simulation, reconstruction and testbeam preparations · simulation. simulation studies focused on...
TRANSCRIPT
Simulation, reconstruction andtestbeam preparationsG.Mavromanolakis, University of Cambridge
� � � ����� �
General
Simulation studies
Clustering algorithm
Testbeam preparation
041110 CALICE-UK meeting, London
General - Activities . Simulation
: survey of various shower models in GEANT4 and GEANT3,systematic studies/comparisons for CALICE proto geometries
. Reconstruction
: develop advanced pattern recognition algorithm for efficientcalorimeter clustering
. Testbeam preparation
: work on implementation of suitable data model
: design graphical user interface for monitoring/event display
Simulation
.simulation studies focused on CALICE ECAL and HCAL prototypes,to support and guide the testbeam program
.in general, studies reveal significant discrepancies amongshower packages, thus preventing model independentpredictions on calorimeter performance and reliable detectordesign optimization
.one of the main goals of the CALICE testbeam program isto resolve the situation and reduce the current largeuncertainty factors
.within this context, a proposal has been made to expose theECAL alone to hadronic beam for inclusive measurement of shower
model tag brief description
G3-GHEISHA : GHEISHA
G3-FLUKA+GH : FLUKA, for neutrons with � � 20 MeV GHEISHA
G3-FLUKA+MI : FLUKA, for neutrons with � � 20 MeV MICAP
G3-GH SLAC : GHEISHA with some bug fixes from SLAC
G3-GCALOR : � � 3 GeV Bertini cascade, 3 � � � 10 GeV hybrid Bertini/FLUKA, � 10 GeV FLUKA,for neutrons with � � 20 MeV MICAP
G4-LHEP : GHEISHA ported from GEANT3
G4-LHEP-BERT : � � 3 GeV Bertini cascade, � 3 GeV GHEISHA
G4-LHEP-BIC : � � 3 GeV Binary cascade, � 3 GeV GHEISHA
G4-LHEP-GN : GHEISHA + gamma nuclear processes
G4-LHEP-HP : as G4-LHEP, for neutrons with � � 20 MeV use evaluated cross-section data
G4-QGSP : � � 25 GeV GHEISHA, � 25 GeV quark-gluon string model
G4-QGSP-BERT : � � 3 GeV Bertini cascade, 3 � � � 25 GeV GHEISHA, � 25 GeV quark-gluon string model
G4-QGSP-BIC : � � 3 GeV Binary cascade, 3 � � � 25 GeV GHEISHA, � 25 GeV quark-gluon string model
G4-FTFP : � � 25 GeV GHEISHA, � 25 GeV quark-gluon string model with fragmentation ala FRITJOF
G4-QGSC : � � 25 GeV GHEISHA, � 25 GeV quark-gluon string model
G3: GEANT3.21 G4: GEANT4.6.1 with hadronic physics list PACK 2.5
ECAL+HCAL scint "response" vs model, � � 10 GeV
N cells hit E deposited
modelG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALORG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALOR
resp
on
se (
no
rmal
ised
)
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
ECAL 10 GeV-π
modelG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALORG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALOR
resp
on
se (
no
rmal
ised
)
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
ECAL 10 GeV-π
modelG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALORG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALOR
resp
on
se (
no
rmal
ised
)
0.6
0.8
1
1.2
1.4
1.6
1.8
2
HCAL (with ECAL in front) 10 GeV-π
scint
modelG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALORG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALOR
resp
on
se (
no
rmal
ised
)
0.6
0.8
1
1.2
1.4
1.6
1.8
2
HCAL (with ECAL in front) 10 GeV-π
scint
� different models predict different calorimeter response� HCAL more sensitive than ECAL� EM discrepancies between frameworks seen by ECAL
ECAL+HCAL scint "response" vs model, � � 1 GeV
N cells hit E deposited
modelG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALORG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALOR
resp
on
se (
no
rmal
ised
)
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
