Plume Modeling

Once we have moved sediment and water over hillslopes and routed it through a river system, we can think about this flux of water and sediment draining into lakes or into the ocean. Typically the riverine freshwater and the suspended sediments will form a so-called hypopycnal sediment plume. We will use a component called PLUME to investigate the behavior of these river sediment plumes. The model is largely based on turbulent jet theory presented by Albertson et al., (1950). In this lab we will look at the river plume characteristics and the depositional patterns for single plume events. To learn more about Plume and its approach you can download this presentation.

The WMT allows you to set up simulations, but once you are ready to run them, you will need an account on the CSDMS supercomputer to submit your job. More information on getting an account can be found here Beach HPCC Access

STEP1 Load the PLUME component

If you have never used WMT, learn how to use it here.
>> Open the WMT
>> Select the PLUME component as a driver.
>> We will use PLUME as a stand-alone model in this lab, so no need to connect it to any ports.

River Characteristics
Plume is designed to model the distribution of sediment draining into any lake or the coastal ocean. In this lab, we will focus on a specific braided river system in West-Greenland. The Watson River drains a small section of the Greenland Ice Sheet, it winds through a narrow braided river plain for about 20km and then drains into Kangerlussuaq Fjord. The photo shows the river braidplain in 2012, which was a peak melt year in Greenland. The river waterdepth on the braidplain is typically shallow (>1m with perhaps a smaller channel section that is deeper ~ 3m).

PLUME briadplainwidth.png QdailyWatson times2.jpg

Question 1
Use the annual hydrograph of the Watson River to estimate discharge events for 2 distinctly different days in the summer of 2012. 
Indicate what your estimated dates and discharges are. The photo above can help to estimate a river mouth width. 
Can you estimate the river flow velocity? Please motivate your answer.

>> Set up two simulations of plume model for the different discharge events you have estimated.
>> Change the default basin geometry to be 4 km wide and 30km long to make the basin more like a fjord
>> Plume is a steady-state model, it calculates a single solution for given conditions, so its timestep is 1 day. However, if you want to create output you will have to ask it to create output every 0.5 day. Switch on the netcdf file output.

>> Download the generated depositional grid for visualization in Panoply: ‘’
>> Alternatively, you can use the 'plume.out' file, it contains ASCII values for the plume grid. This file can be easily imported into Matlab (for example). >> In either case, you may want to use a map visualization with a log-scale, just because the values of the plume sedimentation drop rapidly with distance.

PLUME lab1sim1.png Plume output warning.png

Question 2
How far out did sediment get deposited by the plume during early season conditions and how far out at high melt season conditions? 
How thick was the thickest deposit for each of the two events? Plot your two output grids.

Sediment Settling Rates

PLUME is one of the key components of the marine sedimentation model SedFlux; river plumes are just one of the marine processes that control marine sedimentation. Other processes include waves and tides and turbidity currents and biological production. The stand-alone plume component runs for a single flood event, and it runs for a single suspended sediment load class only. You can imagine that in a more comprehensive framework the plume model runs for several grain-size classes and for a sequence of many different discharge events over time. We will look at the effect of changing the suspended grain-size in the river.
Empirically, sediment removal rates, λ, [1/T] for fine sediment in the plume model are defined as a function of grain-size, D:

λ(D) = 0.222D + 1.573

This relationship was derived from time-lapse image analysis of underwater particle settling for grains between 2 micron and 42 micron.

WMT plume settlingratesmenu.png
>> Run 3 simulations of PLUME. Vary the removal rates systematically.
>> List your parameters for grainsize and removal rate, discharge scenario and basin shape.

Question 3
How much effect does the removal rate have on the overall plume depositional pattern? 
Plot the 3 grids of sediment deposition using Panoply or Matlab. 
Question 4
Compare the empirically derived removal rates with Stokes settling velocities for these grain diameters? 
What would be the removal rates if you assume turbulent conditions during settling? 
Can you explain the difference?

Plumes in an ocean basin

In most real-world cases plumes will enter an ocean basin where other conditions affect plume behavior. One of these effects is the importance of longshore currents. The following experiments show the effect of a persistent longshore current acting on the plume. The plume model can calculate the effect of Coriolis force when small longshore currents act upon the plume

>> Set up and run 2 simulations of PLUME with positive and negative ocean current velocities (1.5 m/s) for a wide 30 by 30 km basin. List your input parameters for the discharge event you are running.
>> Visualize and plot the output grid stacks in Panoply.

Question 5
What effect of longshore currents do you see for the unconstrained open marine basin? 
Hypothesize what the effect of the Coriolis force should be on the plume deflection? 
What would you expect to see at low latitude and at high latitude (i.e. our fjord in Greenland is 66° North). 
And what would you expect to see with positive and negative ocean currents?


  • Albertson, M.L., Dai, Y.B., Jensen, R.A., Hunter, R., 1950, Diffusion of submerged jets. American Society Civil Engineers Trans, v. 115, p. 639-697.
  • Syvitski, J.P.M., Skene, K.I., Nicholson-Murray, K., Morehead, M., 1998a, PLUME 1.1; deposition of sediment from a fluvial plume. Computers & Geosciences, v. 24, 2, p. 159-171. [1]