I want to thank you for your response. It will help a great deal in my understanding of particle momenta in emulsions.
But, my problem is with definition of track, not particle momenta.
Here is a excerpt from the text I was reading through:
Each found track is fitted taking into account multiple scattering in the various materials
between the hits. The track fit yields a χ2-probability P
In this calculation,
the hit resolution is assumed to be σ = 180 μm. Two effects cause the P(χ2)
distribution
to deviate from the expected uniform distribution. First, the momentum of the track can
be unknown. In this case, a momentum of 1GeV/c is assumed. Depending on the real
momentum of the particle, this can over or under estimate the multiple scattering contribution
to the hit resolution. Therefore, a cut on P(χ2)
Introduces a bias on momentum,
selecting preferably high momentum tracks. Second, the hit-residual distribution has
tails with a rms width of about 750μm. These tails are probably due to cross-talk in
the input window of the image intensifier or backscattering in the multi-channel plate.
This leads to reconstructed tracks with very low χ2-probability
If anyone is interested, full text can be found on:
https://openaccess.leidenuniv.nl/handle/1887/12079
(excerpt is from page 70)