the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 14 Jul 2023
A Global Lake/Reservoir Surface Extent Dataset (GLRSED): An integration of HydroLAKES, GRanD and OpenStreetMap
Abstract. Global lake/reservoir surface water extent is the basic input data for many studies. Although there are some datasets at present, there are problems such as incomplete or spatial inconsistency exist among them due to various reasons like different data sources and dynamic change characteristics of the surface water. In this paper, a new Global Lake/Reservoir Surface Extent Dataset (GLRSED) that contains spatial extent and basic attributes (e.g., name, area, lake type and source) of 2.17 million lakes/reservoirs was produced based on HydroLAKES, GRanD and OpenStreetMap. In addition, by overlaying with mountain data, we identified the lakes/reservoirs located in mountain areas. By overlaying with the Global geReferenced Database of Dams (GOODD) and Georeferenced Global Dams and Reserves (GeoDAR) dataset, we partitioned human-managed reservoirs from natural lakes. Lakes/reservoirs on the rivers were identified by overlaying with the SWOT Mission River Database (SWORD). Using the same method, we identified endorheic, glacier-fed and permafrost-fed lakes. Furthermore, the coverage of Surface Water and Ocean Topography (SWOT) ground track to each lake/reservoir in GLRSED was calculated to explore the potential of SWOT for monitoring lakes. These datasets could provide basic data for global lake/reservoir monitoring, enabling the study on the impact of human actions and climate changes on lake/reservoir freshwater availability. The GLRSED database is available at https://doi.org/10.5281/zenodo.8121174 (Bai et al., under review, 2023).
- Preprint
(1726 KB) - Metadata XML
-
Supplement
(378 KB) - BibTeX
- EndNote
Status: closed
-
RC1: 'Comment on essd-2023-216', Anonymous Referee #1, 29 Aug 2023
Review of “A Global Lake/Reservoir Surface Extent Dataset (GLRSED): An integration of HydroLAKES, GRanD and OpenStreetMap”
In this work the authors combine multiple hydrological datasets to improve the spatial representation of lakes and reservoirs as well as linking multiple important physical parameters to each water body. This work provide an important product for further inland water research and could specifically benefit from correction and integration with the upcoming SWAT measurements from the US National Aeronautics and Space Administration (NASA) and the French space agency Centre National d’Etudes Spatiales (CNES). I have some points with this manuscript which are listed hereunder, in general more details and clarity is required.
This work construct a Global Lake/Reservoir Surface Extent Dataset (GLRSED) by combining three main datasets (HydroLAKES, GRanD and OpenStreatMap) through a processing step followed by an integration of spatial extent and physical characteristics. Thereafter additional auxiliary data is derived through combination of GLRSED with additional spatial datasets (Mountains, permafrost, glaciers, endorheic catchments, etc.).
As it is now it is not clear how these steps were preformed. It can be inferred from the manuscript, that the spatial integration (spatial overlapping) was performed by finding the largest spatial extent for an arbitrary lake/reservoir present in the main datasets above. This process with each step need to be clearer described in the text. Furthermore Fig. 1 should be split into two figures and additional clarity and information should be provided. The first figure should show the exact working path used for constructing the GLRSED spatial dataset. Including the step where rivers was removed and how spatial interaction was treated, for example was a water body partly or fully removed when it was overlapped by a river segment? The second figure should show the exact derivation of lake type data from the auxiliary datasets including distances used (example the 1 km used for glacier interaction). The spatial overlapping clarification should also be present in the construction of the auxiliary datasets, it should be clear how lakes completely or partially within the set limit was classified.
As for the permafrost-, glacier fed lakes and endorheic lakes, the evidence provided in the manuscript is to general to tag these water bodies as such. In the text the authors classify a lake to be glacier fed if it is within (not clear exact how, see above) 1 km of glaciers, similar for endorheic and permafrost fed lakes. The evidence is enough to say these lakes/reservoirs are located in these regions, in fact the authors acknowledge this act in figure 6, but the details is not enough to classify how a lake or reservoir is fed.
