ocean surface currents in smaller domains (e.g., lagoons, ports and harbors).
Applications of SAR systems are then presented with case studies in mapping
wind, oil spills and other ocean features. Following this, scatterometers are
described for wind measurement, including a case study of doldrums off the west
coast of the central Americas. Finally, the use of X-band radar for two-dimensional
wave fields is described.
13.2 HF Ocean Radar
13.2.1 Analysis and Classification Techniques
The starting point for analysis of HF ocean radar echoes to extract sea surface
parameters is the Doppler shift spectrum. Examples are shown in Fig. 13.1 for the
two most common types of HF ocean radar systems: phased-array (Fig. 13.1a) and
crossed-loop direction-finding (Fig. 13.1b). In these graphs the dominant peaks are
due to first-order Bragg scatter, and the rest of the energy above the noise background is due to double-scatter mechanisms and higher order non-linearities in the
waves (see Chap. 11). In both systems the radial component of the current towards
the radar station is determined, and a second radar looking at a near-orthogonal
angle is required to produce the surface current vector. The phased-array approach
gives fine spatial and temporal resolution because measurements are made in
individual pixels. In contrast, the direction-finding approach has more compact
antennas but generally gives a broader view of the surface current field.
For long-range HF ocean radars operating up to 200 km, the errors in measuring
the radial components of the surface currents vary on the order of 5–20 cm/s.
Chapman and Graber (1997) compared radar surface currents with those from the
top usable bin of a moored acoustic profiler and found rms differences of 15 cm/s.
Kohut et al. (2006), however, pointed out that some of this difference is due to
shears in the flow between the effective depth of the radar measurements and the
uppermost usable bin on the acoustic profiler, concluding that the radar error was
just 5 cm/s. Cook et al. (2007) estimated the error in crossed-loop systems using a
simulation method and found that the errors were 6 cm/s at short ranges but
increased to over 20 cm/s at long ranges where the signal-to-noise ratios are
reduced. In comparison, the phased-array systems have errors of typically 6–12 cm/
s and are invariant with range.
The spatial resolution of HF radar varies from about 1 km at short ranges,
where it is limited by the operating bandwidth of the radar, to about 8–20 km at
long ranges, where it is limited by the accuracy of the radar azimuth. Temporal
resolutions vary from about 10 min to 3 h depending on the type of radar and the
configuration.
Raw HF radar data are susceptible to inclusion of outliers due to a wide range of
phenomena, including radio interference, spurious echoes from ships, and local
effects like animals or vehicles affecting the performance of individual antennas.
13 Radar Applications
343
Applications of SAR systems are then presented with case studies in mapping
wind, oil spills and other ocean features. Following this, scatterometers are
described for wind measurement, including a case study of doldrums off the west
coast of the central Americas. Finally, the use of X-band radar for two-dimensional
wave fields is described.
13.2 HF Ocean Radar
13.2.1 Analysis and Classification Techniques
The starting point for analysis of HF ocean radar echoes to extract sea surface
parameters is the Doppler shift spectrum. Examples are shown in Fig. 13.1 for the
two most common types of HF ocean radar systems: phased-array (Fig. 13.1a) and
crossed-loop direction-finding (Fig. 13.1b). In these graphs the dominant peaks are
due to first-order Bragg scatter, and the rest of the energy above the noise background is due to double-scatter mechanisms and higher order non-linearities in the
waves (see Chap. 11). In both systems the radial component of the current towards
the radar station is determined, and a second radar looking at a near-orthogonal
angle is required to produce the surface current vector. The phased-array approach
gives fine spatial and temporal resolution because measurements are made in
individual pixels. In contrast, the direction-finding approach has more compact
antennas but generally gives a broader view of the surface current field.
For long-range HF ocean radars operating up to 200 km, the errors in measuring
the radial components of the surface currents vary on the order of 5–20 cm/s.
Chapman and Graber (1997) compared radar surface currents with those from the
top usable bin of a moored acoustic profiler and found rms differences of 15 cm/s.
Kohut et al. (2006), however, pointed out that some of this difference is due to
shears in the flow between the effective depth of the radar measurements and the
uppermost usable bin on the acoustic profiler, concluding that the radar error was
just 5 cm/s. Cook et al. (2007) estimated the error in crossed-loop systems using a
simulation method and found that the errors were 6 cm/s at short ranges but
increased to over 20 cm/s at long ranges where the signal-to-noise ratios are
reduced. In comparison, the phased-array systems have errors of typically 6–12 cm/
s and are invariant with range.
The spatial resolution of HF radar varies from about 1 km at short ranges,
where it is limited by the operating bandwidth of the radar, to about 8–20 km at
long ranges, where it is limited by the accuracy of the radar azimuth. Temporal
resolutions vary from about 10 min to 3 h depending on the type of radar and the
configuration.
Raw HF radar data are susceptible to inclusion of outliers due to a wide range of
phenomena, including radio interference, spurious echoes from ships, and local
effects like animals or vehicles affecting the performance of individual antennas.
13 Radar Applications
343