the calorimeter wasn't originally designed for precision photon measurement, the above-mentioned observables benefit from additional corrections (compared to the original uDST values). As these corrections are quite complicated and non-universal (e.g. not applicable for all analyses) they are not yet implemented in the standard data production chain.
IMPORTANT: remember to discard photons with a preshower pulse >0.12(GeV) if you use any of the correction formulas as these values come from saturated ADCs and cannot be real!
The current best-knowledge is given below:

Single Photons (DVCS)

Energy correction using the preshower pulse for 0.001<PulsPre<0.12GeV:

  • Fehler beim Parsen (Konvertierungsfehler. Der Server („https://wikimedia.org/api/rest_“) hat berichtet: „Cannot get mml. Server problem.“): {\displaystyle E_{\gamma }=E_{\mathrm {calo} }\cdot (0.949861+0.255836\cdot E_{\mathrm {preshower} })+11.240620\cdot E_{\mathrm {preshower} }}


For non-converting photons (see Photons vs. Leptons) the correction formula has the form

  • Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle E_\gamma=0.9255\times E_{\mathrm{calo}}}


NOTE: all energies (calo and preshower) are given in GeV, as in uDST tables.

Multiple Photons (Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle \pi^0, \eta} ,...)

As secondary photons from meson decays usually have much lower energies compared to DVCS photons (and DIS leptons used for the correction formula extraction), they need a different correction formula for the energy. As for single photon case, different formulae are used for preshowering (0.001<PulsPre<0.12GeV) and non-preshowering photons:

  • Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle E_\gamma=E_{\mathrm{calo}}\cdot(0.9278+0.0005802\cdot E_{\mathrm{calo}}+0.1711\cdot E_{\mathrm{preshower}})+11.71\cdot E_{\mathrm{preshower}}}


and for PulsPre<0.001GeV the same formula as for DVCS can be used

  • Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle E_\gamma=0.9255\times E_{\mathrm{calo}}}

Position Reconstruction

The photon positions are measured in the calorimeter and saved in uDST smCluster table, in variables named rXleW and rYleW. These values are only useful if the corresponding z-position is known with a good precision. For all recent productions a value of

  • z=734cm

is recommended for preshowering photons and

  • z=747.5cm

for non-preshowering. NOTE: The previous value of z=729cm is valid for productions with old alignment, with incorrect position of the calo w.r.t. the target.

Monte-Carlo Analyses

Due to certain mismatch between data and MC one cannot use the same set of corrections for both cases. The corresponding formulas for MC are:
for :

  • Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle E_\gamma = E_{\mathrm{calo}} - 0.137 \cdot \sqrt{E_{\mathrm{calo}}}+(10.+0.3\cdot E_{\mathrm{calo}}) \cdot E_{\mathrm{preshower}} }

for DVCS:

  • Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle E_\gamma = E_{\mathrm{calo}}\cdot(0.9255+0.1841\cdot E_{\mathrm{preshower}})+11.57\cdot E_{\mathrm{preshower}}}

HRC-based analyses

For HRC-based analyses one may use the PID-library for selecting leptons, or, optionally, use hard cuts. The latter cut a lot of statistics as they don't take into account time-dependent variations in detector responses such as (mainly) TRD. Usually the following cuts are sufficient to select a reasonably clean lepton sample:

PulsTRD>20 && PulsPre>0.01 && 0.85<E/P<1.15

Intro

General Description

The arrangement of the Calorimeter walls
The dark side of the Calorimeter

The HERMES Electromagnetic Calorimeter is composed of 840 radiation resistant F101 lead-glass (LG) blocks arranged in a symmetric configuration with one wall above and one below the beam, and with photomultipliers (PMTs) viewing from the rear. Each wall is composed of 420 identical lead-glass blocks, stacked in a 42x10 array. Each block has an area 9x9 cm2 and a length of 50 cm (about 18 radiation lengths). This cell size meets the requirement that about 90% of the shower is contained in the cell for an axially-incident positron. The blocks were polished, wrapped with 50 mm thick aluminized mylar foil and covered with a 125 mm thick tedlar foil to provide light isolation.

Functions of the Calorimeter

The electromagnetic calorimeter is the main detector of the four detectors of the HERMES PID system.

Its functions are:

  • to provide a first-level trigger for scattered positrons, based on energy deposition in a localized spatial region;
  • to separate positrons from pions with a rejection factor of more than 10 at the first-level trigger and an additional factor of more than 100 in event reconstruction analysis;
  • to provide a measurement of the energy of DIS positrons;
  • to measure the energy of photons from radiative processes or from Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle \pi^0 } and Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle \eta } decays;
  • to give a coarse position measurement of scattered electrons and photons.