I am missing lake/reservoir depth as a variable, which might be outside the scope of this study. But this work would benefit from the inclusion of this parameter. I leave it up to the authors to decide if they want to include this from the sources below and/or other works.
Choulga, M., Kourzeneva, E., Zakharova, E., and Doganovsky, A.: Estimation of the mean depth of boreal lakes for use in numerical weather prediction and climate modelling, Tellus A, 66, 21295, https://doi.org/10.3402/tellusa.v66.21295, 2014.
Lehner, B. and Döll, P.: Development and validation of a global database of lakes, reservoirs and wetlands, J. Hydrol., 296, 1–22, https://doi.org/10.1016/j.jhydrol.2004.03.028, 2004.
Toptunova, O., Choulga, M., and Kurzeneva, E.: Status and progress in global lake database developments, Adv. Sci. Res., 16, 57–61, https://doi.org/10.5194/asr-16-57-2019, 2019
Short points:
Line 90: add reference for SWOT data (is it SWORD?) and the revisit cycle (21 days?).
Line 91: The reader needs to know how a mountain is defined.
Line 98-99: Why did you use different search words in China and nowhere else? This should be uniform to avoid bias.
Line 130 to 155: Combine and/or extend paragraphs
Fig. 1: redo as described
Fig. 2: Add satellite image as in Fig. 9
Fig. 3: Split into two frames, one with lakes and one with reservoirs
Fig. 4: Normalize lake count and area towards country area, and use the real spatial extent of all countries.
Fig: 6: Simplify legend
Table 2 and S2: Theses tables need more detail, it is hard to follow what they show. Furthermore make sure that Table 2 and Figure. 4a correspond to each other, it looks like Italy shouldn’t be included in the table.
Citation: https://doi.org/10.5194/essd-2023-216-RC1 - AC1: 'Reply on RC1', B. X. Bai, 25 Sep 2023
-
CC1: 'Comment on essd-2023-216', Yaohui Liu, 01 Oct 2023
The manuscript "A Global Lake/Reservoir Surface Extent Dataset (GLRSED): An integration of HydroLAKES, GRanD and OpenStreetMap" have generated an useful data. In particular, the authors have done a great job that adding additional attributes to the data seems to be beneficial for many scientific studies. However, I find the manuscript suffers from pervasive grammatical errors that negatively impact the readability.
In addition, some specific comments may improve the manuscript:
- The author sometimes use “lakes”, sometimes “reservoirs”, sometimes use “lakes/reservoirs”, “lake/reservoirs” or “lake/reservoir”, which feels very chaotic. Like Line 23、25、37、166, could you please clarify?
- The Table S1 seems important, maybe move it from the supplement to main body better.
- Show the continents of area and count together and in order in Table 2.
- Figure 9 (a) is already shown in Figure 2. It is recommended to remove this duplicate information or delete Figure 2.
- Is the reservoir information contained in the OSM field used for setting the reservoir attributes of the new data? If not, it is recommended to integrate this information.
- Explain the distribution anomalies like OSM data in Italy in the manuscript.
Citation: https://doi.org/10.5194/essd-2023-216-CC1 -
RC2: 'Comment on essd-2023-216', Anonymous Referee #2, 06 Oct 2023
In the manuscript with the ID “essd-2023-216”, the authors report a new lake/reservoir dataset, Global Lake/Reservoir Surface Extent Dataset (GLRSED), containing the spatial extent and basic attributes of 2.17 million lake/reservoir by simply merging multiple open-access datasets including HydroLAKES, GRanD and OpenStreetMap, GOODD, and other attribute data. The types of lake/reservoir were identified by overlaying with ancillary data. In addition, the potential of SWOT for monitoring lake/reservoir was explored by intersecting the designed orbit tracks. Though the study is designed with a right motivation to provide a base data for global lake and reservoir monitoring, many fundamental weaknesses existing in the data production and lack of necessary descriptions on the data-preprocessing flows and resulting data attributes prevent the manuscript from being recommended for publication in ESSD.
