Title:Success of collinear expansion in the calculation of induced gluon emission

Abstract: We clarify the confusion in a recent paper by Aurenche, Zakharov and Zaraket (AZZ) \cite{azz} over the procedure and region of validity of the collinear approximation in the twist expansion approach to induced gluon emission in deeply inelastic scattering (DIS) off a nucleus target. We point out that, in this approach to the semi-inclusive spectrum, the transverse momentum $\vec\ell_\perp$ of the induced gluon must be fixed in the collinear expansion in the transverse momentum $\vec k_\perp$ of the initial partons, therefore the result is valid for $<k_\perp^2> \ll \ell_\perp^2 \ll Q^2$. In the twist-four contribution, one can single out the double-hard term corresponding to collinear quark-gluon Compton scattering which can be calculated independently of the collinear expansion approach. We will discuss the connection between the collinear approximation in the twist expansion approach and the small $k_T$ approximation of the results in the Light-Cone Path Integral (LCPI) and Gyulassy-Levai-Vitev (GLV) opacity expansion approach. We point out the misconstrued variable change by AZZ before the $k_T$ expansion in LCPI and opacity expansion approach, without which one obtains the same result for the induced gluon spectrum under collinear approximation as in the twist expansion approach. We also show that corrections beyond the collinear approximation to the transverse momentum integrated gluon spectrum within the static potential model in the GLV approach give rise to a logarithmic factor difference from the result of collinear approximation.