In order to predict the performance of a coastal high-frequency surface-wave radar (HFSWR) when the antenna must be located some distance back from the ocean, measurements of the path loss over beach sand at 13 and 25 MHz were made at several locations on beaches near Rotterdam, Netherlands in September 2014. The measured mixed-media path loss was compared with estimates made using a procedure developed by Millington in 1949 and were found to agree within 1-2 dB over a total range of 30 dB, even though the distances of a few hundred meters were much shorter than the several kilometers usually employed by the Millington method. Using this method along with the measured loss, it is possible to predict the degradation in achievable range for a coastal HF radar when the path to the sea includes a significant segment of low-conductivity soil, such as dry sand. Example calculations are included for 13 and 25 MHz.

High frequency (HF) radar has become an important tool for remotely mapping the spatial distribution and temporal evolution of waves and currents of the nearshore coastal ocean. Its acceptance along ocean coasts has resulted in the development of several commercially available systems and a planned nationwide coastal network to routinely measure coastal currents. Because HF radiation is known to propagate less efficiently over fresh water than seawater, it has been largely overlooked as a viable tool for freshwater application. However, its potential utility in freshwater was clearly demonstrated by a deployment along Lake Michigan as part of the 1999–2001 Episodic Events Great Lakes Experiment. As part of this experiment, the University of Michigan Multi-frequency Coastal Radar consistently produced reliable near surface current measurements to a range of approximately 25km offshore showing strong correlation with both in-situ measurements and numerical hind-casts. This paper provides background on HF radar technology, a summary of the current state of the art with respect to freshwater and describes the results of a recent experiment to measure the propagation of HF radar signal over freshwater using CODAR Ocean Sensors SeaSondes, operating at 5 and 42MHz with 21W and 90W average radiated powers, respectively. The effective offshore range for these radars was found to be 18km at 5MHz and 4–5km at 42MHz. These findings are consistent with currently available models for the prediction of propagation loss, verifying that they can reliably be used to estimate ranges in freshwater settings.

Under Phase I of this SBIR program, CODAR Oceans Sensors and the University of California, Santa Barbara (UCSB) demonstrated the feasibility of using AIS vessel positions to identify vessel echoes in HF Radar Doppler cross spectra and the characteristics of those echoes to produce antenna pattern measurements (APMs). Under Phase II, the primary objective was the development of an operational prototype. To attain this objective, software for processing the AIS and HF radar data was integrated with 3 party commercial AIS hardware to operate with SeaSonde® data collection computers, outputting antenna pattern data in real time. The prototype developed under Phase II is now close to commercial readiness and the first commercial version is on schedule to be offered within three months of the completion of this phase II program. The technical objectives described in the Phase II work plan were attained as follows: 1) software for automated real time APM determination was developed and deployed; 2) quantitative data quality indicators were developed and improved compared to Phase I methods; 3) azimuthal coverage was expanded compared with Phase I results; and 4) these methods were applied to systems operated at other commonly used HF radar frequencies bands. With this augmentation, SeaSonde operators will be able to calibrate antenna patterns in a less expensive and time-consuming manner, which will both save money and foster higher quality data outputs for Search and Rescue, Spill response activities as well as the many other uses identified by the NOAA Integrated Ocean Observing System (IOOS).

CODAR Ocean Sensors, Ltd. and the University of California, Santa Barbara are developing a method by which HF radar antenna response patterns can be calibrated automatically over time. Currently, over 130 HF radar units are providing coastal surface current maps to the public via the U.S. Integrated Ocean Observing System (USIOOS): http://www.ioos.gov/hfradar/. These real-time data are used for Coast Guard Search and Rescue, Hazardous Materials Spills Response, Water Quality Monitoring, Monitoring Harmful Algal Blooms, Fisheries Management, Modeling, Marine Navigation, Ocean Energy Production. Techniques for improved, automated quality assurance of the data provided by coastal radar stations, such as the one discussed here, will improve the efficacy of efforts in these areas. Passing vessels provide a steady supply of targets for which the echoes in the HF Doppler spectra can be used as source signals. The Automatic Identification System (AIS) transmissions from these vessels provide the position and, therefore, bearing to the vessel. By associating the known AIS positions with HF Doppler echoes, a low-cost calibration procedure can be implemented which can reduce or eliminate more labor-intensive alternatives. This method is demonstrated using data from mid-range systems in the Santa Barbara channel that are operating in the 13 MHz band. A prototype package has been deployed on a system monitoring the Gulf of Farallones and the shipping lanes approaching the San Francisco Bay. Performance and data quality metrics for this prototype will be discussed.

