Participating:
 
  Rene Bellwied (Wayne State)
  Tony Johnson (SLAC)
  Dean Karlen (Carleton)
  Homer Neal (Yale)
  Sanjeev Pandey (Wayne State)
  Keith Riles (Michigan)
  Mike Ronan (LBNL)
  Bruce Schumm (UCSC)
  Ron Settles (MPI, Munich)
  Nic Sinev (Oregon)
  Rick Van Kooten (Indiana)
  Wolfgang Walkowiak (UCSC)

Introduction & Overview (Riles)

  The main purpose of this teleconference was to see where
we stand in simulation status and available manpower in order
to decide how best to focus group efforts in the next several
months.

  New baseline tracker configurations have been defined for
the small ("S2") and large ("L2") trackers, based on the
August meeting at SLAC, and with refinements / optimization
by a subgroup led by B. Schumm. Descriptions, drawings and
estimated resolutions for the new designs can be found at
the working group web site. Highlights of recent changes
include:
 
 * The same CCD vertex detector is used for both detectors,
   with a 1-cm beam pipe and the first CCD layer at 1.2 cm.
   (Schumm raised the issue of reducing the beampipe radius still
    further for the small detector, taking advantage of the
    higher magnetic field, in the spirit of pushing each
    detector design to its optimum before comparison. It was
    decided, however, to revisit this issue at the Berkeley meeting
    in March.)

 * The L2 tracker now has five forward disks
   (silicon microstrips with +/- 20 mrad stereo.

 * The L2 TPC's inner radius has been moved out to 50 cm to
   give better momentum resolution at forward angles
   (making room for the disks).

 * An intermediate double-sided silicon microstrip layer is
   placed just inside the TPC for |cos(theta)| < 0.5. The
   immediate motivation is a factor of two improvement in
   momentum resolution at high pt, but the layer may also
   prove beneficial in pattern recognition.

  Schumm has written a full 3-D resolution parametrization
program for deriving 5x5 error matrices as a function of
momentum and cos(theta). The program, based on the Billoir
approach, was used in the above tracker optimization and
includes multiple scattering effects from active and dead
materials. Its momentum resolutions were cross-checked
at a variety of [p, cos(theta)] points with an independent,
brute-force Monte Carlo program (written by Riles) using a
Minuit-based helical fit to randomly generated (& scattered)
tracks. As expected, the Billoir approach gave better resolution,
in general. The agreement between the programs was found to be
good where the Billoir improvements in resolution were expected
to be small.

  As discussed below, the track simulation & reconstruction are
in a transition phase, with some new tools not yet incorporated
into the released software packages. The new S2 and L2 designs
are not yet in the simulation, but soon will be. A major
gap remains the absence of track finding for projective geometries
(e.g., microstrip forward disks and the microstrip barrel option).

Status of LCD simulation & reconstruction software (Johnson)

(Detailed slides were made available on the Web before the meeting.
 Not all of the presentation will be summarized here.)

  The present track reconstruction in JAS is based on Ronan's
pattern finding program (adapted from Babar mini-TPC reconstruction
code) with improvements in optimization and fitting provided
by Sinev. Sinev has also added standardized hit smearing / inefficiency
and random background hit overlays. Still lacking:
 * Track finding for projective geometry (includes all forward disks!)
 * Track finding using Monte Carlo cheat information
 * Final incorporation of R. Kutschke's (FNAL) Kalman-filter track
   fitter. The initial version failed because of residual glitches in
   extracting stand-alone code from the FNAL Mcfast program. Kutchke
   has provided fixes which have not yet been tested, but which should
   be in place by the end of the year.

  Cluster finding in the calorimetry is well advanced with several
available options and tunable parameters. An initial track-cluster
association (written by Ronan) has not yet been incorporated into
the standard JAS release, but soon will be. For all-hadronic
events, track finding requires roughly 5 CPU seconds vs 0.1 second
for calorimeter clustering.

