Kenneth Laws

Simulation-based evaluations of HF radar ocean current algorithms

A computer simulation is used to analyze errors in high-frequency (HF) radar ocean surface current measurements. Two pointing algorithms used for current extraction, a direction finding approach using MUltiple SIgnal Characterization (MUSIC) developed by Schmidt (1986), and conventional beam forming, are compared in terms of the effect of variations in sea state parameters on current measurement error. The radar system parameters used in the simulation were taken from the University of Michigans multi-frequency coastal radar (MCR), which operates on four frequencies from 4.8 to 21.8 MHz and employs an eight-element linear phased array for its receive antenna. Results show MUSIC direction finding to be applicable to phased array systems and to have a better sensitivity to sharp current features, but larger random error than traditional beam forming methods. Also, for cases where beam forming errors are dominated by beam width or low signal to noise ratio, results show MUSIC to be a viable alternative to beam forming.

Estimation and Assessment of Errors Related to Antenna Pattern Distortion in CODAR SeaSonde High-Frequency Radar Ocean Current Measurements

A simulation-based investigation of errors in HF radar‐derived, near-surface ocean current measurements is presented. The simulation model is specific to Coastal Ocean Dynamics Application Radar (CODAR) SeaSonde radar systems that employ a compact, collocated antenna geometry. In this study, radial current retrievals are obtained by processing simulated data using unmodified CODAR data processing software. To avoid limiting the results to specific ocean current and wind wave scenarios, the analyses employ large ensembles of randomly varying simulated environmental conditions. The effect of antenna pattern distortion on the accuracy of retrievals is investigated using 40 different antenna sensitivity patterns of varying levels of distortion. A single parameter is derived to describe the level of the antenna pattern distortion. This parameter is found to be highly correlated with the rms error of the simulated radial currents (r 5 0.94) and therefore can be used as a basis for evaluating the severity of site-specific antenna pattern distortions. Ensemble averages of the subperiod simulated current retrieval standard deviations are found to be highly correlated with the antenna pattern distortion parameter (r 5 0.92). Simulations without distortions of the antenna pattern indicate that an rms radial current error of 2.9 cm s 21 is a minimum bound on the error of a SeaSonde ocean radar system, given a typical set of operating parameters and a generalized ensemble of ocean conditions.

Using HF surface wave radar and the ship Automatic Identification System (AIS) to monitor coastal vessels

We compare the ship detection capabilities of the Automatic Identification System AIS (installed on some ships) and coastal, surface wave HF radars, showing how to use both systems together to enhance ship detection performance in coastal regions. Practical reasons to want better real-time awareness of the location, velocity and type of vessels along coasts include vessel safety, protection of the coastal environment and national security. Our model for the HF radar aspect uses an example radar with significant power and aperture, similar to the Pisces radar. The AIS model is for the high power (12.5 W) AIS unit and a significantly elevated receiver (~ 250 ft asl). The HF system show good capability to ranges of ~ 150 km for small ships to 250 km for large ships. The AIS system shows excellent capability out to a typical horizon of ~ 50 km with irregular coverage beyond using ducted propagation to several hundred km and more. Use of both systems allows monitoring of both AIS and non-AIS equipped ships and enhances probability of detection for situations where both systems are functional.

Identifying ship echoes in CODAR HF radar data: A Kalman filtering approach

Coastal nations have an interest in maritime domain awareness for applications in national security, coastal conservancy, fishery and stewardship of the exclusive economic zones (EEZs) along their coastlines. Maritime situational awareness involves knowing the location, speed and bearing of ships and boats in the EEZ. HF radar is a useful tool in providing ship information in real time. It is especially effective when combined with ship-borne AIS beacons. Our previously developed HF radar and AIS ship detection models estimate signal to noise ratio (SNR) as a function of range, including ducted propagation for the AIS radio signals. However, ship detection is hampered by the high variability of HF echoes from ships. This is due in part to the aspect dependence of ship radar cross-section and to the presence of clutter bands at known Doppler shifts from both the ground and ocean waves. Tracking ships using their HF radar echoes becomes the means for effectively monitoring the presence of ships in the coastal ocean. We explore the application of Kalman filtering to the ship tracking problem, following the techniques described by J. V. Candy. This approach is described and demonstrated with a simple example.

Monitoring of Coastal Vessels Using Surface Wave HF Radars: Multiple Frequency, Multiple Site and Multiple Antenna Considerations

Marine situational awareness is a key factor in coastal activities, e.g. national security (terrorism, drug smuggling, etc.) and environmental protection (marine protected areas, fishery monitoring and regulation, oil spills, etc.). HF surface wave radar is a strong candidate to become a component of any large area, vessel-monitoring network. We discuss estimating the primary metrics for assessing performance for ship tracking radars (probabilities of detection Pd and false alarm Pfa) and their variation with parameters, such as range, azimuth and frequency, and number of observing modes. We discuss several design options for an HF radar, ship monitoring system, such as multiple or single frequency, multiple or single sites or method of target bearing determination (MUSIC or real aperture antenna) and present a model of the SNR for ship detection by HF radar and as well as observational examples of ship detection with single and multifrequency HF radars. Finally we suggest future experiments and draw conclusions.

