validate this assumption experimentally and conclude that the HFR-derived velocity
corresponds to exponentially weighted velocity profile and that the equivalent
velocity value is found at the effective depth z = λ/8π. The effective depth is of the
order of 0.75 m for 16 MHz and 2.3 m for 5 MHz radar.
Mapping the surface velocity field Individual radar samples the surface velocity
over a large area which can extend up to 200 km offshore (5 MHz long-range
radar). This area is divided into a number of directional beams and each beam
contains a large number of bins (range cells). A good quality Doppler signal
requires each range cell to be larger than at least 150 Bragg wavelengths. For this
reason, the bin size becomes larger as the radar frequency decreases. A HFR cannot
unambiguously identify the arrival directions of backscattered signals. Currently,
HFR systems for mapping surface currents may be classified into two types,
beam-forming (BF) and direction-finding (DF), based on the method used to
determine arrival directions.
Beam-forming radars use a linear array of antenna elements that are steered by
adjusting the amplitudes and phases of received signals (e.g., [27]). An early beamforming (BF) radar, used extensively for mapping surface currents, was the ocean
surface current radar (OSCR) [20] with an azimuthal resolution of 10°. More
recently, the Wellen radar (WERA) was developed by Gürgel et al. [10] using
beam-forming configurations with up to 16 antenna elements offering an azimuthal
resolution of ∼3–4°. Such an array often occupies more than 150 m of coastline
which is not always practical.
To achieve smaller footprints, direction-finding (DF) configurations are used.
These radars compare the phases and amplitudes of radio signals received by
closely spaced antenna elements coupled with various direction-finding inversion
algorithms. The most common direction-finding radar is the CODAR SeaSonde [1].
Both algorithms (DF and BF) can be used to identify the arrival direction by
linear-array antenna WERA system. The azimuthal resolution produced by each
method can play a crucial role in analysis of surface circulation patterns in the
regions, where high horizontal velocity gradients are expected. Figure 1 provides
an example of such a difference in spatial structure of current velocity sampled by
12.4 MHz WERA system around the Ushant Island, off the western Brittany coast
of France, by using both DF and BF algorithms. The DF allows small-scale
structures of tidal currents to be better resolved. On the contrary, DF radars often
produce coverage gaps (cf. eastern and southern margin of the radar coverage area
in Fig. 1).
A single HFR measures projection of the current velocity vector on the beam
direction. Therefore, velocity measurements by at least two distant radars are
required for reconstruction of the vector velocity map. At a point of the beam
intersection of two radars, the east–west (u) and north–south (v) components of the
surface current vector may be found by solving a system of two equations involving
measured radial velocities and known beam angles. If a point on the sea surface is
in the coverage area of three or more radars, then u and v can be obtained from a
least-squares solution of the corresponding system of linear equations. These are the
522
A. Sentchev et al.
corresponds to exponentially weighted velocity profile and that the equivalent
velocity value is found at the effective depth z = λ/8π. The effective depth is of the
order of 0.75 m for 16 MHz and 2.3 m for 5 MHz radar.
Mapping the surface velocity field Individual radar samples the surface velocity
over a large area which can extend up to 200 km offshore (5 MHz long-range
radar). This area is divided into a number of directional beams and each beam
contains a large number of bins (range cells). A good quality Doppler signal
requires each range cell to be larger than at least 150 Bragg wavelengths. For this
reason, the bin size becomes larger as the radar frequency decreases. A HFR cannot
unambiguously identify the arrival directions of backscattered signals. Currently,
HFR systems for mapping surface currents may be classified into two types,
beam-forming (BF) and direction-finding (DF), based on the method used to
determine arrival directions.
Beam-forming radars use a linear array of antenna elements that are steered by
adjusting the amplitudes and phases of received signals (e.g., [27]). An early beamforming (BF) radar, used extensively for mapping surface currents, was the ocean
surface current radar (OSCR) [20] with an azimuthal resolution of 10°. More
recently, the Wellen radar (WERA) was developed by Gürgel et al. [10] using
beam-forming configurations with up to 16 antenna elements offering an azimuthal
resolution of ∼3–4°. Such an array often occupies more than 150 m of coastline
which is not always practical.
To achieve smaller footprints, direction-finding (DF) configurations are used.
These radars compare the phases and amplitudes of radio signals received by
closely spaced antenna elements coupled with various direction-finding inversion
algorithms. The most common direction-finding radar is the CODAR SeaSonde [1].
Both algorithms (DF and BF) can be used to identify the arrival direction by
linear-array antenna WERA system. The azimuthal resolution produced by each
method can play a crucial role in analysis of surface circulation patterns in the
regions, where high horizontal velocity gradients are expected. Figure 1 provides
an example of such a difference in spatial structure of current velocity sampled by
12.4 MHz WERA system around the Ushant Island, off the western Brittany coast
of France, by using both DF and BF algorithms. The DF allows small-scale
structures of tidal currents to be better resolved. On the contrary, DF radars often
produce coverage gaps (cf. eastern and southern margin of the radar coverage area
in Fig. 1).
A single HFR measures projection of the current velocity vector on the beam
direction. Therefore, velocity measurements by at least two distant radars are
required for reconstruction of the vector velocity map. At a point of the beam
intersection of two radars, the east–west (u) and north–south (v) components of the
surface current vector may be found by solving a system of two equations involving
measured radial velocities and known beam angles. If a point on the sea surface is
in the coverage area of three or more radars, then u and v can be obtained from a
least-squares solution of the corresponding system of linear equations. These are the
522
A. Sentchev et al.