Aerospace and Electronic Systems Magazine March 2018 - 44

Extraction of Wind Direction from the HF Hybrid Sky-Surface Wave Radar Sea Echoes
where KB = 2K0cos(ϕ0/2), θN is the ellipse normal direction and
φw represents the wind direction, which is considered to equal the
dominate wave direction for the fully developed sea state. Wyatt
et al. [14] suggest using hyperbolic secant [17] as the directional
spreading function, which is written as
g (ϕ ) = 0.5β sech 2 ( βϕ )

(4)

where β is the spreading parameter and φ is an angle referenced to
the mean wave direction. The distributions predict angular ranges
of r values, as shown in Figure 2. Using (3) and (4), the wind direction φw can be written as


N  
w

(5)

where ± indicates the ambiguity existing in the deduced wind direction and θ is the include angle between the wind direction and
the ellipse normal direction, given by the following:

θ=

R1/ 2e βπ − 1
1
ln
2 β 1 − R1/ 2e − βπ

Figure 2.

Bragg peak ratio as a function of angle for the hyperbolic secant model.

(6)

According to (5) and (6), the wind direction of HFHSSWR can be
calculated, but the direction ambiguity still exists.

VERIFICATION EXPERIMENT AND WIND DIRECTION
INVERSION
An experiment was carried out to test the sea state sensing capability of the HFHSSWR May 16-17, 2014. The transmitter and
receiver sites are deployed in Henan and Qingdao, China, respectively, as shown in Figure 3. The receiving antenna array comprises eight elements with an interval 7.5 m. The operating frequency
of this system is from 8 to 12 MHz, which was selected according
to the actual condition of trying to reduce the influence of ionosphere on the sea echoes. In addition, the coherent integration time
is within 2 min to reduce the influence of the ionosphere on the
quality of the data.
To obtain the in situ data and verify the radar measurements, a ship carrying an anemometer, current meter, and wave
buoy arrived and anchored at the appointed site A (35°45′34.92″
N, 120°55′57.00″ E) on May 16 and at site B (35°39′40.98″ N,
121°5′27.72″ E) on May 17 (Figure 3). In addition, the radar beam
directions of sites A and B are about 145° and 140°, respectively,
on the condition that the north azimuth is 0°.
The range-Doppler (R-D) spectrum of site A obtained at
08:25 on May 16 is shown in Figure 4. From this figure, we can
see the Bragg peaks shift, because of the ionosphere influence,
and the two directly arriving waves, which is because of the radio reflected by E-layer and F-layer ionosphere. Meanwhile, we
found that the Doppler spectrum from the E-layer reflection is
more stable than that from the F-layer reflection by analyzing the
experiment data. So the Bragg peaks associated with the E-layer
spectrum is chosen to retrieve the wind direction. Before the inversion of wind direction is carried out, the frequency shift com44

Figure 3.

Map of the transmitter and receiver locations and ship locations.

Figure 4.

R-D spectrum at site A obtained at 08:25 on May 16.

IEEE A&E SYSTEMS MAGAZINE

MARCH 2018

Table of Contents for the Digital Edition of Aerospace and Electronic Systems Magazine March 2018