The radar cross sections (RCS) and frequency spectra of a wind turbine consisting of three conducting blades of 42 m radius, with a vertical mast of 65 m over a perfect ground plane, was estimated using the Numerical Electromagnetics Code (NEC). The NEC input deck generation, parsing the NEC output to select the RCS numbers, plotting the time series, and calculating and plotting the frequency spectra were all done by bash shell scripts. The shell scripts generally set up several environment variables and then called Ruby, Octave or Gnuplot programs to perform the text manipulations, data calculations and plot generation. Calibrated plots of time series and frequency spectra for several cases are included. The effect of a dielectric blade is briefly considered.

A RiverSonde was operated during June-August 2010 along the Hudson River in New Jersey at a location about 140 m from the water's edge with the antenna about 40 m above the water level. With this configuration, usable signals were obtained all the way across the river, out to a range of 1400 m from the radar. This was considerably greater than the 300 m which had been observed in previous experiments. Initial data processing shows that the along-channel velocity had the expected tidal signature with a maximum value of approximately 1 m/s and was nearly in phase with the stage measured about 5 km downstream at a NOAA gaging station.

Two-dimensional river flow patterns have been measured using a pair of RiverSondes in two experiments in the Sacramento-San Joaquin River Delta system of central California during April and October 2007. An experiment was conducted at Walnut Grove, California in order to explore the use of dual RiverSondes to measure flow patterns at a location which is important in the study of juvenile fish migration. The data available during the first experiment were limited by low wind, so a second experiment was conducted at Threemile Slough where wind conditions and surface turbulence historically have resulted in abundant data. Both experiments included ADCP near-surface velocity measurements from either manned or unmanned boats. Both experiments showed good comparisons between the RiverSonde and ADCP measurements. The flow conditions at both locations are dominated by tidal effects, with partial flow reversal at Walnut Grove and complete flow reversal at Threemile Slough. Both systems showed complex flow patterns during the flow reversals. Quantitative comparisons between the RiverSondes and an ADCP on a manned boat at Walnut Grove showed mean differences of 4.5 cm/s in the u (eastward) and 7.6 cm/s in the v (northward) components, and RMS differences of 14.7 cm/s in the u component and 21.0 cm/s in the v component. Quantitative comparisons between the RiverSondes and ADCPs on autonomous survey vessels at Threemile Slough showed mean differences of 0.007 cm/s in the u component and 0.5 cm/s in the v component, and RMS differences of 7.9 cm/s in the u component and 13.5 cm/s in the v component after obvious outliers were removed.

Two-dimensional river flow patterns were measured using a pair of RiverSondes in an experiment at Threemile Slough in the Sacramento-San Joaquin River Delta system of central California during October 2007. An earlier experiment at Walnut Grove in April 2007, on the migration path of juvenile fish, revealed complex flow patterns during periods of tidally-induced flow reversals, but data there were limited by low winds. Consequently, a follow-on experiment was conducted at Threemile Slough where wind conditions and surface turbulence historically have resulted in abundant data from a single RiverSonde. The experiment at Threemile Slough included ADCP near-surface velocity measurements from unmanned survey vessels. Quantitative comparisons between the RiverSondes and the ADCPs showed mean differences of 0.007 cm/s in the cross-channel component and 0.5 cm/s in the along-channel component, and RMS differences of 7.9 cm/s in the cross-channel component and 13.5 cm/s in the along-channel component after obvious outliers were removed. Interpolation and integration of the velocity vectors revealed complex trajectories of simulated particles during times of flow reversal and smooth trajectories during the remainder of the tidal cycle.