Performance of the Calorimeter

The performance of the calorimeter can be summarized as follows:

  • uniformity of the response of all counters within 1%;


Lincur.png
New calo delx y.png
  • linearity of the response to positrons within 1% over the energy range 1-30 GeV;


  • resolution

for a 3x3 array of counters and

    • Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle \frac{\sigma (E)}{E} = \frac{ 5.1 \pm 1.1 } { \sqrt{E}} + (2.0 \pm 0.5) + \frac{10.0 \pm 2.0}{E} }

(with E expressed in GeV) for the whole calorimeter operating in the spectrometer, including the effect of pre-showering of the positrons in a preshower detector before the calorimeter;


  • position reconstruction with resolution of about 0.7 cm;


Ped.png
Pi0eta ale.png
  • stability in time of the response within 1%;


  • no observed degradation of performance due to radiation damage,

within the accuracy of the measurements;


  • a hadron rejection factor exceeding 10 at the trigger level, and

a further off-line rejection factor of about 100;


  • reconstruction of Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle \pi^0} and Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle \eta} masses in agreement with the PDG values.

Calibration of the Calorimeter

Described in full details in the calibration internal note. Basic principles are the following:

  • the concept of total energy absorbsion in EM calorimeters is exploited, e.g. electrons' (positrons) measured energy should equal it's momentum (assuming 18+2 radiation lengths)
  • for each calo block a lepton sample is selected using hard cuts (see above)
  • the raw ADC values (after pedestal subtraction) for the cluster blocks are summed up, converted to energy units, and the previous calibration constant for the central block applied
  • the distribution of the energy calculated in the previous step divided by momentum (E/p) is constructed and fitted with a gaussian+polinomial
  • the reverse mean value of the gaussian is convolved with the initial value of the calibration constant and saved as a new value for the given block for a subsequent iteration
  • the whole procedure is repeated till converged - normally 2 iterations are sufficient, an extra iteration is done "just in case"

Essentials

Year Dependence

Specific issues related to certain years of data taking
Year Issue
96: Events with excessively large number of clusters (up to 100), obviously false. The reason hasn't been found, some clues indicate that the 50Hz noise could be responsible. The suggestion is to apply an upper cut on cluster multiplicity of 8 or 10 in the analysis.
97:
98: Some productions where by mistake the same setup for calibration was used as for previous (positron-beam) years. These are redone and obsolete.
99:
00: Signal cable swapped during data taking (see figure), treated by a special mapping file and repeated calibration.
02:
03:
04:
05: Some productions where by mistake the same setup for calibration was used as for previous (positron-beam) years. These are redone and obsolete.
06: Misplaced connector in splitter, resulted in data loss for one block closest to beam pipe. Treated with a special mapping and repeated calibration.
07: Common for 06+07 (recoil running): a significant gain drop for periods where the recoil magnet was on. Cured via separate calibration of periods of magnet-on and magnet-off runs. Similarly, a gradual gain drop with time was observed for both years. Partially cured by calibration in relatively small chunks (5000 runs).

In some periods of data taking, where due to signal cable or other reasons there was a missing block in the calorimeter, a special data quality bit (n.31) is introduced, to allow semi-inclusive analyses under the condition of having at most 1 dead block in calo. If the analysis depends heavily on photons, using this data is deprecated and the usual badbit 17 should be used instead.

Energy Measurement in the Calorimeter

Cluster search algorythm

In the calorimeter the energy of showering particles (electrons/positrons and photons) can be measured with a high precision. Due to the fact of the electromagnetic showering, these particles lose almost all of their energy in the 18 rad.-length depth of the lead glass, as opposed to pions which lose energy only due to hadronic showers. To further improve the efficiency of the showering, the calorimeter was preceded by a passive 11mm thick (~2 rad. lengths) lead preshower hodoscope. Since the shower development happens in transverse direction as well as in the longitudinal, the energy for each particle was measured by summing the energies deposited in a 3x3 array (usually called 'a cluster') of lead-glass blocks with a maximum of energy deposited in the central block. Such a construction ensures that a large fraction (>99% for central orthogonal impact) of the shower energy is contained in the cluster.

