A detailed analysis is presented of the diffractive deep-inelastic scattering process $ep\to eXY$, where $Y$ is a proton or a low mass proton excitation carrying a fraction $1 - \xpom > 0.95$ of the incident proton longitudinal momentum and the squared four-momentum transfer at the proton vertex satisfies $|t|<1 {\rm GeV^2}$. Using data taken by the H1 experiment, the cross section is measured for photon virtualities in the range $3.5 \leq Q^2 \leq 1600 \rm GeV^2$, triple differentially in $\xpom$, $Q^2$ and $\beta = x / \xpom$, where $x$ is the Bjorken scaling variable. At low $\xpom$, the data are consistent with a factorisable $\xpom$ dependence, which can be described by the exchange of an effective pomeron trajectory with intercept $\alphapom(0)= 1.118 \pm 0.008 {\rm (exp.)} ^{+0.029}_{-0.010} {\rm (model)}$. Diffractive parton distribution functions and their uncertainties are determined from a next-to-leading order DGLAP QCD analysis of the $Q^2$ and $\beta$ dependences of the cross section. The resulting gluon distribution carries an integrated fraction of around 70% of the exchanged momentum in the $Q^2$ range studied. Total and differential cross sections are also measured for the diffractive charged current process $e^+ p \to \bar{\nu}_e XY$ and are found to be well described by predictions based on the diffractive parton distributions. The ratio of the diffractive to the inclusive neutral current $ep$ cross sections is studied. Over most of the kinematic range, this ratio shows no significant dependence on $Q^2$ at fixed $\xpom$ and $x$ or on $x$ at fixed $Q^2$ and $\beta$.
Details of systematic errors from the Data sample taken in 1999 and 2000. D_UNC ==> uncorrelated systematic error. D_LAR ==> LAr hadronic energy scale. D_ELE ==> SPACAL electromagnetic energy scale. D_THETA ==> scattered electron angle. D_NOISE ==> calorimeter noise treatment. D_XPOM ==> reweighting the simulation in X(pomeron). D_BETA ==> reweighting the simulation in BETA. D_BG ==> background subtraction. D_PLUG ==> plug energy scale. Q**2 ==> reweighting the simulation in Q**2. SPA ==> SPACAL hadromic energy scale.
Forum