Ching et al. (2014) describe the inability for high resolution models to reliably simulate convectively induced secondary circulations (CISCs; e.g., convective rolls) owing to the disparity between the smaller spatial scale related to the CISCs and the model’s larger grid length.