ECAL 1 GeV-π
modelG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALORG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALOR
resp
on
se (
no
rmal
ised
)
0.5
0.6
0.7
0.8
0.9
1
1.1
1.2
1.3
1.4
1.5
ECAL 1 GeV-π
modelG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALORG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALOR
resp
on
se (
no
rmal
ised
)
0.5
1
1.5
2
2.5
3
HCAL (with ECAL in front) 1 GeV-π
scint
modelG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALORG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALOR
resp
on
se (
no
rmal
ised
)
0.5
1
1.5
2
2.5
3
HCAL (with ECAL in front) 1 GeV-π
scint
� same pattern as at 10 GeV case, even more pronounced� ECAL standalone may have some discriminating power
HCAL rpc – HCAL scint
N cells hit shower width
modelG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALORG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALOR
res
po
ns
e (
no
rma
lis
ed
)
0.6
0.8
1
1.2
1.4
1.6
1.8
2
HCAL (with ECAL in front) 10 GeV-π
scintrpc
modelG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALORG4-FTFP
G4-LHEP
G4-LHEP-BERT
G4-LHEP-BIC
G4-LHEP-GN
G4-LHEP-HPG4-QGSC
G4-QGSP
G4-QGSP-BERT
G4-QGSP-BIC
G3-GHEISHA
G3-FLUKA+GH
G3-FLUKA+MI
G3-GH SLAC
G3-GCALOR
sh
ow
er
rad
ius
(n
orm
ali
se
d)
0.6
0.8
1
1.2
1.4
1.6
1.8
2
HCAL (with ECAL in front) 10 GeV-π
scintrpc
� HCAL rpc less sensitive to low energy neutrons than HCAL scint
discrepancies between frameworks at em level
N cells hit E deposited
# of ECAL cells0 10 20 30 40 50 60 70 80 90 100
cou
nts
0
0.02
0.04
0.06
0.08
0.1
Number of ECAL cells hit per event
GEANT4
GEANT3
FLUGG
1 GeV-e
BinWidth: 1e+00Number of ECAL cells hit per event
E deposited in ECAL (MeV)0 10 20 30 40 50 60 70 80 90 100
even
ts
0
0.02
0.04
0.06
0.08
0.1
0.12
ECAL E deposited per event
GEANT4
GEANT3
FLUGG
1 GeV-e
BinWidth: 1e+00ECAL E deposited per event
GEANT3 14% higher than GEANT4 GEANT3 14% higher than GEANT4
FLUGG 24% higher than GEANT4 FLUGG 30% higher than GEANT4
0
2.5
5
7.5
10
12.5
15
17.5
20
0 2.5 5 7.5 10 12.5 15 17.5 20
π−
ECAL+HCAL rpcECAL+HCAL scintECAL
E incident (GeV)
E r
econ
str
(GeV
)
� combine ECAL+HCAL to study hadronic shower development� is it worth exposing ECAL alone to lower energy hadrons?
shower longitudinal profile by W block + ECALW block ��� � ��� � ����� ��� � W block ��� � ��� � ����� � ��� �
ECAL layer0 5 10 15 20 25 30
N c
ells
hit
(a.
u.)
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
G3-FLUKA GH
G3-FLUKA MI
G3-GCALOR
G3-GHEISHA
ECAL 5 GeV-π
BinWidth: 1e-00
ECAL layer0 5 10 15 20 25 30
N c
ells
hit
(a.
u.)
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
G3-FLUKA GH
G3-FLUKA MI
G3-GCALOR
G3-GHEISHA
) + ECALIλW(0.9 5 GeV-π
BinWidth: 1e-00
ECAL layer0 5 10 15 20 25 30
N c
ells
hit
(a.
u.)
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
G3-FLUKA GH
G3-FLUKA MI
G3-GCALOR
G3-GHEISHA
) + ECALIλW(1.8 5 GeV-π
BinWidth: 1e-00
� differences among models start to show up at the end of ECAL� use W block to have similar shower development as in ECAL! models clearly distinguishable with block of W in front of ECAL! combine "photos" to reveal inclusive longitudinal profile
shower transverse profile by W block + ECALW block ��� � ��� � ����� ��� � W block ��� � ��� � ����� � ��� �
x (mm)-100 -80 -60 -40 -20 0 20 40 60 80 100
N c
ells
hit
(a.
u.)
0
10000
20000
30000
40000
50000
60000G3-FLUKA GH
G3-FLUKA MI
G3-GCALOR
G3-GHEISHA
ECAL 5 GeV-π
BinWidth: 1e+01
x (mm)-100 -80 -60 -40 -20 0 20 40 60 80 100
N c
ells
hit
(a.
u.)
0
10000
20000
30000
40000
50000
60000G3-FLUKA GH
G3-FLUKA MI
G3-GCALOR
G3-GHEISHA
) + ECALIλW(0.9 5 GeV-π
BinWidth: 1e+01
x (mm)-100 -80 -60 -40 -20 0 20 40 60 80 100
N c
ells
hit
(a.
u.)