1-The authors emphasize the drawbacks of existing open-access datasets, including HydroLAKES, OSM, JRC GSW, etc., overcoming these disadvantages is thus necessary. A rough collection of these datasets may solve a few problems, for example, generating a fuller spatial distribution of waterbodies (lake and reservoirs), less omission errors, or a more spatial representation of reservoir count and area. However, a simple merge of these datasets cannot generate a convincible and valuable new dataset. Instead, it may induce more commission errors and larger spatial and temporal inconsistency from different data sources, which greatly weaken the application value of the new dataset. For instance, the vector data of HydroLAKES were derived from multiple sources (e.g., SWBD, Canadian national dataset, MODIS images and so on) but have explicit descriptions on the original places for each polygon. Similarly, the OSM data set have multiple sources (satellite images, manual delineation and others) which represent inconsistent cartographic time periods. The merge of these lake/reservoir polygons make the data source and the time period of water boundary more complicate to search.
2-Attributing the lakes from different features (mountain lake, endorheic lake, or permafrost-fed, etc.) is an important thing to be done. Yet, the processing rules as presented in this study cannot generate high-value and robust attribution data. More specifically, it is not reasonable to define the types of glacier-fed and permafrost-fed lakes merely relying on the calculating the buffer distance of 1km around the glacier or permafrost extent. For example, the downstream lakes with supply of permafrost/glacier melt water may be classified as the opposite group, particularly for the dense distribution of arctic lakes. The processing of mountain lakes and endorheic lakes by overlaying with the mountain or basin boundary cannot produce any added value for audiences. Instead, many key attributes in the original datasets, such as the storage, reservoir function, water depth, are missing in the new dataset.
3-The authors mentioned that “there are also geometric differences in features between the different datasets”. In this study, a new lake/reservoir dataset, GLRSED, was generated based on HydroLAKES, GRanD and OpenStreetMap; however, how were the differences in spatial boundaries and temporal inconsistencies among the three datasets solved? Especially, there could be some situations in the spatial intersection of multiple polygons from different sources. Generating a uniform set of water body boundaries would be more convincing.
4-The methodology section is not described in enough detail. The flowchart in Figure 1 is not clear. As introduced in Section 3, the final processed OSM contains data on a total of approximately 0.85 million lakes. How many lakes does OSM contain in draft version? How many reservoirs does OSM contain? How to tackle with the overlapping issues with HydroLAKES data. Moreover, are the “cuo” and “pond” water bodies classified as lakes or reservoirs? Are there lakes with other terminology rules missed in the initial filtering? None of these issues were mentioned.
5-Both the HydroLAKES and GRanD datasets provide the attribute of lake and reservoir capacity. By reviewing the new dataset GLRSED, lake/reservoir from either the HydroLAKES or GRanD datasets account for more than half of the total number (1,412,011), and the authors may consider adding storage capacity information for this portion of the lake/reservoir.
6-The results and discussion section are too thin and still needs further improvement, for example, the dataset description (4.2 Attribute Table of the GLRSED). The basic information of the dataset, such as the total number of lakes and reservoirs, the total area, and the number of lakes and reservoirs, respectively, and the total area, should be described with more spatial and grouping details, despite that it is a data paper.
Minor comments:
Line 12-15: Where are the data source of the most component of reservoir type identification for the GLRSED, from GRanD, GOODD or GeoDAR?
Line 32-33: How about the definition of lake and reservoir expansion and the processing of inconsistence of their spatial boundaries in this study?
Line 41: There is one more “period” mark.
Line 44-48: All water mappings cannot avoid such issues, including the datasets used in this study. The problems are overcome to some degree in the JRC GSW dataset.
Line 68: The superscript number for the area unit.