Prototype autonomous mini-buoy for use in a wireless networked, ocean surface sensor array

We report the design, prototype construction and initial testing of a small minibuoy that is aimed at use in a coordinated, wireless networked array of buoys for near-surface ocean sensing. This vehicle is designed to fill the gap between larger ocean surface vessels and/or moored buoys and subsurface gliders. The size and cost is low enough that these versatile sensor platforms can be deployed easily and in quantity. Since these minibuoys are mobile, they can keep station in currents as large as 25 cm/s or move as an adaptive, coordinated sensor array for high resolution in both time and space. The buoy is about 74 cm (29 in) long, 41 cm (16 in) wide (max) and weighs about 14.5 kg (32 lbs); hence, it can be deployed easily from small craft. Deployment times are about 1 to 2 days or more - longer with solar power. The buoy structure is fiberglass and PVC with two 2 W DC motors. Control is done with GPS and magnetic heading sensors and a PID scheme to maintain course. Communication is via a 900 MHz system with a range of 1 to 2 km and plans for a longer range HF/VHF or satellite system. The initial sensor system is designed for ocean hyperspectral observations as surface truth for airborne system calibration and validation and other ocean color applications. Acoustic, wave, air & water temperature sensors as well as GPS are included. The Mark I prototype has been successfully tested in a pool with manual control.

Using multifrequency HF radar to estimate ocean wind fields

HF radar has become an important tool for mapping surface currents in the coastal ocean and has been used to determine wind direction. Here the author investigate further the ability of multifrequency HF radar to measure the vector wind field and the impact that such measurements have on the measurement of coastal wind fields over both the land and sea. In this study the author use data collected over Monterey Bay, California from Jan. to Aug. 2001. Their multifrequency coastal radars (MCRs) operated at 4.8, 6.8, 13.4 and 21.8 MHz, measuring currents at effective depths of about 2.5, 1.8, 0.9 and 0.6 m respectively. Here, the author move beyond their preliminary reports by examining the durability of their HF wind vector measurements over a seven-month data set. The results over this longer time span indicate standard errors of prediction (SEPs) of 1.2 and 1.1 m/s for the U and V wind speed components respectively with biases less than 0.15 m/s. The author also investigate a Kalman filtering modification to their partial least squares algorithm. Further, the author demonstrate the beneficial impact of multifrequency HF radar, wind field measurements, on estimation of the coastal wind field over both land and sea.

Monitoring coastal vessels for environmental applications: Application of Kalman filtering

Maritime domain awareness is important for coastal nations in terms of applications to coastal conservancy, security, fishery and stewardship of their exclusive economic zones (EEZs). Maritime situational awareness involves knowing the location, speed and bearing of ships and boats in the EEZ. HF radar is a useful tool in providing ship information in real time. It is especially effective when combined with information from ship-borne AIS beacons. Our previously developed HF radar and AIS ship detection models estimate signal to noise ratio (SNR) as a function of range, including ducted propagation for the AIS radio signals. However, ship detection is hampered by false targets related to wave echoes, interference and the high variability of HF echoes from ships. This is due in part to the aspect and frequency dependence of ship radar cross-section and to the presence of clutter bands at known Doppler shifts from both the ground and ocean surfaces. Distinguishing ship echoes from false alarm echoes is significantly aided by identifying radar targets with ship-like behavior. Thus, tracking ships using their HF radar echoes becomes an important means for effectively monitoring the presence of ships in the coastal ocean. We demonstrate the application of Kalman filtering to the ship-tracking problem with examples using data from the COCMP HF radar network along the California coast. As with other radar tracking problems, the Kalman approach proves effective in this application as well.

A system trade model for the monitoring of coastal vessels using HF surface wave radar and ship automatic identification systems (AIS)

Coastal nations have an interest in maritime domain awareness for applications in national security, coastal conservancy, fishery and stewardship of the exclusive economic zones (EEZs) along their coastlines. Using our previously developed HF radar and AIS ship detection models we find signal to noise ratio (SNR) as a function of range, including ducted propagation for the AIS radio signals. We use these SNR estimates to find probability of detection Pd and then explore multiple systems and stations at variable spacings along the coast. Our example HF radar has significant power and aperture, similar to the Pisces radar. The AIS model is for high power (12.5 W) AIS and a significantly elevated receiver (≈ 250 ft asl). A combined system of HF radar and AIS shows good capability (Pd > 0.9) to ranges of ≈ 125 km for small ships and to 200 km for large ships. Considering a system of sites separated by 100 km we find that a Pd of > 0.9 can be maintained to a distance off shore of 130 km even for small, 120 ton, ships.

Error assessment of HF radar-based ocean current measurements: An error model based on sub-period measurement variance

Data from CODAR-type ocean current sensing radar systems are used here to evaluate the performance of an error indicator provided as part of the available radar data. Investigations are based on data from pairs of radar systems with over-water baselines. Approximately year-long time series are used. The radar data are the typical hourly radial measurements provided by CODAR systems. These measurements are actually the median (or mean) of anywhere between 2 and 7 sub-hourly measurements collected by the radar system. The error indicator under examination is based on the standard deviation (std) of the sub-hourly radials, divided by the square root of the number of sub-hourly radials. These values are recorded in the hourly data files produced by recent versions of the CODAR data processing software. Examination of the model demonstrates a positive correlation between the model and the measured baseline difference std for all baseline pairs examined. The predictive capability of the error model is demonstrated by presenting its use as a data discriminator and by examination of time series of sliding boxcar samples of radar data. Baseline difference std for data rejected by a threshold based on the error model is shown to be significantly higher than for the data retained. The results presented here demonstrate potential to improve assessment of the HF radar current measurement uncertainty. Such improvement has potential to benefit all applications of HF radar data, including for example, Lagrangian particle tracking and surface current assimilation into numerical models.