NCCS-Hosted Simulations Probe the Interactions of the
Freshwater Yukon River and the Salty Arctic Ocean


The Enhanced Thematic Mapper Plus on the Landsat 7 satellite acquired this natural-color image of the Yukon Delta in southwestern Alaska on September 22, 2002. Looking a little like branching and overlapping blood vessels, the delta’s rivers and streams flow through circuitous channels toward the sea, passing and feeding a multitude of coastal ponds and lakes. A new NASA simulation probes how freshwater from the Yukon River interacts with the salty Bering Sea. Image and caption adapted from NASA Earth Observatory.

The NASA Center for Climate Simulation (NCCS) Discover supercomputer powered a model NASA Goddard Space Flight Center scientists developed to simulate the physical properties and transport of water in the lower Yukon River and Northern Bering Sea — water that ultimately reaches the freshest of the world’s major oceans, the Arctic Ocean.

“The Yukon River is one of the largest rivers in the world, and a lot of the freshwater and potentially dissolved material that enters the ocean from the river is transported into the Arctic,” said J. Blake Clark, an assistant research scientist in NASA Goddard’s Ocean Ecology Laboratory and affiliated with GESTAR II through the University of Maryland, Baltimore County. “We undertook this study to understand the balance between freshwater input and external forces such as wind and tides on the structure of the freshwater Yukon River plume and the transport of water in the coastal ocean.”

Clark collaborated on the computational study with his postdoctoral mentor, Antonio Mannino, a NASA Goddard Ocean Ecology Laboratory research oceanographer. Their study appears in the journal Frontiers in Marine Science.

Study co-authors were (left to right) J. Blake Clark and Antonio Mannino, scientists in NASA Goddard Space Flight Center’s Ocean Ecology Laboratory.

For this research, Clark and Mannino used the open-source Finite Volume Community Ocean Model (FVCOM). They implemented the FVCOM model for a region spanning from 200 kilometers (km) upstream of the coast at Pilot Station, Alaska, to south of the Yukon delta northward to west of Nome, Alaska, encompassing nearly all of Norton Sound and the Yukon delta (see figures A-C below).

(A) The Arctic Ocean showing the Yukon River watershed in purple; (B) Alaska and western Canada showing the Yukon watershed with the extent of the Yukon River FVCOM hydrodynamic model domain; (C) the FVCOM model domain with Alaska Department of Fish and Game fishery survey stations as red markers and the stations sampled under a 2019 NASA Remote Sensing of Water Quality project as cyan markers. The color contours are the average surface salinity predicted by FVCOM for 2019. Figure adapted from Clark and Mannino, 2022.

YukonFVCOM, as the researcher team calls this implementation of the model, ran at spatial resolutions ranging from roughly 100 meters in shallow areas of the coasts and deltas to roughly 2.5 km offshore. The simulation covered 7 years (2004–2005, 2015–2019) of mostly ice-free months (April–September) to capture varying river flow and weather conditions. Model input data came from the National Oceanic and Atmospheric Administration (NOAA) North American Regional Reanalysis (weather) and the U.S. Geological Survey National Water Information System (river flow and temperature).

Clark and Mannino ran YukonFVCOM on the NCCS Discover supercomputer, using 1,120 ”Skylake” cores for approximately 89,740 core hours. Open access to model output data (including variables such as temperature, salinity, and velocity) is available on the NCCS DataPortal.

“This model has over 700,000 grid cells, and it would not be feasible to run the model on a lab-scale computer system,” Clark said. “NCCS and the Discover supercomputer allowed me to efficiently perform all aspects of the research, including model development and testing, output analysis, and big data processing.”

Impact: This study improves understanding of coastal ocean physics in a region that is rapidly shifting due to climate change and provides a modeling system for NASA to continue studying physics, chemistry, and biology into the future under different scenarios.


The YukonFVCOM simulation shows that phenomena occurring on differing timescales largely determine the position, current speed, and strength of the river plume. On short timescales (days to weeks) the dominant factor is wind direction and speed. On interseasonal and interannual timescales, the dominant factor is the timing and volume of incoming river freshwater as it flows over heavier ocean saltwater.

“Wind can steer the freshwater plume in the ocean for hundreds of miles, which will affect where sediment and carbon is transported and processed in the ocean,” Clark said. “There is a very large ‘bulge’ of freshwater [see left panel in figure below] that develops to the west before currents push water back towards the coast and then further north to the Bering Strait.”

The panels show seven-year depth-averaged May–July (A) salinity and (B) sea surface height (η) with the current velocity field represented by streamlines. In (A), one noteworthy feature is a large bulge of freshwater (colored blue) near the Alaska coast that coincides with higher water elevation and a large eddy (B). Figure from Clark and Mannino, 2022.

In contrast, “the freshwater input drives these strong currents on longer timescales due to the lighter freshwater flowing over heavier saltwater in the bottom layer,” Clark said. Validation for the YukonFVCOM model results comes from observations of sea surface temperature taken by NASA MODIS and NOAA VIIRS satellite instruments.

In a follow-on NASA Carbon Cycle Science project, Clark and Mannino have linked the YukonFVCOM hydrodynamic model to a full carbon cycle model and prepared a soon-to-be published paper describing the transport and transformation of carbon in the Yukon River delta and plume. The team is running the updated model on Discover and will expand their simulations into other years and test additional scenarios. “We will produce estimates on how the river and ocean carbon cycle will look in the future under various climate change scenarios such as increased temperature and atmospheric carbon dioxide,” Clark said.

Related Link


Jarrett Cohen, NASA Goddard Space Flight Center
June 14, 2022