  R. Cassell (SLAC) is developing a test suite (histogramming
and statistics) to allow rapid, standard checking of new
software releases for early bug detection. He plans to turn
his attention to endcap track finding when finished, with
assistance from Ronan. Ronan suggests that a good starting
point would be to treat the stereo doublet hits as true 3-d
space points in the hope that ambiguities will be manageable.

  The reconstruction program will be made stand-alone to allow
reconstruction from standard input/output files (SIO) with output
that can be analyzed in ROOT or JAS. As a default, reconstruction
will in the future be run after simulation at the CPU farms so
that end users need not run it themselves. Reconstruction code
has been moved to a CVS repository for universal access to code.
Those wanting to update code must contact Johnson to get an account.

  M. Iwasaki-Abe (Oregon) is working on a 5-dimensional interpolation
program to incorporate Schumm's error matrix estimates into
the fast Monte Carlo. The Babar track parameter convention
is used in the LCD simulation / reconstruction.

  Two sets of track classes will be defined in JAS, allowing for
both beam-constrained and non-beam-constrained tracks to be
use in analysis. The latter are needed for vertexing, b-tagging, etc.
Two-hit resolution effects will not be included in detector
simulation, but can be included in reconstruction in the same
way as single-hit smearing & inefficiency.

 
           Status reports from individual groups
 
  Several groups whose R&D proposals were approved in the summer
were asked to report. Specifically, they were asked to
report on manpower availability, restate their simulation interests,
and comment on their ability / willingness to assist with developing
simulation and/or reconstruction infrastructure.

U.C. Santa Cruz (Schumm)

  A new postdoc (Walkowiak) has been hired (started Nov 1) to spend
half-time on linear collider studies.

  The group is interested in many tracking issues, including:
  * Doublets vs singlets in the S tracker
  * Ensuring that pattern recognition is even-handed for the
    S and L detectors.
  * Benefits of stereo and whether one could restrict its
    use to only the outer tracking layer to allow very long,
    low-material microstrip ladders for the barrel. (The vertex
    detector would provide z information inside the tracker.)
  * The benefits / tradeoffs in better impact parameter resolution
  * (Longer term) Pulse shaping techniques and tradeoffs between
    resolution and background suppression

In general, the group is open to helping with infrastructure and
will look hard at Johnson's to-do list to see where they can be
useful.

Wayne State U. (Bellwied)
 
  An MOU has been signed with Fermilab, making available funds for
half of a postdoc starting December 1. The other half of funding
will be provided by Wayne State funds to work on the STAR or CLEO
experiments (hopes for a BNL contribution ended with a recent DOE
directive concerning new BNL initiatives). The position has not been
filled, but advertisements are out. Interested candidates are welcome
to apply.

  The group is most interested in pursuing simulation issues related
to silicon drift detectors (SDD) for the S detector barrel and for the
intermediate layer of the L detector. In particular, they want to
study pattern recognition benefits in detail. (Schumm offered to
help in this effort.) They intend to use ROOT rather than JAS to
carry out their studies.

  They would be happy to contribute to improving the infrastructure
of the SDD simulation at the detailed hit level, once the postdoc
is on board.

Indiana U. (Van Kooten)

  An MOU with Fermilab is in preparation and should be completed
soon to make available funds for a half-postdoc. Present postdocs
in the group are too busy with D0 work. A new person must be hired.
The other half of the support has been requested from DOE. University
funds may serve as a fallback. A realistic timescale for hiring a
new person is March-April.

  The group plans to study the benefits of an intermediate tracker for
the L detector and of an outer, precise layer, just inside the
calorimeter. Several different technologies will be explored at
the simulation level, including straw chambers and scintillating
fibers. They will begin with resolution parametrizations based
on Schumm's program, but will eventually carry out detailed
simulations to address pattern recognition issues. (Ronan suggested
that many issues can be addressed via on-the-fly hit simulators
during reconstruction.)

  They would be happy to contribute to simulation/reconstruction
software, work that would be carried out primarily by the new
postdoc, while students can help in other areas.

