|In this manuscript, Tourian et al. produced a new global dataset on water cycle HydroSat. This dataset was compiled using existing satellite data and their derived products. They also estimate the uncertainty of this product. This observation-based product is potentially useful for understanding the water cycle, improving hydrological models, and assessing freshwater availability. |
I think this is an interesting paper and fits into the scope of ESSD. While it has the potential to be published, I have some major concerns on the clarity and novelty and suggest a significant revision on these issues.
The title and abstract are confusing, potentially misleading given the broad readership of ESSD. The water cycle includes several fluxes (Precipitation, ET, surface/subsurface runoff) and stocks (snow, glaciers, lakes and reservoirs, soil moisture, groundwater). The compile dataset only includes estimates for limited components of the water cycle. Why other components are excluded even through relevant satellite observations are available? The spatial and temporal coverage of this dataset should be explicitly mentioned in the abstract.
The novelty of this dataset is unclear to me in the current version. I agree that satellite observations provide a new dimension for understanding the water cycle. Some satellite-based products have been generated. For example, the Hydroweb dataset provides historical and operational water levels for lakes and rivers and GRACE-derived TWS has also been reported in some papers. The authors should highlight why we need a new dataset in the abstract and potentially who are the targeted users in the main text.
The improvements of this new dataset upon existing datasets seem to be unclear. Additionally, the literature is not up-to-late, which also somehow prevents the understanding of the novelty. See my specific comments below.
Line 10: “..act as inputs to hydrological models”. Hydrologic models generally use climate forcing data, terrain and land cover data as inputs. How this dataset can be used as the inputs is not clear to me.
Line 25: not clear about what’s known vs unknown. At which spatial and temporal scales?
global water cycle is a big topic. The current view is not extensive enough. I would recommend incorporate the insights of existing papers on this topic. One example is given below:
Rodell, M., Beaudoing, H. K., L’ecuyer, T. S., Olson, W. S., Famiglietti, J. S., Houser, P. R., ... & Wood, E. F. (2015). The observed state of the water cycle in the early twenty-first century. Journal of Climate, 28(21), 8289-8318.
A review of existing products is missing. Without it, it would be difficult to understand the need of a new product.
Line 105: the difference between SR and HR products is not clear. Please clarify
Line 110: improved temporal resolution compared to what? What’s the exact temporal resolution? Does the dataset cover all lakes and rivers or a subset? If a subset, any filtering steps on lakes and rivers?
Line 258: There are more studies on generating area time series from Landsat, such as
“Yang, K., Yao, F., Wang, J., Luo, J., Shen, Z., Wang, C., Song, C., 2017. Recent dynamics of alpine lakes on the endorheic Changtang Plateau from multi-mission satellite data. J. Hydrol. 552, 633–645.
Yao, F., Wang, J., Wang, C., Crétaux, J.-F., 2019. Constructing long-term high-frequency time series of global lake and reservoir areas using Landsat imagery. Remote Sens. Environ. 232, 111210.”
Line 115: I appreciate the reported validations for individual cases. But a global-scale evaluation makes more sense to me. Have you compared the coverage (spatially and temporally) and accuracy with existing products?
Line 125: same comment as above for the lake products
Table 2: this list does not reflect the up-to-late status. A more comprehensive literature review is required. Just name a few excluded studies:
“Pickens, A. H., Hansen, M. C., Hancher, M., Stehman, S. V., Tyukavina, A., Potapov, P., ... & Sherani, Z. (2020). Mapping and sampling to characterize global inland water dynamics from 1999 to 2018 with full Landsat time-series. Remote Sensing of Environment, 243, 111792.”
Yao, F., Wang, J., Wang, C., Crétaux, J.-F., 2019. Constructing long-term high-frequency time series of global lake and reservoir areas using Landsat imagery. Remote Sens. Environ. 232, 111210.
Zhao, G., Gao, H., 2018. Automatic correction of contaminated images for assessment of reservoir surface area dynamics. Geophys. Res. Lett.
Line 363: The original resolution of GRACE is 3 degree. How you downscaled the resolution should be introduced. Any cautions should be paid when using this downscaled product?
Table 4: same comment as on Table 2. For example,
“Yao, F., Wang, J., Yang, K., Wang, C., Walter, B.A., Crétaux, J.-F., 2018. Lake storage variation on the endorheic Tibetan Plateau and its attribution to climate change since the new millennium. Environ. Res. Lett.”
“Wang, J., Song, C., Reager, J.T., Yao, F., Famiglietti, J.S., Sheng, Y., MacDonald, G.M., Brun, F., Schmied, H.M., Marston, R.A., Wada, Y., 2018. Recent global decline in endorheic basin water storages. Nat. Geosci. 11, 926–932.”
Line 405: I would apologize if I missed anything. How you estimated the uncertainty of the storage anomaly?
The conclusion seems to be abstract. For example, how many lakes and rivers have been covered in the dataset? I expect to see more quantitative summaries and highlights on the improvements upon existing products.