0
10000
20000
30000
40000
50000
60000G3-FLUKA GH
G3-FLUKA MI
G3-GCALOR
G3-GHEISHA
) + ECALIλW(1.8 5 GeV-π
BinWidth: 1e+01
" front face # $ % # $ &�' ( segmented into # % # &�' ( cells,sufficient to record shower’s core
� differences more pronounced with a block of matter in front of ECAL
Calorimeter clustering . particle flow paradigm
: highly granular em and hadr calorimeters to allow very efficientpattern recognition for excellent shower separation and pidwithin jets to provide excellent jet reconstruction efficiency
.: software development (calorimeter clustering, track finding,
track-shower matching) to play a very crucial role
. current activity
: algorithm based on MST theory has been developed toexploit a "top-down and then bottom-up" approach tocalorimeter clustering
Introduction – theory minimal spanning tree
: a tree which contains all nodes with no circuits andof which the sum of weights of its edges is minimum
properties
: unique for the given set of nodes and the chosen metric
: deterministic, no dependence on random choices of nodes
: invariant under similarity transformations that preserve themonotony of the metric
0 10 20 30 40 50 60 70 80 90 1000
10
20
30
40
50
60
70
80
90
100
nodes, path & circuit
0 10 20 30 40 50 60 70 80 90 1000
10
20
30
40
50
60
70
80
90
100
spanning tree
0 10 20 30 40 50 60 70 80 90 1000
10
20
30
40
50
60
70
80
90
100
minimal spanning tree
0 10 20 30 40 50 60 70 80 90 1000
10
20
30
40
50
60
70
80
90
100
clustering
clustering
MST Clustering – Main steps . metric
define a metric : metric defines configuration space
: not necessarily euclidean
fill distance matrix : lower triangular ) * matrix
. MST
construct the MST : apply Prim’s algorithm
. clustering
set cut : "proximity" bound between nodes belongingto the same cluster
find clusters : single linkage cluster analysis
: i.e. go through MST and cut brancheswith length above cut
nodes minimal spanning tree clustering
x (mm)
-500-400
-300-200
-1000
100200
300400
500
y (mm)
0
200
400
600
800
1000
1200
1400
z (mm
)
-500
-400
-300
-200
-100
0
100
200
300
400
500
x (mm)
-500-400
-300-200
-1000
100200
300400
500
y (mm)
0
200
400
600
800
1000
1200
1400
z (mm
)
-500
-400
-300
-200
-100
0
100
200
300
400
500
x (mm)
-500-400
-300-200
-1000
100200
300400
500
y (mm)
0
200
400
600
800
1000
1200
1400
z (mm
)
-500
-400
-300
-200
-100
0
100
200
300
400
500
+ , - , . ,
x (mm)
-500-400
-300-200
-1000
100200
300400
500
y (mm)
0
200
400
600
800
1000
1200
1400
z (mm
)
-500
-400
-300
-200
-100
0
100
200
300
400
500
x (mm)
-500-400
-300-200
-1000
100200
300400
500
y (mm)
0
200
400
600
800
1000
1200
1400
z (mm
)
-500
-400
-300
-200
-100
0
100
200
300
400
500
x (mm)
-500-400
-300-200
-1000
100200
300400
500
y (mm)
0
200
400
600
800
1000
1200
1400
z (mm
)
-500
-400
-300
-200
-100
0
100
200
300
400
500
x (mm)
-500-400
-300-200
-1000
100200
300400
500
y (mm)
0
200
400
600
800
1000
1200
1400
z (mm
)
-500
-400
-300
-200
-100
0
100
200
300
400
500
x (mm)
-500-400
-300-200
-1000
100200
300400
500
y (mm)
0
200
400
600
800
1000
1200
1400
z (mm
)
-500
-400
-300
-200
-100
0
100
200
300
400
500
x (mm)
-500-400
-300-200
-1000
100200
300400
500
y (mm)
0
200
400
600
800
1000
1200
1400
z (mm
)
-500
-400
-300
-200
-100
0
100
200
300
400
500
Top-down and then bottom-up clustering in brief
: use MST clustering algorithm with loose cut to performcoarse clustering
: go through MST clusters found in previous step and refineusing a cone like clustering algorithm
advantages of top-down and then bottom-up approach
: speed because of preclustering, vital for a verygranular calorimeter even if its occupancy is low
: geometry independence (or at least no strict bindings)
: efficiency (hopefully)
x (mm)
-2400 -2200 -2000 -1800 -1600 -1400 -1200 -1000 -800 -600
y (m
m)
-2400
-2200
-2000
-1800
-1600
-1400
-1200
-1000
-800
-600
x (mm)
-2400 -2200 -2000 -1800 -1600 -1400 -1200 -1000 -800 -600
y (m
m)
-2400
-2200
-2000
-1800
-1600
-1400
-1200
-1000
-800
-600
x (mm)
-2400 -2200 -2000 -1800 -1600 -1400 -1200 -1000 -800 -600
y (m
m)
-2400
-2200
-2000
-1800
-1600
-1400
-1200
-1000
-800
-600
true clusters after mst clustering final reconstructed clusters
x (mm)
400 500 600 700 800 900 1000 1100 1200 1300 1400
y (m
m)
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
x (mm)
400 500 600 700 800 900 1000 1100 1200 1300 1400
y (m
m)
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
x (mm)
400 500 600 700 800 900 1000 1100 1200 1300 1400
y (m
m)
1000
1100
1200
1300
1400
1500
1600
1700
1800
1900
2000
true clusters after mst clustering final reconstructed clusters
x (mm)
-200 0 200 400 600 800 1000 1200 1400 1600
y (m
m)
1200
1400
1600
1800
2000
2200
2400
2600
x (mm)
-200 0 200 400 600 800 1000 1200 1400 1600
y (m
m)
1200
1400
1600
1800
2000
2200
2400
2600
x (mm)
-200 0 200 400 600 800 1000 1200 1400 1600
y (m
m)
1200
1400
1600
1800
2000
2200
2400
2600
true clusters after mst clustering final reconstructed clusters
Performance evaluation .
