The large momentum transfer provides the hard scale necessary for the application of perturbative QCD (pQCD) models. The results are compared to two theoretical predictions: a fixed order calculation in which the hard interaction is approximated by the exchange of two gluons and a LL calculation in which it is described according to the BFKL evolution.

The t dependence of the e p -> e rho Y cross section is presented in Fig. 1. Both the experimental result and the theoretical predictions are normalised to unity. Over the measured t range, the data are well described by the BFKL model (solid line) in contrast to predictions of the two-gluon model with fixed (dotted line) or running (dashed-dotted line) alpha_S.

The spin density matrix elements characterise the helicity states of the rho meson and the intermediate photon. The Fig. 2 presents the dependences versus |t| of the r⁰⁴₀₀, Re[r⁰⁴₁₀] and r⁰⁴_1-1 matrix elements extracted from the H1 data together with previous results from the ZEUS Collaboration. The small values of r⁰⁴₀₀ indicate that the probability of producing a longitudinally polarised rho meson from a transversely polarised photon is low. The non-zero values of Re[r⁰⁴₁₀] and r⁰⁴_1-1 indicate however a violation of s-channel helicity conservation (SCHC) between the intermediate photon and the produced rho meson. The two-gluon predictions (dotted lines) are unable to describe the measured spin density matrix elements. For the BFKL predictions (solid lines), r⁰⁴₀₀ is well described but the prediction for r⁰⁴_1-1 is too negative and the wrong sign for Re[r⁰⁴₁₀] is predicted. Although, unlike the two-gluon models, the BFKL model is able to describe a predominantly longitudinal rho meson production, both considered models fail to describe the full set of spin density matrix elements.