I would like to clarify at the outset that the decision of “reconsider after major revisions” should not be interpreted as a rejection of the manuscript. At this stage, there is a technical issue for which I would like to receive clarification from the authors before asking the reviewers to undertake another full review. Following the editorial procedure, I therefore need to select “reconsider after major revisions” to allow the authors an opportunity to respond to and address this issue. If the authors' response and corresponding revisions satisfactorily resolve the concern described below, the revised manuscript can then proceed to another round of review.

An issue that remains unclear is the pronounced symmetry of the simulated fields about the domain centerline, y=10 km. In particular, the 13-h time-averaged surface precipitation exhibits this symmetry across the simulations, and a similar symmetry appears in the forward water-vapor trajectories. Given the stated periodic boundary condition in the y-direction, the domain has no physically distinguished location in y; periodicity does not, by itself, select y=10 km as a preferred symmetry axis. Moreover, random perturbations are introduced to initiate turbulence.

While convective organization and cold-pool dynamics can certainly produce spatially heterogeneous and organized precipitation, it remains unclear why these processes would consistently produce reflection symmetry specifically about y=10 km across the simulations and over a 13-h averaging period. The authors should explain what physical or numerical mechanism selects the domain centerline.

Please also clarify the implementation of the y-direction boundary condition and whether any additional treatment (e.g., damping, relaxation, sponge layers, or other boundary-related forcing) is applied. In addition, the forward trajectories that reach a y-boundary appear to terminate rather than re-enter through the opposite boundary. Please explain how periodic boundary crossings are handled in the trajectory calculation and whether this behavior is solely a trajectory post-processing issue or reflects the boundary treatment used in the LES.

Please revise the manuscript carefully in response to the reviewers’ comments and suggestions. Upon successful revision, I believe the manuscript will be ready for publication.

The manuscript will be accepted once the following point is addressed.

In the statement, “we use the base-state and perturbation fields directly output by CM1. A model variable can be expressed as phi(t,z,x,y) = phi_0(t,z,x,y) + phi_{pert}(t,z,x,y), ...,” the authors should explain more clearly how the base-state and perturbation fields are defined in CM1. In particular, they should describe what constitutes the base state in the model and how the perturbation field is computed relative to that base state. The relationship among the total, base-state, and perturbation fields should be specified clearly enough that the decomposition can be understood from the manuscript without relying on knowledge of CM1's internal implementation.

The functional dependence of the base-state field should also be verified and reflected correctly in the equation. If the base state is independent of time, t should not appear as an argument of phi_0. Likewise, given that the LES is initialized with conditions homogeneous in the y direction, the authors should clarify whether the base state is also independent of y, and, more generally, which coordinates the base state actually depends on. The notation should be revised accordingly.

I also recommend using the more conventional ordering (x,y,z,t), rather than (t,z,x,y), throughout this discussion.