Line 73-79: What does the time period of lake data from OMS represent?
Line 88-89: The literature on SWOT should be added.
Line 91: What exactly does “The mountain data” refer to?
Line 110-112: The details should be added, e.g., how to treat the inconsistency and commission errors?
Line 115: Please explain in detail why the authors used a 270 m buffer to eliminate the impact of reservoir position deviations.
Line 117: The simple definition of buffer zone distance is not sound in generating data product although it was used in prior study.
Lines 123-124: How the number of orbits passing on each lake is counted, and how many points fallings into the lake are considered valid orbits.
Figures 3-5: It is recommended that the lake/reservoir pattern be shown separately.
Line 143-155: Merging and reorganizing the sections 4.4 & 4.5 could make more sense for lake/reservoir type and other attribute identification.
Line 156-174: For dataset comparison, GLRSED is obtained by fusing HydroLAKES, GRanD and OSM, where most of the lakes and reservoirs are from HydroLAKES and OSM. Therefore, there is no need to compare GLRSED with HydroLAKES and OSM. Instead, it should be compared with other independent data sets.
Line 168: Section 4.6 mentions validation with remotely sensed imagery. Is this a direct overlay image validation? Why not consider taking samples of lakes and reservoirs or other datasets (GLAKES or continental/national-scale lake products well inventoried, e.g., US, Europe, China) for careful manual inspection and accuracy evaluation?
Line 190-201: There are no sufficient quantitative information on the dataset description, including the total number of lakes and reservoirs and other type attributes in the new dataset.
SI-Line 5: The inconsistent spatial boundaries are far limited to this case in Figure S1. How about the cases of one polygon in Data-A containing multiple polygons in Data-B, or multiple polygons in Data-A intersecting multiple polygons in Data-B???
SI-Line 10: It is recommended that lakes and reservoirs from OSM be shown in different colors in Figure S2.
SI-Line 15: There are multiple definitions of mountain areas on the earth, corresponding to multiple data products. Why is this one (Korner et al., 2017) chosen in this study, and how to handle the spatial relationship due to inconsistent spatial scales (boundary square sizes)? Besides, how to process the reservoir identification when there are discrepancies among these three datasets?
SI-Line 25: Meaningless table! BTW, how to calculate the areas of lakes that are located in transboundary areas of two or more nations, e.g., the Great Lakes, Lake Victoria, Caspian Sea, etc.?
The language expression in the manuscript still needs to be improved, especially the method and result section, and further improvement is suggested to clarify the method logic.
Citation: https://doi.org/10.5194/essd-2023-216-RC2
Status: closed
-
RC1: 'Comment on essd-2023-216', Anonymous Referee #1, 29 Aug 2023
Review of “A Global Lake/Reservoir Surface Extent Dataset (GLRSED): An integration of HydroLAKES, GRanD and OpenStreetMap”
In this work the authors combine multiple hydrological datasets to improve the spatial representation of lakes and reservoirs as well as linking multiple important physical parameters to each water body. This work provide an important product for further inland water research and could specifically benefit from correction and integration with the upcoming SWAT measurements from the US National Aeronautics and Space Administration (NASA) and the French space agency Centre National d’Etudes Spatiales (CNES). I have some points with this manuscript which are listed hereunder, in general more details and clarity is required.
This work construct a Global Lake/Reservoir Surface Extent Dataset (GLRSED) by combining three main datasets (HydroLAKES, GRanD and OpenStreatMap) through a processing step followed by an integration of spatial extent and physical characteristics. Thereafter additional auxiliary data is derived through combination of GLRSED with additional spatial datasets (Mountains, permafrost, glaciers, endorheic catchments, etc.).
