1. In section 2.3, it is important to note that the Y-axis in figure 2b represents V(T) and not f(TV), and the equation for the blue line is Wang 2017, not Wang-Engel (1998). Additionally, the plots in figure 2 should be labeled as a and b, and the Y-axis labels should match the text. Furthermore, the manuscript's novelty is the new heat stress function, but it is not adequately explained. Therefore, it is essential to provide a detailed scientific explanation of the function to help readers better understand it.
2. In Figure 3, it is recommended to include subplot numbers (a, b, c) to avoid confusion.
3. In line 190, it should be noted that "O" should not be capitalized in the MODIS abbreviation.
4. In line 206, the CONUS abbreviation's full form is incorrect; it should be "Conterminous" instead of "CONtiguous."
5. In line 219, it is necessary to clarify that there is no equation 7 in the manuscript and refer the readers to the comments in section 2.3.
6. In line 230, the sentence is ambiguous, and it is unclear what the author means. According to the given numbers, the model seems to have overestimated the yields. Therefore, it is recommended to rephrase the sentence to make it clear that the model overestimated the yields.
7. In section 3.1, it is essential to specify whether any other parameter tweaking was done during the single-point calibration and validation, besides GDD and Tavg for planting.
8. In Figure 6, the plot is overcrowded, and the colorbar is too small. To improve visibility, it is suggested to stack the three separate plots vertically and include a bigger horizontal colorbar. Furthermore, the description of Figure 6 needs improvement.
9. In Figure 9, it would be clearer to separately plot the default and temperature stress results. The figure will now have three rows. Also, due to the marker type, it is challenging to see the improvement during the temperature stress function.

• It is necessary to provide a good scientific explanation of the P90 threshold and the temperature stress function being added.
• In line 326 with the temperature stress simulation, the sentence claims that the stressed yield results correspond to heatwave events in 2002, 2006, and 2012, which mostly agree with the observations. However, it is not evident in Figure 10 that the temperature-stressed simulated yield matches the observations by comparing the overall trend. For instance, in 2012, the default yield is much closer to the observed yields. Therefore, it is recommended to provide more clarity and detail.
• In line 397, “For example, Siebert et al. (2014) claimed that the differences between applying canopy temperature and air temperature in crop models under heat stress simulations.” It is unclear what is being said about the differences between applying canopy temperature and air temperature in crop models under heat stress simulations. Therefore, the sentence needs to be rephrased to provide more clarity.
• In line 437, it is necessary to mention whether the conclusion is for the region being studied. If the answer is yes, it is recommended to include this information in the sentence.

Crop growth and yield simulations in land surface models are crucial for both crop yield projections and the land-atmosphere interactions. Noah-MP is the latest generation of Noah land model. The authors added spring wheat growth (including dynamic planting/harvest and the temperature stress) into Noah-MP which improved its capabilities in application over the area where the spring wheat is the major crop. The manuscript is well written and there are only minor comments before considering for publication.

Minor Comments:

Even though the authors mentioned that the 10 ^oC is based on a global synthesis of planting and harvest dates, the thresholds in equation 3 and 4 are quite arbitrary. I suggest to do some sensitivity analysis by adjusting the thresholds and show the current thresholds are reasonable for the region.

For the regional simulation, the authors used 4-km resolution atmosphere forcing data. Could you prove the 4-km resolution yield a better LAI and yield simulation than other coarse resolution forcing, such as CRUNCEP? I’m curious for the regional averaged crop LAI and yield, will the coarse forcing showed a similar result as the 4-km high resolution.

In figure 5, adding the spring wheat model improved LH, but showed poor soil moisture simulations than the default model. Please comment on how to fix this problem in Noah-MP.