To properly assess the impact of global climate change and plan for remediation, accurate regional climate models are necessary, functioning on 10 km size scales rather than the typical 100 km scales. Examples of 10 km size scale phenomena are coastal wind eddies at 10-40 km scales, studied over Monterey Bay,California by Archer,Ludwig et al. using shore and buoy anemometers and satellite images. Most frequent in the evening and early morning hours, these, typically cyclonic, eddies are often responsible for fog in the Santa Cruz California area. We have previously demonstrated the ability of multifrequency HF radar (4.8 to 21.8 MHz) to map the ocean wind field. Observations over a year time span indicate standard errors of prediction of 1.7 m/s for wind speed and 25deg for direction with biases of 0.1 m/s and 0.3deg respectively. Combining HF radar wind vector estimates with shore based anemometer data in the WOCSS surface wind field model allows formation of detailed (3 to 5 km resolution) images of wind eddies over Monterey Bay and observation of their creation and decay. We discuss requirements for wind field observations with HF radars and demonstrate how multiple radar sites can be used to produce HF radar wind field maps. We report observations of 10-20 km cyclonic eddies in the northern and eastern parts of Monterey Bay.

Two RiverSondes were operated simultaneously in close proximity in order to provide a two-dimensional map of river surface velocity. The initial test was carried out at Threemile Slough in central California. The two radars were installed about 135 m apart on the same bank of the channel. Each radar used a 3-yagi antenna array and determined signal directions using direction finding. The slough is approximately 200 m wide, and each radar processed data out to about 300 m, with a range resolution of 15 m and an angular resolution of 1 degree. Overlapping radial vector data from the two radars were combined to produce total current vectors at a grid spacing of 10 m, with updates every 5 minutes. The river flow in the region, which has a maximum velocity of about 0.8 m/s, is tidally driven with flow reversals every 6 hours, and complex flow patterns were seen during flow reversal. The system performed well with minimal mutual interference. The ability to provide continuous, non-contact two-dimensional river surface flow measurements will be useful in several unique settings, such as studies of flow at river junctions where impacts to juvenile fish migration are significant. Additional field experiments are planned this year on the Sacramento River.

Coastal wind eddies at the 10-40 km scale have been studied over Monterey Bay, California by Archer, Ludwig et al. using shore and buoy anemometers and satellite images. Most frequent in the evening and early morning hours, these, typically cyclonic, eddies are often responsible for fog in the Santa Cruz area. We have previously demonstrated the ability of multifrequency HF radar (4.8 to 21.8 MHz) to map the ocean wind field. Observations over a year time span indicate standard errors of prediction of 1.7 m/s for wind speed and 25deg for direction with biases of 0.1 m/s and 0.3deg respectively. By combining HF radar wind vector estimates with shore based anemometer data in the WOCSS surface wind field model we are able to form detailed (3 to 5 km resolution) images of eddies over Monterey Bay and to follow their development and decay. We discuss the requirements for making wind field maps with HF radars and demonstrate how the changing array of multiple radar sites can be used to produce HF radar wind field maps. We report observations of 10-20 km cyclonic eddies in the northern and eastern parts of Monterey Bay.

A UHF RiverSonde radar system was deployed in an intertidal salt marsh during November and December 2005 near Georgetown, South Carolina. The purpose of this experiment was to evaluate radar performance in a coastal plain salt marsh, where the channels are highly sinuous and the flow is dominated by tides. To help evaluate radar performance, an acoustic current profiler was placed in a large subtidal channel located in the field of view of the radar. Comparisons indicated that the radar accurately reproduced both the magnitude and phasing of the surface current. Spatial coverage was generally good in the channels but much more sporadic over the marsh platform. Temporal coverage in the channel was also good and correlated with wind speed. Reliable measurements were obtained under wind speeds as low as 2-3 m s(-1), which are common at this site. In addition, some unexpected signals were seen at near zero Doppler shift, which were attributed to the saltwater vegetation Spartina alterniflora, which acted as effective radar targets. However, these signals were uncorrelated and likely can be filtered.

A UHF RiverSonde radar system was deployed in a tidal marsh during November and December 2005 near Georgetown, South Carolina. The water flow in the marsh was dominated by tidal effects, and the How pattern was complicated due to the sinuous water channels. A single radar was used, so only the radial component of velocity was observable, but one of the main channels was aligned nearly in the radial direction from the radar antenna so that interpretation of the flow in that channel is relatively straightforward. Acoustic flow measurements are also available for comparison. The radar system performed well, and usable data were obtained under wind speeds of 2-3 m/s. Occasionally some unexpected signals also were seen.

