- Proposal report (A Proposal of the DOM response implementation method
for the Photonics table)(ps)
(pdf)
 Purpose
of the Monte-carlo
In the last collaboration meeting at Delaware, the data structure of
the photonics is discussed for next release.
Applying the averaged sensitivity to all OMs, current version of the
photonics stores the number of photo-electrons at every distance between
a light source to an OM and the angle of incidence, in order to save
computer memory. It convolutes absolute quantum efficiency at the photo
cathode of PMT, acceptance of arrival photons at the surface of PMT,
photon absorption and reflection in the OM glass and gel, and the wave
length spectrum after the propagation of given distance in non-homogeneous
ice. The size of table is up to 7GB and will grow up in the next release
as the detector volume increases from the AMANDA to the IceCube.
On the other hand, the optical properties of individual PMT are measured
carefully (see our PMT
page) which has indicated that a local but systematic deviation
of the tube gain and the collection efficiency
on a photocathode surface is not negligible and is indeed a major
factor in PMT response.
These behavior also appears differently from a tube to tube and it
would be important to implement these measured results to the next
detector Monte-Carlo simulator in order to reduce the uncertainty
of the DOM
response. Adding these informations to the photonics may cause
considerable expansion of table size, however, no one would be able
to load the table on
his/her computer
any more.
In order to find the good solution, we argued following possibilities:
- Use same data structure as in the AMANDA simulation chain
- Let the photonics calculate number of the photons arriving at
individual DOM location instead of number of photo-electrons
- Add new field
for the wave length spectrum of the arrival photons
The first one is just a back up plan. The second choice assuming
100% efficiency to all DOMs in the photonics table, then the difference
of each DOMs are simulated in the detector Monte-carlo side.
Remarkable advantage is that the method doesn't require any additional
field to the photonics itself while an individual response of DOMs
can be simulated in resonable manner in the DOM simulator module. Obviously
it will be the most favorable solution, however
there is one drawback: we cannot know the wavelength of the arrival
because the wavelength factor is convoluted in the photonics, although
the response
of DOMs depend on the wavelength.
The last choice may cover this drawback by paying relatively small
penalty about table size.
Before going to the third choice, we tried to figure out whether
we can recover the drawback without increasing table size by assuming
analytical
wavelength spectrum after the propagation inside pole ice. A toy
Monte-Carlo program is developed for this purpose and its details and
several results
are reported in a proposal report "A Proposal
of the DOM response implementation method for the Photonics table".
The systematic shift of the number of photo-electrons by assuming an
analytical
wavelength spectrum would be 5%, which would be better than
the uncertainty due to our incomplete knowledge of IceCube ice profile.
[top]
Document
- Proposal report (A Proposal of the DOM response
implementation method for the Photonics table)
[top]
Figures
- Number of photons after propagation in ice vs wavelength (eps)
(pdf) (gif)
- Absolute QE curve (measured by Hamamatsu Photonics) (eps) (pdf)
(gif)
- Wavelength spectrum at different distances (eps) (pdf)
(gif)
- Wavelength spectrum at different distances with QE curve convolution(eps)
(pdf)
(gif)
- Number of photo-electrons (Model / MC) (eps) (pdf)
(gif)
[top]
|