As it is now it is not clear how these steps were preformed. It can be inferred from the manuscript, that the spatial integration (spatial overlapping) was performed by finding the largest spatial extent for an arbitrary lake/reservoir present in the main datasets above. This process with each step need to be clearer described in the text. Furthermore Fig. 1 should be split into two figures and additional clarity and information should be provided. The first figure should show the exact working path used for constructing the GLRSED spatial dataset. Including the step where rivers was removed and how spatial interaction was treated, for example was a water body partly or fully removed when it was overlapped by a river segment? The second figure should show the exact derivation of lake type data from the auxiliary datasets including distances used (example the 1 km used for glacier interaction). The spatial overlapping clarification should also be present in the construction of the auxiliary datasets, it should be clear how lakes completely or partially within the set limit was classified.
As for the permafrost-, glacier fed lakes and endorheic lakes, the evidence provided in the manuscript is to general to tag these water bodies as such. In the text the authors classify a lake to be glacier fed if it is within (not clear exact how, see above) 1 km of glaciers, similar for endorheic and permafrost fed lakes. The evidence is enough to say these lakes/reservoirs are located in these regions, in fact the authors acknowledge this act in figure 6, but the details is not enough to classify how a lake or reservoir is fed.
I am missing lake/reservoir depth as a variable, which might be outside the scope of this study. But this work would benefit from the inclusion of this parameter. I leave it up to the authors to decide if they want to include this from the sources below and/or other works.
Choulga, M., Kourzeneva, E., Zakharova, E., and Doganovsky, A.: Estimation of the mean depth of boreal lakes for use in numerical weather prediction and climate modelling, Tellus A, 66, 21295, https://doi.org/10.3402/tellusa.v66.21295, 2014.
Lehner, B. and Döll, P.: Development and validation of a global database of lakes, reservoirs and wetlands, J. Hydrol., 296, 1–22, https://doi.org/10.1016/j.jhydrol.2004.03.028, 2004.
Toptunova, O., Choulga, M., and Kurzeneva, E.: Status and progress in global lake database developments, Adv. Sci. Res., 16, 57–61, https://doi.org/10.5194/asr-16-57-2019, 2019
Short points:
Line 90: add reference for SWOT data (is it SWORD?) and the revisit cycle (21 days?).
Line 91: The reader needs to know how a mountain is defined.
Line 98-99: Why did you use different search words in China and nowhere else? This should be uniform to avoid bias.
Line 130 to 155: Combine and/or extend paragraphs
Fig. 1: redo as described
Fig. 2: Add satellite image as in Fig. 9
Fig. 3: Split into two frames, one with lakes and one with reservoirs
Fig. 4: Normalize lake count and area towards country area, and use the real spatial extent of all countries.
Fig: 6: Simplify legend
Table 2 and S2: Theses tables need more detail, it is hard to follow what they show. Furthermore make sure that Table 2 and Figure. 4a correspond to each other, it looks like Italy shouldn’t be included in the table.
Citation: https://doi.org/10.5194/essd-2023-216-RC1 - AC1: 'Reply on RC1', B. X. Bai, 25 Sep 2023
-
CC1: 'Comment on essd-2023-216', Yaohui Liu, 01 Oct 2023
The manuscript "A Global Lake/Reservoir Surface Extent Dataset (GLRSED): An integration of HydroLAKES, GRanD and OpenStreetMap" have generated an useful data. In particular, the authors have done a great job that adding additional attributes to the data seems to be beneficial for many scientific studies. However, I find the manuscript suffers from pervasive grammatical errors that negatively impact the readability.
In addition, some specific comments may improve the manuscript:
- The author sometimes use “lakes”, sometimes “reservoirs”, sometimes use “lakes/reservoirs”, “lake/reservoirs” or “lake/reservoir”, which feels very chaotic. Like Line 23、25、37、166, could you please clarify?
- The Table S1 seems important, maybe move it from the supplement to main body better.
- Show the continents of area and count together and in order in Table 2.
- Figure 9 (a) is already shown in Figure 2. It is recommended to remove this duplicate information or delete Figure 2.
