I am interested in your great work on snow phenology analysis. I have several questions that need your clarification.

• According to numerous previous works and my analysis (https://doi.org/10.1016/j.isprsjprs.2024.07.018), the uncertainty in snow phenology is mainly derived from the accuracy of daily snow cover products. You always mentioned there are potential bias when using the fixed threshold to determine snow phenology. Through your whole text, you didn’t specifically illustrate this type of uncertainties. May I ask you to give some examples to clarify this uncertainty what you mentioned?
  
  ---In the Abstract, “Previous studies commonly employed … leading to potential biases of snow phenology.” This description is too general.
  
  --- Lines 65-66, “The fixed threshold for snow phenology fails to account for the variations in snow cover across the NH”. Please give us more explanation on this. Do you have any related references to support your statement? Additionally, the following sentence “In fact, snow cover increase …. Especially on the TP” cannot support your statement on the uncertainties due to using fixed threshold.
  
  --- Lines 68-69: “Snow conditions are variable, … underestimation or overestimation of snow phenology.” Please add more explanations on this underestimation and overestimation. Is there any published works to support this statement?
• Line 14-15: “At low and middle latitudes, the snow cover duration (SCD) extends, the snow cover onset day (SCOD) advances, and the snow cover end day (SCED) delays, …” This conclusion is quite different from what we got up to now. Does your analysis conclusion tell us we have more snow in this region in a warming world?
• MOD10C2 products provide a maximum snow cover extent during this 8-days period. That means you would potentially overestimate the snow cover phenology metrics (snow cover days, snow onset date and snow end date) when you used this data. “For the SCF dataset, … after the last identified snow cover.” (Line 112-113) Please cite references to support this processing’s reasonability.
• 1: which part of data do you use to plot this figure? 1989-2018 or 2000-2018? Single data and which data? Additionally, this curve is like for vegetation growth instead of snow cover evolution. Please check it again. The fact is that Summer is no snow (DOY=180 days). Finally, how can I understand the term “percentage of snow depth”? For your reference, here is a in-situ observation of snow evolution within a snow cover year (10/1 – 9/31) https://www.climatehubs.usda.gov/hubs/northwest/topic/30-year-normals. Regarding only snow depth variable used in your method, I am confused how you use snow cover products (IMS and SCF) to calculate snow phenology metrics.
• Through your manuscript, you used a 5% or 10% to determine snow phenology metrics. Is this another type of “fixed threshold” method. Additionally, you didn’t include some comparisons, validations, and evaluations. How do you convince our readers of the responsibility (not accuracy) of your snow phenology results? Because your method quite largely different the method we usually used.
• Based on your analysis, I wonder if your approach means that snow phenology metrics will vary depending on the study area used. For example, the snow depth in the Northern-Xinjiang is generally higher, while it is lower in the TP. ---Case1: Only TP data is used to analyze snow phenology metrics for TP. ---Case2: The data both in the TP and Northern-Xinjiang are used to analyze snow phenology metrics for TP. Are the snow phenology metrics in the TP region different between Case 1 and Case 2?

Minor comments:

Change “on” to “in” in line 67.

Change “24 hours” to “daily” in line 95

Change “the hydrological year” to “a hydrological year” in line 111

How do I understand the “snow index” in Line 132? Snow depth? Snow cover days? NDSI?

Line 101-102: “This data has been validated against … less than 5 cm account for approximately 65% of all the data” Please add the reference for this statement to support this “5cm - 65%” number pair.

Line 141: Why do you use “30-day” moving window? Any specific reasons?

Section 2.1: how did you handle the different spatial resolution of these snow products?

Line 246-247: “Given that the zonal variations … compared to a fixed threshold.” It is not clear.

Are the days in Fig 2 and Table 2 DOYs (Day of Year)? If it is, 66 of SCOD in table 2 means the snow starts in March. Please check the whole paper’s figures again.

The authors introduce a method of defining start dates, end dates, and durations of snow cover over Northern Hemisphere lands that differs from fixed definitions for these variables.  This reviewer is left with considerable uncertainties as to how the author's dynamic method is any more useful than the fixed method.  One major concern has to do with the use of microwave derived snow depth data to "drive" this approach.  Many uncertainties remain to this day regarding the accuracy of this spectral region to accurately estimate depth.  This can especially be true at shallow depths, with wet snow, and with deep snow that may have ice lenses, depth hoar, etc. within the pack.  Unless the authors can explain how these issues do not impact their dynamic phenology method, I would place greater faith in the IMS and SCF estimates.  This brings into question just how specific one can get with timing using an 8-day window for the SCF.  Not that the IMS is a full-hemisphere daily evaluation of snow cover given persistent cloud cover that may mask surface conditions for multiple days.    These issues are mainly more challenging in the fall for all sources, in part due to greater cloudiness then than in the spring melt season.  Also, due to shallower depths, potential wet conditions, and ephemeral snow cover that can be found everywhere during seasonal snow onset (not that it doesn't exist throughout the season at lower elevation, low-mid latitude regions.  Also, the challenges of using microwave data to map snow cover over the Tibetan Plateau has long been recognized, in part, as the authors suggest, due to its shallow depth (over non-mountainous areas) and ephemeral nature.

I will go no further with this evaluation as  I see earlier reviewers have posted excellent comments regarding manuscript specifics, to which I agree.  At this point, I will conclude that while the manuscript addresses an interesting topic and employs the key snow cover extent data products at regional to continental scales (IMS and MODIS), I am not convinced that the SD data and the subsequent conclusions built upon it are fully supported.