This note is a summary of my understanding of the status of some
neutrino working group analysis results and planned activities for the
next few months. Two meetings of the working group were held during
the collaboration meeting. At the first meeting, new
analyses on data since March 1994 were presented in detail. At the
second meeting, plans for proceeding in the future were discussed in
detail and agreed upon by all present. This is my account of those
discussions.
Analyses
--------
Three analyses were presented at the Gran Sasso meeting and it is
my understanding that a fourth analysis is underway at Indiana. The
status of the presented analyses:
Okada: Most (but not all) attico data analyzed. Analysis technique
does not depend on traditional track fitting in ST's. Rather, the
scintillator positions are used to define a road which must contain a
minimum number of ST hits which varies depending on the geometry. Cuts
on raw TDC values, number of scintillator tanks, numbers of ST's,
requirements of ST's in various planes and minimum path length are all
made to reduce backgrounds. Allowed geometries include both 3
scintillator plane and 2 scintillator plane events with tracks
crossing through either the bottom or attico. Hence, no requirement of
minimum muon energy has been imposed and contained vertex events (at
least of some classes) are allowed. A smooth zenith distribution is
obtained. A comparison to a MC prediction which does not include
contained vertex or low-energy backscattered events was made and the
two look similar. A number of corrections for inter-ERP, south face
and a few other tanks have been made to the standard data base in
order to remove background problems. These corrections should be
easily transferred into the standard data base. "Private"
muon-astronomy-type DST's have been used based on the "old" production
program. Due to various bug fixes which have been made to that code,
use of the new mu-ast DST's may prove to be the most economical way to
assure understanding of possible effects. Surprisingly (to me at
least) the acceptance near horizontal in this analysis is still
significantly smaller than near the vertical. This will be studied.
Sanzgiri: Using the more traditional approach of reconstructing a ST
track and requiring it to point through scintillator box hits,
Sanzigiri has used a subset of the "Okada DST's" to look for upgoing
muons. As in the Okada analysis, both 2 and 3 plane events are allowed
without requirement of lower detector absorber. Due to background
problems, some specific boxes, south face and interERP events have
been elminated from the final analysis. A smooth zenith distribution
is obtained and compared with MC but the MC here also does not yet
include events resulting from contained vertex or backscattered events
in the attico.
Montarulli/Ronga: Most of the existing data have been analysed using
an approach similar to the 6-month run. New geometries including the
attico have been added but not yet compared to MC due to the lack of
the aforementioned upgrades which are needed. A zenith distribution
based on requirements similar (same?) to the 6-month run was performed
and compared both to earlier data and MC (minimum 3 m pathlength in
the absorber part of the detector). No interERP events were used and
some various corrections have been applied to the standard data base
in order to remove some backgrounds. The features of the earlier data
appear to remain, including the large deficit compared to expectation
near the vertical. All of the zenith bins area consistent with the
previous measurements when scaled for livetimes and
acceptance. Several events of the contained-vertex variety were shown
but no comparison to expectations was made. One event with a downgoing
muon and a not-clearly-related upgoing track was shown. This event
remains not understood.
Hawthorne: Not presented at the meeting. I don't know the status of
this.
Spurio: Events with downgoing muons with clearly related backscattered
tracks in both the lower and upper detector were shown. Due to the
very low energy threshold, it was argued that such events could
provide significant background for attico-only upgoing events.
Summary: Due to the wide variation in runs considered, analysis
approach, database differences and geometries considered, no useful
comparison of analyses was possible at this time. Clearly, it is
essential to define constrained data sets and analysis approaches
which can be compared. This was done at the second group meeting.
Goals for the next few months
-----------------------------
The group agreed that the highest priority goal is to produce a new
analysis on atmospheric neutrinos which is ready to release by the
time of Neutrino 96 (late June in Helsinki). This requires that the
basic result be prepared in time for the next MACRO collaboration
meeting (April 22-24). A secondary goal (not considered very realistic
by several people) is to produce a journal letter ready for submission
at the time of the conference *if* the new results combined with old
data make MACRO data statistically significantly in agreement or
disagreement with expectations on atmosperhic neutrinos. Although some
work may continue on it in the next months (primarily DiCredico) it
was agreed that the group does not have time to properly address new
neutrino astronomy results prior to the time of the Venice conference
given the demands of the primary goal.