: / 0 pairs at 10 GeV or 5 GeV on CALICE ECAL+HCAL RPC prototypes
: ECAL, HCAL cellsize 1 1 1 cm * , cell threshold = 0.5 mip
: satisfactory performance given the fact that the algorithm is seedlessand both ECAL and HCAL hits are treated as digital in clustering
distance (cm)0 5 10 15 20 25
sepa
ratio
n ef
ficie
ncy
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
ECAL+HCAL rpc
5 GeV-π 10 GeV-π
separation quality
: use V.Morgunov, A.Raspereza concept
: shoot pairs of particles with fixed energy and at given distancefrom each other, how well the algorithm reconstructs the showers
: "separation quality = fraction of events in which reconstructedenergy of the shower lies in the range 2436587:9<; ��= , where = isthe nominal energy resolution of the shower without aclose by shower"
Testbeam data model
.proposing a data model for the CALICE testbeam program> persistency> flexible implementation> simple user interface> efficiency
.use LCIO and ROOT frameworks for some simple testimplementation and benchmarking
.general conversion scheme (from raw/simulation datato analysis data) has been discussed/decided
LCIO event data entities (v01-03)
CALICE testbeam data model
LCObject
LCGenericObject
TBEcalHit TBTrackerHit TBPIDData
TBEcalAux TBTrackerAux TBPIDAux
LCIOUserDefined
alternative ...
TObject
TTree
treeTBEcalHits
treeTBTrackerHits
treeTBPIDData
TBTrackerAux
TBPIDAux
ROOT
UserDefined
TBEcalAux
instances of TTree
Benchmarks . configuration
> machine: Linux P4 2.66 GHz / 512 MB RAM> libs: ROOT v4.00/08 and LCIO v01-03> task: write/read 1 ROOT tree or 1 LCIO collection
of N events ? 100 hits (1 hit = 3 integers + 3 floats)
L C I O R O O T
100k events size (MB) 28 4time write (sec) 64 9time read (sec) 71 19
500k events size (MB) 139 19time write (sec) 365 48time read (sec) 328 95
Data flowchart
Cell MappingCalibration ConstantsDetector Alignment
TestbeamRAW Data
SimulationData
Noise - Pedestals
EXE
EXE
Testbeam Datafor Analysis
Simulation Datafor Analysis
A n
a l
y s
i s
Graphical User Interface
.cross-platform GUI application under development for@ online histogramming@ basic analysis@ monitoring@ event display
.using ROOT GUI classes based on Xclass’95 widget library
( ACBCBCD EGFCFIHKJMLMN<OCO PQOMRISUTVJ�WIXYR�T�ZVW P\[VWIB )
.basic layout has been implemented and tested onRedhat 7.3 (gcc 3.2.3 or 2.96) with ROOT 4.00 or 3.10
GUI: histograms and action widgets
GUI: tabs for expanded info, settings
GUI: status table
GUI: active pads to invoke ROOT actions
Summary . Simulation
: current poor understanding of hadronic shower developmentleads to significant discrepancies among available packagesand does not permit reliable model independent predictions
: testbeam data will help to test-validate-improve simulation code
. Reconstruction
: calorimeter clustering algorithm is being developed andfirst performance tests are satisfactory
. Testbeam preparation
: work on testbeam related software to support the running andanalysis phase has started and rapid progress is expected