10 - H1 - Desy

144 Conclusions and outlook
is consistent with the one made in the previous H1 publication [35]. The improvement can
be associated with the lower NLO corrections in the exclusive case, which may suggest
that the higher order corrections for inclusive prompt photon cross sections may still be
significant.
The cross sections in bins of x LO
γ and x LO
p variables were also measured and found to be
consistent with the predictions within errors.
Correlation between the prompt photon and the hadronic jet were studied in the part
of the phase space with enhanced contribution from direct events (x LO
γ > 0.8) and the
complementary one (x LO
γ < 0.8). In both cases, the predicted rate of events with the
photon not being fully balanced by the leading hadronic jet is below the measured cross
sections. Higher order corrections are expected to introduce additional imbalance, possibly
improving the calculations. The fraction of events in the tails of the studied distributions
may also be increased by introduction of the additional intrinsic 〈k T 〉 of the partons
entering the hard interaction. The comparison of the measured cross sections in the
direct enhanced phase space with the QCD calculations corrected for the 〈k T 〉 effects
leads to the conclusion that though the improvement is clearly visible, the intrinsic 〈k T 〉
can not solely remove the observed discrepancy.
The difference in the azimuthal angle between prompt photon and the accompanying jet
∆Φ in the exclusive phase space with additional requirement x LO
γ < 0.8 shows vulnerability
of the FGH calculation. The back-to-back region, ∆Φ → 180 ◦ , is sensitive to multiple
soft gluon radiation which limits the validity of fixed order calculations. The LZ calculation
includes multiple gluon radiation in the initial state before the hard subprocess and
describes problematic region significantly better.
The prompt photon measurement presented in this thesis may still be improved. Most
significantly, higher statistics of monitor sample which consists of pure photons is needed
to improve the cluster shape description and decrease the leading systematic error. The
limited statistics of deeply virtual Compton scattering events in the kinematic range of
this analysis demands finding more suitable source of the photon monitor sample. The
errors of this measurement are low enough to resolve the deficits in the QCD calculations
applied for the comparison.
The conclusions drawn from this study may become stronger by improving the theoretical
understanding of the prompt photon production. Especially, better understanding of the
hadronisation model would help in decreasing the error of the predictions. The new parton
density functions coming from combined H1 and ZEUS data will help with calculation of
more precise predictions. The sizeable NLO corrections lead to the conclusion that higher
orders should be investigated in future. The box diagram, significantly contributing to
the prompt photon cross sections, on the level of 10%, should be taken into account in all
the calculations, as well as contribution from the quark-to-photon fragmentation function.
The latter should also be studied in more details.