Preshower Energy Correction

Preshower energy correction

The passive lead radiator placed in front of the calorimeter serves as a preshower, so that an early development of the shower ensures it's full contamination in the lead glass later. As the TDR studies showed, such a construction improves the trigger efficiency significantly. As a side effect, though, the preshower also reduces the energy resolution as all the charged particles produced in the lead of the preshower also lose energy due to ionisation. The number of such (charged) particles may reach 50-60, and the total energy loss in the lead could be as much as 1GeV (depending on the shower development profile). This energy was being neglected when calculating the energy deposition in the preshower, causing a bias both during the calibration and analysis. First observed in 98, this effect was discussed in Arne Vandenbroucke's thesis, with a detailed study and parameterization of the energy loss in the preshower. One of the advantages of the study was that it used photons for the extraction of the formula. This became possible as the study used Monte-Carlo to compare the generated and reconstructed values of the photon energy. Unfortunately, the resulting formula didn't completely eliminate the energy and preshower-pulse dependence of the energy measured by the calorimeter in real data. An analogous study was performed later using leptons from real data (see talks). The resulting formula resulted in 10-30% resolution improvement for lepton energy, Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle \pi^0} mass and DVCS missing mass peaks. As a drawback, the peak positions for and DVCS missing mass got shifted from their expected positions, despite the resolution improvements. The currently recommended formula for lepton and DVCS photon energy correction is:

Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle E_\gamma=E_{\mathrm{calo}}\cdot(0.949861+0.255836\cdot E_{\mathrm{preshower}})+11.240620\cdot E_{\mathrm{preshower}}}

under the condition of having the preshower pulse 0.001<PulsPre<0.12GeV . For non-converting photons (see Photons vs. Leptons) the correction formula has the form

Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle E_\gamma=0.9255\times E_{\mathrm{calo}}}

Though the analytical form of the formula is chosen empyrically, there is still some logic behind. It is obvious that the (ionization) losses in the passive lead radiator depend on the number of (charged) particles created in it, which is proportional to the pulse registered in the scintillator, as each lepton created leaves a signal roughly equal to a MIP (~2MeV). Hence the last term of the formula should be interpreted as Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle C(GeV)\cdot N_{\mathrm{particles}}} , where Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle N_{\mathrm{particles}}=\frac{E_{\mathrm{preshower}}}{E_{\mathrm{MIP}}}} . The ionization losses depend highly on the longitudinal position of the particle creation, which depends on the energy of the initial particle - that's reflected in the middle term. Since the bias in the energy measurement affects the calibration process which relies on the energy being equal to momentum for leptons, the calibration procedure had to be repeated with the corrected energy. This became necessary as the average momentum distribution is different for each block, reaching ~18GeV in the central region of the calorimeter yet hardly 2-3GeV in the edges (see Momentum distribution). All recent productions use the correction formula for leptons, hence no extra correction is required if your analysis use lepton energy directly (usually not needed). The photon energy in uDSTs is untouched since for different analyses different corrections might be necessary.

Calibration Issues

Log-likelyhood (red) and Chisquare (green) fits of E/P

All productions made before May, 2002 (e.g. 00c and older), were calibrated using log-likelyhood method. As some studies showed this is not efficient, especially for blocks with large statistics and tails in the distributions. All subsequent productions have been calibrated using the chisquare fit method.

Productions made before August, 2009 use the "raw" calibration method (e.g. using the energy in calorimeter without correcting for losses in preshower), as opposed to subsequent ones:
98e, 99d, 00e, 02d, 03d, 04d, 05d, 06e, 07c. Currently there is no production available with preshower energy correction for 96, 97 data taking years.

Magnetic Field Influence

With the start of Recoil running in 2006 it turned out that despite the large distance (>7 meters) between the recoil solenoid and the calorimeter PMTs a significant gain change (mainly drop) is observed when the recoil magnet is switched on. During the maintenance shutdown in May-June 2006 a dedicated study has been performed with magnetic field measurements in the vicinity of the PMTs for the configurations with "Recoil OFF, Spectrometer OFF", "Recoil OFF, Spectrometer ON" and "Recoil ON, Spectrometer ON". The results had been reported during subsequent onsite meetings and can be briefly summarized as follows:

  • Despite the distance and the large amount of steel between the solenoid and the PMTs the magnetic field is sizeable, as the results of the measurements showed.
  • The magnetic field in the PMT area is non-homogeneous, and so is the resulting gain change: in very few spots this change is small and positive (1-2%), while in average it's negative (gain drop) of ~5-6%, sometimes reaching 10%.
  • As no reliable shielding was feasible for the last years of data taking, the proposed cure was to treat the magnet-on data separately during calibration. The resulting deviation between (Recoil) magnet-on and magnet-off data is marginal (link?), since the calibration process treats every block individually, and the inhomogenity of the magnetic field and the resulting gain drop is thus handled properly and eliminated.

