This manuscript represents a significant contribution in the form of an extensive application ready lake clarity data product covering small and large lakes of China over a period of 35 years.  This data product, LAWSD30 (and associated derivations), will be a valuable asset to future endeavours that seek to understand the drivers of lake clarity change in China.

Specific comments:

1. Include the normalised RMSE in addition to the RMSE; the normalised RMSE summarises the explanatory power of a model in a manner that is independent of scale, and so enables a ready comparison across modelling efforts.

2. Line 118 : replace 'product' with 'application ready data product'.

3. Lines 147 - 48: it is not clear how 'user' and 'producer' map to the terminology used in Peckel et al 2016. Please clarify or re-word.

4. Line 211: replace 'Following him' with 'Following Chen et al. (2021)', assuming that is what is meant?

5. Line 215: for the convenience of the reader, please briefly describe the constituent BAP criteria and how the BAP was calculated, rather than pointing to Chen et al. 2021 (which then on-refers the reader to White et al. 2014 for further details).  

6. Line 227: please replace the word 'proven' with 'demonstrated', here and elsewhere in the document.  Likewise, replace 'proves' with 'demonstrates'.  'Proven' and 'proves' should be reserved for contexts where there is little room for interpretation.

7. Line 316: replace ‘product’ with 'LAWSD30 data set'.

8. Line 339: replace 'researches' with ‘studies’  (here and elsewhere in the manuscript).

9. Line 348: replace 'Similarly, some studies found the same SD ....' with 'Other studies found the same SD ...'.

10. Line 462: replace 'hydrometeor' with 'weather'.  (hydrometeor has a different meaning to that which I assume is intended)

11. Lines 523 - 24: replace 'to reflect the comprehensive conditions in water bodies' with 'of water quality'

12. Conclusion: include commentary that alerts the reader to the notion that an informed interpretation of the lake clarity patterns presented in this manuscript will require an understanding of the relative importance of local and regional driving forces; and that such an informed interpretation will be required in order to prioritise management actions. 

For example, the work of Yu and Zhai (https://doi.org/10.1038/s41598-020-71312-3) demonstrate that compound drought and heat events have impacted China more severly since ca. the year 2000.  Have such events led to changes in SD patterns in some lakes, for example by making wind-driven resuspension more or less likely? 

This paper developed a 30 m long-term Lake Water Secchi Depth (SD) dataset (LAWSD30) of China (1985–2020) using the robust water-color-parameter-based SD model. The LAWSD30 dataset can be used to study the temporal and spatial changes of Lake SD. Therefore, it is of great significance to the study of long-term trend of Lake SD and lake ecological environment management.

Specific comments

1. The RMSE was used for evaluating the accuracy of the model. However, in my opinion, this index is not very appropriate. For example, when the real value is equal to 2 or 20, although the RMSE value is 0.2, the accuracy of the model is very different. Thus, this index does not show how close the real value is to the estimated value I suggest using MAPE or similar indicators to evaluate the accuracy of the model.
2. Line 125-126. “The summer months were chosen because the water clarity is relatively stable in this season and suitable for monitoring with remote sensing imagery”. I understand that calculating SD in the same season can enhance the comparability of data, but I don’t think the clarity is relatively stable in the summer, because heavy rainfall and algal bloom often occur in summer, resulting in the change of suspended solids and therefore affecting the SD.
3. Line 163-164. The collected SD measurements were within seven days of satellite overpasses. I suggest the meteorological conditions should be considered since both heavy rain and strong wind could affect the SD.
4. Figure 4. When SD is less than 2, the covariance relationship between in-situ SD and LAWSD30 is very weak because many data are vertical lines. The accuracy should be analyzed in this situation.
5. Line 332. Why Selinco Lake and Hongze Lake were chosen to illustrate the LAWSD30?
6. Figure 10. The figures present the different SD trends of lakes with area<=1 km² and >1km², in different region of China. The five regions have different socio-economic, geological and climatic conditions, should the driver factors of SD changes be further explained?