- Is the reservoir information contained in the OSM field used for setting the reservoir attributes of the new data? If not, it is recommended to integrate this information.
- Explain the distribution anomalies like OSM data in Italy in the manuscript.
Citation: https://doi.org/10.5194/essd-2023-216-CC1 -
RC2: 'Comment on essd-2023-216', Anonymous Referee #2, 06 Oct 2023
In the manuscript with the ID “essd-2023-216”, the authors report a new lake/reservoir dataset, Global Lake/Reservoir Surface Extent Dataset (GLRSED), containing the spatial extent and basic attributes of 2.17 million lake/reservoir by simply merging multiple open-access datasets including HydroLAKES, GRanD and OpenStreetMap, GOODD, and other attribute data. The types of lake/reservoir were identified by overlaying with ancillary data. In addition, the potential of SWOT for monitoring lake/reservoir was explored by intersecting the designed orbit tracks. Though the study is designed with a right motivation to provide a base data for global lake and reservoir monitoring, many fundamental weaknesses existing in the data production and lack of necessary descriptions on the data-preprocessing flows and resulting data attributes prevent the manuscript from being recommended for publication in ESSD.
1-The authors emphasize the drawbacks of existing open-access datasets, including HydroLAKES, OSM, JRC GSW, etc., overcoming these disadvantages is thus necessary. A rough collection of these datasets may solve a few problems, for example, generating a fuller spatial distribution of waterbodies (lake and reservoirs), less omission errors, or a more spatial representation of reservoir count and area. However, a simple merge of these datasets cannot generate a convincible and valuable new dataset. Instead, it may induce more commission errors and larger spatial and temporal inconsistency from different data sources, which greatly weaken the application value of the new dataset. For instance, the vector data of HydroLAKES were derived from multiple sources (e.g., SWBD, Canadian national dataset, MODIS images and so on) but have explicit descriptions on the original places for each polygon. Similarly, the OSM data set have multiple sources (satellite images, manual delineation and others) which represent inconsistent cartographic time periods. The merge of these lake/reservoir polygons make the data source and the time period of water boundary more complicate to search.
2-Attributing the lakes from different features (mountain lake, endorheic lake, or permafrost-fed, etc.) is an important thing to be done. Yet, the processing rules as presented in this study cannot generate high-value and robust attribution data. More specifically, it is not reasonable to define the types of glacier-fed and permafrost-fed lakes merely relying on the calculating the buffer distance of 1km around the glacier or permafrost extent. For example, the downstream lakes with supply of permafrost/glacier melt water may be classified as the opposite group, particularly for the dense distribution of arctic lakes. The processing of mountain lakes and endorheic lakes by overlaying with the mountain or basin boundary cannot produce any added value for audiences. Instead, many key attributes in the original datasets, such as the storage, reservoir function, water depth, are missing in the new dataset.
3-The authors mentioned that “there are also geometric differences in features between the different datasets”. In this study, a new lake/reservoir dataset, GLRSED, was generated based on HydroLAKES, GRanD and OpenStreetMap; however, how were the differences in spatial boundaries and temporal inconsistencies among the three datasets solved? Especially, there could be some situations in the spatial intersection of multiple polygons from different sources. Generating a uniform set of water body boundaries would be more convincing.
4-The methodology section is not described in enough detail. The flowchart in Figure 1 is not clear. As introduced in Section 3, the final processed OSM contains data on a total of approximately 0.85 million lakes. How many lakes does OSM contain in draft version? How many reservoirs does OSM contain? How to tackle with the overlapping issues with HydroLAKES data. Moreover, are the “cuo” and “pond” water bodies classified as lakes or reservoirs? Are there lakes with other terminology rules missed in the initial filtering? None of these issues were mentioned.
5-Both the HydroLAKES and GRanD datasets provide the attribute of lake and reservoir capacity. By reviewing the new dataset GLRSED, lake/reservoir from either the HydroLAKES or GRanD datasets account for more than half of the total number (1,412,011), and the authors may consider adding storage capacity information for this portion of the lake/reservoir.
