N ov 2 00 1 Phenomenological analysis of experimental data on η - photoproduction on protons ∗

The results of linear regresion of differential cross sections, Σ-and T-asymmetries of η-photoproduction on protons in energy region from threshold up to 1 GeV are presented. Serious contradictions between angular distributions measured in different laboratories are revealed. The energy dependance of regression coefficients may be due to transition from energy region of S 11 During the past years η-photoproduction on protons has attracted increasingly high interest. This is due not only because this item is a new physical phenomenon different from photoproduction of pions but mainly because η-photoproduction should proceed through the small number of nucleon resonances. Even in energy region up to 1 GeV there will be not too many overlapping resonances that permits to extract reliable information on resonance parameters from experimental data. Complete phenomenological analysis of experimental data on photoproduction, as a rule, encounters a number of problems, e.g. solving of nonlinear equations, removal of continuous and discrete theoretical ambiguities, elimination of experimental ambiguities, etc. Analysis of experimental data may be naturally divided in two stages [1]. The first is the linear regression which provides the information about the number of partial waves that contribute to the measured experimental characteristics of process and provides information on the resonances concerned. The linearity of the model used ensures that the estimates of regression coefficients are unbiased. The second is to determine the multipole amplitudes. This paper is confined to the first stage of analysis. We have analysed all known experimental data on differential cross sections (angular distributions) of process γp → ηp [2–4], and also the data on polarization observables, i.e. angular distributions of asymmetry Σ, measured with linear polarized beam [5] and angular distributions of asymmetry T, measured on a polarized target [6]. The energy independent analysis consist in expanding angular distribution of the observables at definite energy using Legendre polinomials. To find how many terms in this expansion provide the best description of data standart statistical procedures including the Fisher criterion were used. Unlike the energy dependent