The manuscript proposed new hydrological parameterization for ELM in the Lower Mississippi river basin by constraining ELM predicted water table depth with predictions from a calibrated 3D subsurface hydrology model. Then the manuscript conducted simulations during the period of Hurricane Ida using both default and new parameterization with ELM and compared the simulated hydrological and nitrogen responses to Ida with observations. The study concludes runoff process is important to predict hydrological and nutrient responses to heavy rainfall events.

Overall, the paper is easy to read. However, One thing is unclear to me - what is the key evidence that calibrated ELM is better than default ELM? It seems to me the biggest difference is that default model does not predict changes in WTD but calibrated model predicted shallower WTD after Ida. The conclusion states that "the calibrated ELM was able to simulate the increased nitrate-nitrogen runoff leaching during Hurricane Ida" (Line 472-473) but both versions predict increases in runoff and N Loss (Fig. 5e and Fig. 6a) and it is not discussed which one is closer to the observation. Given this is a model evaluation paper, more direct and quantitative model-data comparison would be helpful

Some additional comments:

Line 195-205. The whole training workflow reads a little confusing to me. WTD is a model ouptut depending on fmax and fdrai but it is used to predict fmax and fdrai... Doesn't it make more sense to create a model/emulator for WTD based on ELM inputs and parameters and then find the best fmax/fdrai to get WTD values close to Tran2024 WTD?

Line 285. For nitrogen and runoff loading response, it seems only one station has significant response but the other one (07381600) did not? Any explanations? In addition, is the model able to explain the decline of Nitrogen and Runoff Loading before Ida?

Line 370. Fig. 6 panel a, if the value is cumulative sum, the unit should just be gN/m2 without per month.

Line 410. There is not panel g in Fig. 8

Line 432-433. What are the "valuable insights"?

The Manuscript developes a new soil hydrology parameterization for ELM for the Lower Mississipi River Basin (LMRB) 

in an effort to better understand nitrate loss. This proposed calibrated model predicts water table depth from a 3D subsurface 

hydrology model. The authors then conducted simulations of the Hurricane Ida of period and compares default ELM and calibrated 

model simulated results with observations. 

I think the manuscript is relatively easy to read and understand and proposes a novel way to improve our understanding of nitrate leaching. However, I struggle to understand which model performs better in the context of the observations. Both versions predict an increase nitrate runoff but having these comparisons with direct observations on the figure would help the reader understand how model performance compared. 

Additional comments

Figure 1 make the USGS points larger to improve clarity

In figure 3f, I understand the overprediction but why are there a bunch of points at just less than 4m?

Comparing the WTD of figure 5C to observations in 4A was difficult and model results and observations seem very different.

line 232 Specify what exactly those insight are

Line 382 Cod"e"

The biggest issue with the paper was understanding whether the default model or the calibrated model performed better.