INFN Padova - Italy
Data reduction techniques and extraction of physics parameters in
the ICARUS detector
A liquid argon time projection chamber (LAr-TPC) working as an electronic
bubble-chamber, continually sensitive, self-triggering, able to provide 3-D
imaging of any ionizing event together with a good calorimetric response was
first proposed by C. Rubbia in 1977. In order to verify the feasibility of such
a detector, the ICARUS collaboration started in 1985 an intensive R&D program
aiming to solve the main technological problems. The satisfactory results
obtained on small scale tests allowed us to start in 1989 the construction of a
3 ton prototype which is presently working at CERN under stable conditions
without interruptions since May 1991, collecting events from cosmic rays and
monochromatic gamma ray source.
The talk describes the working principles of such a detector, showing the
results of the research program applied in the present prototype.
Some relief is then given to the read-out apparatus, a custom designed
VME-based multichannel waveform recorder. The huge number of channels and the
high sampling rate essential to achieve the high resolution in the track
detection make the data acquisition architecture a crucial point for event
detection rate and self-triggering capabilities. At present a complete signal
analysis is performed to define those parameters needed to tune up algorithms
and filtering methods that will become necessary in the future step of ICARUS
detector (hundred to thousands of tons of argon, i. e. tens of thousands of
readout channels). A particular effort is made to optimize track localization
algorithms, in order to achieve high efficiency data reduction by storing only
those portions of signal which contain the track (thus lowering up to three
orders of magnitude the event size) .
From a human interface point of view, the data presentation relies on an event
imaging very similar to the old fashioned bubble chambers. Besides from giving a
direct feeling of what happens inside the detector, this kind of representations
opens a way to the wide panorama of image processing applications: tasks as
detail enhancement, pattern recognition or 3D reconstruction will helpfully
inherit support from more general procedures.