The paper details a study of preliminary implementation of lateral groundwater flow in the land component of the E3SM model. The goal of the model developments is to allow for subsurface flow along hillslopes to rivers. The formulation of the model allows for water table depth variation such that the slope of the groundwater table may vary in time in response to water infiltration across the land model grid cell and the elevation of water in the connecting river channel segment. Figure 9 of the manuscript highlights the large discrepancy between the default (current) model treatment of groundwater elevations and the observed spatial variability of water table along a hillslope. The ability to more realistically simulation water table depths and seasonally varying lateral groundwater fluxes has the potential to significantly improve the representation of the terrestrial hydrological dynamics in the E3SM model.

The manuscript is well written and clear. I do not have major revisions to request, however, I believe the paper’s value could be enhanced with additional details regarding the observed differences in LGFs between some of the cases. For example, case 9 has a high LGF raising the elevation of the downslope boundary would intuitively seem to reduce the gradient of the WTP. However, Figures 9 and 10 show that the water rose significantly and apparently steepened. Is the higher LGF only in response to the increase in water table slope or also due to the thickness of the zone discharge at the river or due to more flow in the shallower soil layers which may have higher Ksat? Is there a strong transient response in this case for the WT to rise?

The plotted slope for case 6 on Figures 9 and 10 is confusing. The plotted slope looks much steeper than all other slopes, but the case is supposed to be the surface slope. On Table 4 the slope is listed as 0.07 which is identical for the other cases.  

Is it possible from the analysis to quantify how including LGF would impact river discharges across a basin seasonally? Or is the most significant potential change related to the discussion of soil moisture and hydrological feedbacks on plants.

Is there seasonal observational data that can be shown as well? Or is the actual groundwater table at the modeled site truly time invariant. If so, why is the model seasonally responsive but the actual system is not? Can the authors provide any discussion on why all of the modeled cases have WTD much greater than the observed? Is this related to the selection of Ksat and/or the vertical infiltration values?

Minor comments:

Line 91: WTD not defined yet.

Line 291: “Case 6 produced the largest WTD (  19.9m) due to its large LGFs (caused by high WTG, 0.07)I”  Can more detail be provided? The WTG for case 6 is the same for 3, 7, 8, 9, and 10.

Figure 10: the title at the top says February but the results are for August.