Plans for meeting the primary goal
----------------------------------
In order to meet the primary objective stated above, much work in
all of the analyses and comparison of the analyses is required. In
order to make this possible significant organization and
standardization is essential. The following issues were discussed and
agreed in the second working-group meeting:
1. Data Sets: It was agreed that it is not essential to have exactly
the same DST's but the production methods for DST's must be clearly
described and any differences understood. Some bug-fixes to the old
muon astronomy DST production were discussed and it appears that there
could be some significant impact on the Okada DST's. Okada will study
the issue and may choose to use the new DST's as the easier way to
ensure that the acceptance is understood.
2. Run List: It was agreed that the run list would begin with 7473
(April, 1994 when all attico ST's were functional) and end with the
last run on Dec. 31, 1995. People doing analyses were assigned the
task of studying the impact on the statistics if full 6 SM runs only
are used or various SM combinations are allowed. A decision will be
made on which geometries to use based on these studies. The analysers
should propose detailed run number lists which are to be compared and
combined to form standard run numbers which are to be used. Runs
appearing in some lists but not others will be questioned for the
reason for non-inclusion. Bad reasons (eg. my dog ate the data tape)
will not be considered valid for exclusion of data unless the impact
on statistics remains very small. In order to place the focus on
checking data quality rather than quantity, DGM (at least) plans to
not consider any data beyond the end of 1995 in any of the analyses
for presentation this summer.
3. Calibration database: A single calibration data base will be used
by all analyses. Jim Musser and Alice Hawthorne (not present at the
meeting) will be consulted to decide the best way to produce the data
base which folds together all of the knowledge gained by people working on
analyses. Those who have knowledge of specific problems and/or how to
correct such problems are strongly encouraged to help out by
forwarding their detailed information to Jim and Alice. Complaints of
the generic "it should be better" variety are discouraged.
4. Mini-DST's: Each analysis should produce a mini-DST which contains
all events with track segments which reconstruct with 1/beta<0. These
DST's will be used by the group in general and other analysers to
check consistency, variation in acceptance, background differences,
etc. The mini-DST's should be produced with the consistent data sets,
run lists and calibrations mentioned above.
5. Definition of upgoing-muon geometry: Due to many issues about
acceptance and background for attico-only upgoing events, it was
agreed that the main focus of analysis in the next few months will be
only those events which pass through at least two meters of the lower
half of the detector and which are clearly through-going events. The
specific definitions of such events are:
A: The distance between the points defined by the intersection of
the event track and the plane defined by the middle of two
scintillator layers in the lower part of the detector must be
at least 2 meters.
B: The event must contain either a B, E or W scintillator hit in
the lower part of the detector or at least one ST hit "on the
track" which is in the bottom plane of ST's or in at least one
of the three outer planes of ST's on the E and W lower part of
the detector.
6. Upgoing-muon lists: Based on analysis cuts, analysers will produce
lists of all upgoing muons as defined above. The lists should
include information such as boxes hit, ST track information,
positions in scintillator boxes versus ST location, energy in
scintillator, raw scin information, etc. Such lists are invaluable
in comparing analyses and understanding why some events are
included in one or more analyses but not others. DGM will produce
a more specific list and format for such information.
7. Monte-Carlo: A common MC data set (fully simulated) will be
produced by Maurizio Spurio. An initial data set based on the same
flux criteria as used for the 6-month run will be produced. A
second set based on updated flux inputs (cross-sections, QE events,
etc) will be produced once various improvements are agreed upon.
New flux tables will be produced by Mikheyev to include energy of
parent neutrinos in order to permit easy oscillation analysis.
8. Acceptance checks and studies: All analyses must include checks on
acceptance and efficiencies. In particular, the following should be
provided:
A: Comparison of flux of down-going muon intensity versus depth
for each analysis to be compared with that calculated from the
published result with table provided by Ernesto Lamanna.
B: Comparison of acceptance versus zenith on downgoing muons for
ST versus scintillator.
9: Written MACRO memos: Written MACRO memos describing analyses should
be distributed by the end of February. These memos are essential
for careful understanding of comparison between analyses. They are
necessary by the end of February to allow enough time for
comparison and iteration prior to the April collaboration meeting.
10: Single MACRO/pub: A single MACRO/pub will be compiled based on the
results of the several analyses.
11: Improvements in acceptance at large zenith: All analysers are
urged to study the issue of improved acceptance near the
horizontal. Our existing data should permit the angular
distribution to be of interest but if the acceptance near the
horizontal is too small the result may not be well received,
regardless of how many arguments we make about how well we
understand it.
12: Meetings: I order to improve communications, more regular meetings
will be scheduled in the next few months. The next joint Italian-
American meeting will be held (nominally) on January 16th with
video connections. A meeting in the U.S. will nominally be planned
for December 19 or 20 depending on availablity of people and video
connections.