In this work, we apply wavelet-based resolvent analysis to study intermittent bursts in turbulent channel flow at Re_\tau = 2000. It has been argued that the behavior of these bursts can be explained by the Orr mechanism, i.e. the transient linear amplification and decay of velocity perturbations via their forward-tilting by the mean shear. Such a mechanism cannot be represented by individual Fourier modes in time, which are traditionally used to define resolvent modes. The wavelet-based formulation of resolvent analysis has thus enabled this linear technique to study time-localized phenomena like the Orr mechanism. In this work, we compute transient resolvent modes that target bursts in channel flow and compare the time scales of the response modes to bursting in turbulent direct numerical simulations (DNS). Furthermore, we project a time series of a channel simulation onto the resolvent forcing mode to probe its ability to predict bursting in DNS. Our analysis shows that the transient resolvent modes seem to accurately capture the coupling between tilting angle and energy growth of velocity perturbations characteristic of the Orr mechanism.