Greenland liquid water discharge from 1958 through 2019

• Published in: Earth System Science Data
• Main Authors: Mankoff, K., Noël, Brice, Fettweis, Xavier, Ahlstrøm, A., Colgan, W., Kondo, K., Langley, K., Sugiyama, S., van As, D., Fausto, R.
• Format: Article in Journal/Newspaper
• Language: English
• Published: Copernicus Publications 2020
• Subjects: Physical; chemical; mathematical & earth Sciences; Earth sciences & physical geography; Physique; chimie; mathématiques & sciences de la terre; Sciences de la terre & géographie physique; Greenland; Ice Sheet
• Online Access: https://orbi.uliege.be/handle/2268/252579; https://orbi.uliege.be/bitstream/2268/252579/1/essd-12-2811-2020.pdf; https://doi.org/10.5194/essd-12-2811-2020

peer reviewed Greenland runoff, from ice mass loss and increasing rainfall, is increasing. That runoff, as discharge, impacts the physical, chemical, and biological properties of the adjacent fjords. However, where and when the discharge occurs is not readily available in an open database. Here we provide data sets of high-resolution Greenland hydrologic outlets, basins, and streams, as well as a daily 1958 through 2019 time series of Greenland liquid water discharge for each outlet. The data include 24 507 ice marginal outlets and upstream basins and 29 635 land coast outlets and upstream basins, derived from the 100 m ArcticDEM and 150 m BedMachine. At each outlet there are daily discharge data for 22 645 d – ice sheet runoff routed subglacially to ice margin outlets and land runoff routed to coast outlets – from two regional climate models (RCMs; MAR and RACMO). Our sensitivity study of how outlet location changes for every inland cell based on subglacial routing assumptions shows that most inland cells where runoff occurs are not highly sensitive to those routing assumptions, and outflow location does not move far. We compare RCM results with 10 gauges from streams with discharge rates spanning 4 orders of magnitude. Results show that for daily discharge at the individual basin scale the 5 % to 95 % prediction interval between modeled discharge and observations generally falls within plus or minus a factor of 5 (half an order of magnitude, or +500 %/−80 %). Results from this study are available at https://doi.org/10.22008/promice/freshwater (Mankoff, 2020a) and code is available at http://github.com/mankoff/freshwater (last access: 6 November 2020) (Mankoff, 2020b).