This function has been provided by I. Tomalin and introduced in ALPHA 122.44 in March 1998. This is the so called ``sagitta correction". It corrects particle momenta for the effects of residual distortions in the central tracking detector. It depends upon cos() and the year of data taking.
Every year, even after the detector alignment is finished and corrections have been made for field distortions etc., it is found that Ebeam/P in Z0 events is not precisely 1, presumably because of residual distortions. (Typically, in the region | cos()| > 0.9, Ebeam/p is about 0.94 for +ve tracks and 1.06 for -ve tracks. Elsewhere, Ebeam/p is usually consistent with 1 to within a percent or so. The effect is not quite forward-backward symmetric).
The relative bias in momentum is proportional to P, so most people analysing hadronic events can ignore it. Exceptions include analyses using the ECAL electron identifiers in the region | cos()| > 0.9 or analyses which are very sensitive to systematic biases in the momenta (e.g. jet charge, polarization).
QPCORR provides a correction for the momenta based upon Ebeam/P measurements in Z0 events. It assumes that the corrections for -ve and +ve particles are equal in size, but of opposite sign. This is observed to be true, apart from a constant offset, Ebeam/p = 1.002, which is also present in the MC and so not corrected for.
In principle, the correction depends upon your track selection cuts, but providing that the corrections have a small effect on your analysis, you can ignore this. Arguement NVDET does correct for this to first approximation however.
Input arguments:
Important remark: data card PCOR allows to call QPCORR automatically: