Geophysical granular flows generate seismic signals known as 'slidequakes' or 'landquakes', with low-frequency components whose generation by mean forces is widely used to infer hazard-relevant flow properties. Many more such properties could be inferred by understanding the fluctuating forces that generate slidequakes' higher frequency components and, to do so, Arran et al. (2021, https://doi.org/10.1029/2021JF006172) (A21) compared the predictions of pre-existing physical models to the forces exerted by laboratory-scale flows. However, A21 was unable to establish whether the laboratory flows exhibited basal slip, and the conditions for applying its results are therefore unclear. Here, we describe discrete-element simulations that examined the fluctuating forces exerted by steady, downslope-periodic granular flows on fixed, rough bases that prevented basal slip. We show that, in its absence, A21's results do not hold: simulated basal forces' power spectra have high-frequency components more accurately predicted using mean shear rates than using depth-averaged flow velocities, and can have intermediate-frequency components which we relate to chains of prolonged inter-particle contacts. We develop a 'minimal model', which uses a flow's collisional properties to even more accurately predict the high-frequency components, and empirically parametrize this model in terms of mean flow properties. Finally, we demonstrate that the bulk inertial number determines not only the magnitude ratio of rapidly fluctuating and mean forces on a unit basal area, consistent with A21, but also the relative magnitudes of the high and intermediate-frequency force components.Plain Language SummaryAny geophysical granular flow-such as a landslide, rockfall, or debris flow-exerts fluctuating forces that cause the ground to vibrate, in a 'slidequake' that can provide useful information about the flow. Here, we examine simulated slidequakes: computer models of the individual particles within idealized flows, the collisions between them, and the rapidly fluctuating forces they exert on the flow's base. By recording particle and collision properties throughout the flow, we examine pre-existing models for the fluctuating forces; develop, test, and simplify a new model; and relate ratios between forces to an 'inertial number' that characterizes different flows. Our results differ from those of laboratory experiments that previously investigated slidequakes, but the two sets of results can be combined to provide information about real geophysical flows.