Greenland liquid water runoff from 1979 through 2017

We provide high-resolution datasets of Greenland hydrologic outlets, basins, and streams, and a 1979 through 2017 time series of Greenland liquid water runoff for each outlet. Outlets, basins, and streams are derived from traditional hydrologic routing algorithms over the surface of a 100 m ArcticDEM digital elevation model (DEM) twice: Once to the ice margin and once to the coast. We then partition liquid water runoff from both ice and land from two regional climate models (RCMs; MAR and RACMO) into each basin and at each outlet location. The data include 18903 ice basins and outlets (614 basins 5 greater than 10 km2), 30241 land basins and outlets (958 basins greater than 10 km2), major streams in each basin, and daily runoff water volume flow rate at each outlet from each of two RCMs. We perform a sensitivity study of outlet location change for every ice sheet location over a range of hydrologic routing assumptions and data sets. Annual runoff from the ice ranges from ~136 km3 in 1992 to ~785 km3 in 2012. Daily maximum ice runoff from one basin is as large as 4380 m3 s-1. Both ice runoff magnitude and variability increase over the time series. Land runoff contributes an additional ~35 % to the ice runoff. 10 Comparison with 9 basins instrumented with stream gauges shows a range of (dis)agreement from poor to excellent between our estimated discharge and observations. As part of the journal’s living archive option, and our goal to make an operational product, all input data, code, and results from this study will be updated as needed (when new input data are available, as new features are added, or to fix bugs) and made available at doi:10.22008/promice/data/freshwater_runoff/v01 (Mankoff, 2020) and at http://github.com/mankoff/freshwater. 15

1. The result section does not highlight the main contribution of this study very well. It includes numerous numbers of basins, outlets, streams, runoff but their importance is not well demonstrated. Furthermore, this section focuses on the total ice and land runoffs which can be easily derived from RCMs and have been well reported in previ-C1 ous studies. I suggest the result section should focus on what we can learn from runoff partitions in different basins, which is the new contribution of this study.
2. The discussion section is too long and not easy to follow, particularly "6.2 Validation against observations". Most parts of section 6.2 should be removed to the result section. I suggest the authors only highlight the most important implications of their datasets and shorten this section.
3. It is important to mention that moulins are not identified so stream networks are delineated to continuously flow from inland to ice edge outlets. Therefore, the stream product may not represent the actual hydrological environments where moulins are widely distributed and fragment drainage networks, such as southwest GrIS. In contrast, the stream product may reasonably predict northwest GrIS drainage pattern since no moulins form there. Moreover, the contributing area threshold should be better illustrated since it determines the extent of streams. It may be useful to state that the derived stream product aims to represent the general meltwater flow pattern rather than the actual spatial distribution of supraglacial rivers and streams.
4. The quality of the main figures should be improved. Currently they are not satisfactory for publishing. Also, the main point of each figure should be highlighted.

C2
Minor comments: P1 L17, in this paragraph, I think it is necessary to say surface runoff contributes very importantly to Greenland mass balance (along with ice discharge).
P2 L4, it is not straightforward to understand "liquid runoff form surface melt, condensation, and rainfall".
P2 L22, why is 100 m ArcticDEM used to do the analysis?
P3 L5, it may be worthy to mention that weathering crust of bare ice layer can store meltwater. Citation is required for this sentence.
P3 L7, citation is required for this sentence.
P3 L10, it is not common to use the term "hydrologic head elevation".
P3 L11, it is unclear how outlets are determined.
P3 L13, it is unclear what "major streams" means, some specific channel initiation thresholds (i.e. contributing area thresholds) are used to extract streams? It may be useful to call these "major streams" as rivers.
P3 L17, why is 1 km2 used as threshold to merge small basins?
P3 L20, "When this value is negative, it indicates submarine (subglacial) discharge", this sentence is not clear.
P3 L21, this section is too long. I suggest it should be shorten or some parts can be put into SUPP.
P4 L24, see my general comment, more explanations should be provided for the stream product.
P5 L8, it is not easy to understand what these numbers mean and why they are important.
P5 L23, which basin? also report the similar value in Lewis and Smith (2009).
P5 L26, Mt. Pinatubo eruption, add a citation to support this result.
P5 L27, the land runoff is considerably large. It is useful to further illustrate its meaning.
P6 L7, "Routing with a 5 km DEM is likely to cause some basins and outlets to drain into an incorrect fjord", what is the reason for this? Interactive comment on Earth Syst. Sci. Data Discuss., https://doi.org/10.5194/essd-2020-47, 2020.