U. Michigan (Riles)

  The Michigan group is NSF funded, and no FY99 funds were available
for linear collider studies (contrary to earlier expectations).
A proposal was submitted to NSF in the summer, but target date
restrictions prevented NSF's sending it out for review until mid
September. If FY00 funds are approved, the hiring process will
begin immediately, but probably no earlier than February, given
the time required for the ongoing external review.

  The group plans to study the impact of the tracker on Higgs physics,
focussing on three particular issues:
 * Effect of momentum resolution on the ZH -> l+ l- b bbar channel
   (purity, mass resolution)
 * Effect of tracker on b tagging purity/efficiency for mass-based
   vertexing selection
 * 3-D vs 2-D tracking technology (jet reconstruction, machine
   background suppression)
 
  Most of the above work will not get done, however, until a
postdoc is on board. In regard to infrastructure, we would be
willing to help develop / refine track finding algorithms for
projective geometries.

Yale (Neal)

  The Yale group did not request postdoc support in the tracking
proposal, but is active in tracking R&D.

  The group is focussed on optimization of the S tracker's size
from the global viewpoint of tracking & calorimetry tradeoffs.
In particular, they have been looking at effects on supersymmetry
and top physics of outer tracker radius.

  An important part of this work requires an effective vertex
finding program. Although several persons have expressed
interest and done some work in developing such software,
it was not clear when a program will be ready for general use in
LCD analysis. After some discussion of priorities, Neal volunteered to
coordinate the work to incorporate a viable vertex finder into
the standard LCD reconstruction package. Ronan offered to help
identify persons already working in this area.

LBNL (Ronan)

  Ronan reported on the recent Obernai linear collider meeting.
The Europeans are turning more attention to full reconstruction
in their analyses, raising the possibility of combining European
and North American efforts for a common analysis framework.
An interesting physics issue raised at Obernai is reconstructing
ZHH events, which provide a measure of Higgs self-coupling. The
analysis puts a premium on b tagging. Another physics-driven tracking
issue, one raised at Sitges by Japanese physicists, is the effect of
overlapping two-photon events on Higgs reconstruction. Fast timing in
tracking [O(1 ns)] would help in suppressing this background.
Van Kooten pointed out that an intermediate tracker based on
scintillating fibers could provide very good timing.

  Ronan also reported that the European detector's design now looks
very similar to the L2 design (not surprising, since L1 was highly
derivative of the original European design -- KR). In addition,
Active Pixel Sensors are being studied seriously in Europe as a
candidate vertex detector.
 
  Ronan announced the imminent release of two papers concerning
the JAS framework and tracking code. They will soon be available
at his home page. He suggested that the time has come to upgrade
the JAS event display program to allow more graphics manipulation and
information display. He suggested the increasingly popular WIRED
program as a replacement and plans to pursue this. Johnson mentioned
that he hopes to put a student to work on this soon, too.

Carleton U. (Karlen)

  The Carleton group is working on a Gas Electron Multipler (GEM)
readout for a TPC. The GEM readout has the potential to improve both
single-particle spatial resolution and two-particle separation
by significant factors. If the potential is realized, these
improvements could allow a smaller outer radius for the TPC
for a given desired momentum resolution, making the L detector
more affordable and perhaps allowing a finer-grained calorimeter.
R&D is underway, and a proposal for further work has been submitted
to NSERC.

  Ronan raised a concern that 200 MHz sampling would be needed for
a large number of readout channels. Karlen stated that 100 MHz
sampling might well be sufficient. Ronan also remarked that
recent European readout designs are also much improved, with the
potential of reaching the diffusion limit of the chamber in 2-track
separation. He offered to work with the Carleton group.

A.O.B.
 
  Van Kooten requested that Higgs event samples be produced in
the next large-scale event generation at one of the CPU Farms.
Neal requested that large two-photon event samples be generated.
Even a 100K two-photon event sample is not adequate for some studies.
As a fallback, it was suggested that very large 4-vector samples be
made available without detector simulation / reconstruction.
Johnson reminded everyone that event sample requests should be
sent to Gary Bower.

  The next tracking group teleconference is scheduled tentatively
for late January or early February.