Conventional measurements of river flows are costly, time‐consuming, and frequently dangerous. This report evaluates the use of a continuous wave microwave radar, a monostatic UHF Doppler radar, a pulsed Doppler microwave radar, and a ground‐penetrating radar to measure river flows continuously over long periods and without touching the water with any instruments. The experiments duplicate the flow records from conventional stream gauging stations on the San Joaquin River in California and the Cowlitz River in Washington. The purpose of the experiments was to directly measure the parameters necessary to compute flow: surface velocity (converted to mean velocity) and cross‐sectional area, thereby avoiding the uncertainty, complexity, and cost of maintaining rating curves. River channel cross sections were measured by ground‐penetrating radar suspended above the river. River surface water velocity was obtained by Bragg scattering of microwave and UHF Doppler radars, and the surface velocity data were converted to mean velocity on the basis of detailed velocity profiles measured by current meters and hydroacoustic instruments. Experiments using these radars to acquire a continuous record of flow were conducted for 4 weeks on the San Joaquin River and for 16 weeks on the Cowlitz River. At the San Joaquin River the radar noncontact measurements produced discharges more than 20% higher than the other independent measurements in the early part of the experiment. After the first 3 days, the noncontact radar discharge measurements were within 5% of the rating values. On the Cowlitz River at Castle Rock, correlation coefficients between the USGS stream gauging station rating curve discharge and discharge computed from three different Doppler radar systems and GPR data over the 16 week experiment were 0.883, 0.969, and 0.992. Noncontact radar results were within a few percent of discharge values obtained by gauging station, current meter, and hydroacoustic methods. Time series of surface velocity obtained by different radars in the Cowlitz River experiment also show small‐amplitude pulsations not found in stage records that reflect tidal energy at the gauging station. Noncontact discharge measurements made during a flood on 30 January 2004 agreed with the rated discharge to within 5%. Measurement at both field sites confirm that lognormal velocity profiles exist for a wide range of flows in these rivers, and mean velocity is approximately 0.85 times measured surface velocity. Noncontact methods of flow measurement appear to (1) be as accurate as conventional methods, (2) obtain data when standard contact methods are dangerous or cannot be obtained, and (3) provide insight into flow dynamics not available from detailed stage records alone.

Strong evidence for 10-40 km scale eddies in the surface wind field over Monterey Bay, California has been presented by Archer, Ludwig et al. using shore and buoy anemometers and satellite images. These cyclonic eddies are frequently present in the evening and early morning and are responsible for fog in the Santa Cruz area. We have previously demonstrated the ability of multifrequency HF radar (4.8 to 21.8 MHz) to map the ocean wind field. Observations over a year time span indicate standard errors of prediction of 1.7 m/s for wind speed and 25deg for direction with biases of 0.1 m/s and 0.3deg respectively. Here we report observation of a 10-20 km cyclonic eddy at the north end Monterey Bay. By combining HF radar wind vector estimates with shore based anemometer data in the WOCSS surface wind field model we are able to form a detailed (5 km resolution) image of an eddy over Monterey Bay and to follow its development and decay. This particular eddy contrasts with previous eddy observations in terms of season and meteorological setting.

This paper is an extension of other work that addresses the use of radar echoes from ships of opportunity to determine the proper phase corrections for small-loop phased-array antennas used within high-frequency (HF) ground-wave radar systems. This technique also yields estimates for unknown ship bearings that (for cases where there is adequate signal-to-noise ratio of 20 dB or more) are consistent to within 2deg-3deg among measurements from independent radar frequencies. Within this paper, phase corrections gathered from actual ships of opportunity are compared to phase corrections gathered during a calibrated transponder run, in which the ship bearing is known. The phase corrections derived from the ship of opportunity presented in this paper were consistent with the known phase corrections to within 13.2deg (for the worst case). Furthermore, the estimates of the ship bearings collected from the two usable radar frequencies were consistent to within 1deg of each other