6-The results and discussion section are too thin and still needs further improvement, for example, the dataset description (4.2 Attribute Table of the GLRSED). The basic information of the dataset, such as the total number of lakes and reservoirs, the total area, and the number of lakes and reservoirs, respectively, and the total area, should be described with more spatial and grouping details, despite that it is a data paper.
Minor comments:
Line 12-15: Where are the data source of the most component of reservoir type identification for the GLRSED, from GRanD, GOODD or GeoDAR?
Line 32-33: How about the definition of lake and reservoir expansion and the processing of inconsistence of their spatial boundaries in this study?
Line 41: There is one more “period” mark.
Line 44-48: All water mappings cannot avoid such issues, including the datasets used in this study. The problems are overcome to some degree in the JRC GSW dataset.
Line 68: The superscript number for the area unit.
Line 73-79: What does the time period of lake data from OMS represent?
Line 88-89: The literature on SWOT should be added.
Line 91: What exactly does “The mountain data” refer to?
Line 110-112: The details should be added, e.g., how to treat the inconsistency and commission errors?
Line 115: Please explain in detail why the authors used a 270 m buffer to eliminate the impact of reservoir position deviations.
Line 117: The simple definition of buffer zone distance is not sound in generating data product although it was used in prior study.
Lines 123-124: How the number of orbits passing on each lake is counted, and how many points fallings into the lake are considered valid orbits.
Figures 3-5: It is recommended that the lake/reservoir pattern be shown separately.
Line 143-155: Merging and reorganizing the sections 4.4 & 4.5 could make more sense for lake/reservoir type and other attribute identification.
Line 156-174: For dataset comparison, GLRSED is obtained by fusing HydroLAKES, GRanD and OSM, where most of the lakes and reservoirs are from HydroLAKES and OSM. Therefore, there is no need to compare GLRSED with HydroLAKES and OSM. Instead, it should be compared with other independent data sets.
Line 168: Section 4.6 mentions validation with remotely sensed imagery. Is this a direct overlay image validation? Why not consider taking samples of lakes and reservoirs or other datasets (GLAKES or continental/national-scale lake products well inventoried, e.g., US, Europe, China) for careful manual inspection and accuracy evaluation?
Line 190-201: There are no sufficient quantitative information on the dataset description, including the total number of lakes and reservoirs and other type attributes in the new dataset.
SI-Line 5: The inconsistent spatial boundaries are far limited to this case in Figure S1. How about the cases of one polygon in Data-A containing multiple polygons in Data-B, or multiple polygons in Data-A intersecting multiple polygons in Data-B???
SI-Line 10: It is recommended that lakes and reservoirs from OSM be shown in different colors in Figure S2.
SI-Line 15: There are multiple definitions of mountain areas on the earth, corresponding to multiple data products. Why is this one (Korner et al., 2017) chosen in this study, and how to handle the spatial relationship due to inconsistent spatial scales (boundary square sizes)? Besides, how to process the reservoir identification when there are discrepancies among these three datasets?
SI-Line 25: Meaningless table! BTW, how to calculate the areas of lakes that are located in transboundary areas of two or more nations, e.g., the Great Lakes, Lake Victoria, Caspian Sea, etc.?
The language expression in the manuscript still needs to be improved, especially the method and result section, and further improvement is suggested to clarify the method logic.
Citation: https://doi.org/10.5194/essd-2023-216-RC2
Data sets
A Global Lake/Reservoir Surface Extent Dataset (GLRSED) Bingxin Bai, Lixia Mu, Ge Chen, Yumin Tan https://doi.org/10.5281/zenodo.8121174
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 621 | 258 | 43 | 922 | 50 | 34 | 29 |
- HTML: 621
- PDF: 258
- XML: 43
- Total: 922
- Supplement: 50
- BibTeX: 34
- EndNote: 29
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1