Energy Deposition in the Calorimeter

Calorimeter energy vs momentum

Generally, the Calorimeter is a useful tool for particle identification. Light leptons (e+/e-) produce a shower, which is almost fully contained in the lead glass and detected by the PMT. By comparing the measured energy with a complementary momentum measurement for charged tracks one may select leptons as those with a ratio of energy/momentum close to unity (see the figure). Hadrons in HERMES energies practically never produce showers, except for hadronic showers, which only convert a small fraction of the initial hadron energy to light. In fact, many hadrons lose energy purely by ionisation, which only depends on the charge and material length, producing a minimum ionizing peak (MIP) at around 800 MeV/c.

Position Measurement

As follows from the definition of a calo cluster as a 3x3 group of blocks, the position of the particle that created the shower, or more precisely, the x/y coordinates of the shower maximum, can be acquired by composing a weighted sum of the energies deposited in each individual block. Such an algorithm has been proposed and used since the beginning of HERMES, as shown in publications [1] and [2]. That algorithm used Fehler beim Parsen (Konvertierungsfehler. Der Server („https://wikimedia.org/api/rest_“) hat berichtet: „Cannot get mml. Server problem.“): {\displaystyle {\sqrt {E}}} as energy weight and produces somewhat biased results, as later studies by J.Ely have shown. Alternatively, a Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle log E} weight with a treshold parameter has been proposed, which improved the position resolution significantly. Productions made after 2002 (e.g. all recent productions) use the new position reconstruction algorithm, with the old values kept in HRC/uDST tables (see uDST docu, new variables: rXlEw and rYlEw).

An important feature of the analysis that use these values (e.g. photon-related analysis like DVCS or pi0) is the importance of the Z-distance of the shower center from the vertex. Initial studies relying partly on MC have resulted in a value of 747.5cm for the average distance of the shower center from the target center (Z=0cm in HERMES coordinate system), corresponding to a distance of 472.5cm from the center of magnet (with z=275cm). These values were later revisited, apparently taking into account the passive preshower in front of the calo (which effectively moves the shower closer to the inner edge of the blocks), and the corresponding z-distance values were updated to 738cm (from target) and 463cm (from magnet center), correspondingly.

Since 2004, some studies initiated by DVCS analysis indicated that the latter values were not precise enough either. Three independent investigations, comparing the lepton track parameters determined by the tracking chambers and the calorimeter, concluded that a more precise value of the z-distance of the shower center from the target center is z=729cm for intermediate energies (~10GeV). This value was then recommended for the DVCS analysis, under the assumption that DVCS photons have to be converted in the preshower (e.g. produce a signal of PulsPre>0.001GeV), hence behaving as leptons in the calorimeter (see Photons vs. Leptons).

A separate issue that further complicated the exact determination of the mean z-distance of the shower in the calorimeter was the misalignment of the detector, observed since 2000 and later studied in detail (link to Sasha's presentation(s)). In short, the mylar-paper wrapping of the blocks was not accounted in the geometry file, hence the distance between the centers of adjacent blocks was wrong by ~1mm. These errors add up as one moves further away from the beam pipe, reaching a disposition of 2cm in the furthermost blocks. Additionally, the bottom detector had a shift of ~0.5cm in Y-direction. These misalignment issues had only a minor effect on the z-distance calculation as the studies using leptons have been performed with the actual geometry file, and hence all the dispositions were effectively diminished by the probe. Therefore, for all productions (done before 2009), using the old alignment, the value of z=729cm is still valid.

In 2009, after an improved alignment became available, including precise positions of calorimeter blocks in HERMES coordinate system, the studies of the exact determination of the mean shower center position were repeated. For all the recent productions (put list here) using the new alignment and new calibration algorythm, the old value of z=729cm is not valid anymore, due to changed position of the calorimeter in the geometry file. The new study is still in progress (put the link to the talk file), with an aim of providing a momentum dependent value for the mean z-position of the shower.

