AIRS Department Home Page

   William J. Plant  
      Principal Research Scientist  
      AIRS Department  
      Applied Physics Laboratory  
      University of Washington  

   U.S. Geological Survey  
   National Science Foundation  

We are developing techniques for the long-term monitoring of surface velocity at the mouth of the Columbia River with microwave Doppler radars. Through our past efforts with the USGS and ONR, we have developed an X-band Doppler radar which we call RiverRad that has proven valuable for the long-term monitoring of river surface currents and for determining discharge in stable streams (Plant et al., 2005; Costa et al., 2006). Our recent work with a similar CW microwave system called Riverscat has strongly suggested that discharge can also be determined on unstable streams, those with frequently changing beds, using microwave Doppler sensors.

RiverRad operating on the Snohomish River near Everett, Washington.

Intensity image from RiverRad overlaid on a visible image of the same region.

When RiverRad has been operated on straight sections of rivers, as opposed to the complex region shown above, it has proven to yield an effective measurement of the surface velocity. The upper black and white plot (seen below) shows a profile of along-river surface velocities across the Cowlitz River near Castle Rock, WA as measured by RiverRad and by an acoustic system called BoogieDopp. Clearly, the agreement is very good. RiverRad also yields cross-river velocities but no measurements have been available to date with which to compare them. The lower black and white plot (seen below) shows along-river currents in the center of the Cowlitz River from both RiverRad and the UHF system CODAR along with USGS stage measurements. Again the agreement between the two current measuring systems is very good and both show a high correlation with the measured stage.

These examples show that RiverRad is capable of mapping surface currents and roughness across rivers. If the bottom of the river is stable, RiverRad has also been shown to yield a discharge that is comparable to that obtained from standard USGS measurements of stage (Costa et al., 2006). Recent work about to be submitted for publication suggests that this is also true on unstable rivers.

RiverRad measurements of surface currents on the Cowlitz River near Castle Rock, WA compared with those from an acoustic sensor.

RiverRad surface currents compared with those from the UHF CODAR sensor; the line is the stage of the river.


In the next two years, we will set up our existing RiverRad on the north side of Columbia River just east of the Astoria Bridge to monitor surface currents and roughness. We will also work to develop a smaller, solar-powered radar to go on the south side.

In the first year, we will set up RiverRad. The system will run unattended, collecting, processing and storing data on surface currents and roughness. We will work on developing a method for extracting the processed files from RiverRad's disk drive and posting it on the internet. Unfortunately, RiverRad requires line power to operate.

RiverRad's maximum range is 4 km, about half the distance across the Columbia River at this point. We will develop a second system based on the microSAR of Dr. David Long at Brigham Young University (see photo, below). We have borrowed one of these systems from Dr. Long and will soon test its ability to emulate RiverRad. At present, no processing of microSAR data is done in real time and the processing software available to use the data off line is designed to produce images when the system is flown on an airplane. Thus to use the microSAR as a RiverRad, we will first have to first repackage it, then verify that it has sufficient signal-to-noise ratio, and finally write software to process the data in a manner to extract surface velocities in real time.

If all of this is successful, we will deploy the microSAR on the other side of the river so that currents and surface roughnesses can be mapped over the entire width of the river. Dr. Long is presently working with an east coast company to sell the microSAR for $30K. Assuming that our tests with the borrowed system are successful, our plan would be to purchase one microSAR to modify in the first year and a second in the second year and thus have one operating on each side of the river after two years.

The microSAR developed by Dr. David Long at BYU. For scale, the antennae (rectangles at the top of the picture) are each 30 cm long.

These systems are sufficiently low power that they can be made to run off of solar panels. Thus, by the end of the project, perhaps three years out if all goes well, we will have a portable river surface current measuring system that will run independently and automatically at any desired site. If this project is successful, it will produce a true transformation of the way rivers and estuaries are routinely monitored. The final system will yield much more information than present gaging stations and will be much easier to set up.


We will operate RiverRad in the Columbia River estuary to monitor surface currents and roughness. Simultaneously, we will develop a low-cost, low-power, relatively portable microwave Doppler radar to do the same job. We will set up one on each side of the river and use their output to produce river discharge.


Plant, W.J., W.C. Keller, and K. Hayes, Measurement of river surface currents with coherent microwave systems, IEEE TGRS, 43 (6), 1242-1257, doi: 10.1109/TGRS.2005.845641, 2005.

Costa, J.E., R.T. Cheng, F.P. Haeni, N. Melcher, K.R. Spicer, E. Hayes, W.J. Plant, K. Hayes, C. Teague, and D. Barrick, Use of radars to monitor stream discharge by non-contact methods, Water Resources Research, 42, W07422, doi:10.1029/2005WR004430, 2006.