Photons vs. Leptons

In the relatively high energy domain the dominant interaction with matter is the Bremsstrahlung process for leptons and the pair production for photons (see the diagrams below).

lepton Lepton
photon Photon

In our setup (preshower+calorimeter) this effectively results in photonic showers to be in average 1 rad. length delayed w.r.t. the leptonic showers. The spatial development of the shower has an impact on both the energy and the reconstructed position of the detected particle. The energy measurement is affected through (at least) two different mechanisms with opposite effects: a late (deep) start of the shower may potentially result in some particles reaching the edge of the lead-glass block, that will reduce the measured energy. In the meantime the Cerenkov light produced closer to the PMTs is less attenuated by the residual opacity of the glass and produces larger signal. This concurent processes complicate the precise energy determination for photons. Initially, the photon energy was used in the analysis after applying a correction factor of 0.97 (obtained via MC studies, in agreement with the Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle \pi^0} peak position in data).
In practice, though, two distinct cases can be considered: photons that do convert in the preshower lead and those that don't. The converted photons can often be treated like leptons, e.g. their measured energy is directly used without further corrections. Photons that do not produce a pair in the preshower (up to 20% depending on the energy), tend to create a larger pulse in the PMT and hence need to have their measured energies corrected by a factor of 0.9255.
Similarly, the shower mean z-distance from the target center is different for the converted and non-converted photons. For the converted photons the same value for the z-distance as that of leptons can be used (e.g. z=729cm for the productions with old alignment). The shower development for non-converted photons happens much later (deeper in the lead glass) and hence has a wider spread with a mean at z=747.5cm.

Work in Progress

Several studies are still being finalized for reliable photon reconstructions. These are:

  • Energy (and angle?) dependent position reconstruction for preshowering and non-preshowering photons
  • Proper energy correction formula for the full energy range (e.g. universal for DVCS and Fehler beim Parsen (MathML mit SVG- oder PNG-Rückgriff (empfohlen für moderne Browser und Barrierefreiheitswerkzeuge): Ungültige Antwort („Math extension cannot connect to Restbase.“) von Server „https://wikimedia.org/api/rest_v1/“:): {\displaystyle \pi^0} photons)
    • Study of the slight inconsistency in E/P spectrum between DIS and regular lepton samples
  • Energy correction and position determination for Monte-Carlo

Issues

  • (Resolved) The photon spectrum shows a step function at 1 GeV (uncorrected smCluster value)
    • See comparison of lower energy photons from pi0 candidates in DIS for the given data productions (plot includes an energy correction factor of 0.97))
    • This issue has been resolved (as of 12.01.2011). It has been found that the ace package from the production chain applies a selective cut on clusters controlled vie flags "-wrsparse 1.0" and "-wrkeep 0.5", where the latter is effective for events with number of charged tracks >2. In the final productions of 06/07 these cuts will be set to 0.5GeV for all events.

More Info

Related Publications

Longitudinal electromagnetic shower profile in the HERMES calorimeter, Internal Note 07-007, by A.Vandenbroucke

Simulation of the HERMES Lead Glass Calorimeter Using a Look-Up Table Conference Proceedings 06-081 (talk file), by A. Vandenbroucke, C.A. Miller

Hermes Calorimeter Position Reconstruction Study Internal Report 01-056, by J. Ely

Performance of the Electromagnetic Calorimeter of the HERMES experiment Publication (not HERMES Collaboration) 98-067, by H. Avakian, N. Bianchi, G.P. Capitani, E. De Sanctis, P. Di Nezza, A. Fantoni, V. Giourdjian, R. Mozzetti, V. Muccifora, M. Nupieri, A.R. Reolon, P. Rossi, J.F.J. van den Brand, M. Doets, T. Henkes, M. Kolstein, A. Airapetian, N. Akopov, M. Amarian, R. Avakian, A. Avetisian, V. Garibian, S. Taroian Nucl. Instr. and Meth., A417 (1998) 69-78; hep-ex/9810004

Calibration, monitoring and simulation of the HERMES electro-magnetic calorimeter Internal Report 97-038, H. Avakian, P. Di Nezza, A. Fantoni

The HERMES Electromagnetic Calorimeter Conference Proceedings 96-037, by A. Fantoni, 6th International Conference on Calorimetry in High-Energy Physics (ICCHEP 96), Rome, Italy, Jun 8 - 14, 1996

Costruzione del calorimetro a contatori di vetro al piombo dell'esperimento HERMES per la misura delle funzioni di struttura di spin dei nucleoni by A. Fantoni, Ph.D. Thesis

Talks/Presentations

E.Avetisyan at Collab. Meetings Mar'08, Oct'07, Sep'07, Dec'04

A. Vandenbroucke at CALOR06, Dec'05, Oct'05

D.Zeiler at Mar'05.

F. Ellinghaus at collaboration meeting Oct. 2007 (?) about the z-position of the calorimeter for photons

Code Repository

Abgerufen von „https://hermes-wiki.desy.de/index.php?title=Photon_Reconstruction&oldid=12493“