Aerospace and Electronic Systems Magazine December 2017 - 16

OTH Radar Phenomenology
recurrence, with energy being restored to the primary wave before the process repeats. Exactly this
behaviour is evident in Figure 16, where the recurrence length is around 10-13 km, which compares
well with the predicted range 8-18 km, under the
prevailing conditions. Nevertheless, the recurrence
interpretation must be treated with caution, though
explanations such as atmospheric gravity waves
and tidal interference patterns were shown not to
be the cause. The other leading candidate explanation - nonlinear effects in wave group evolution
- is not entirely convincing, given the length and
time scales involved. But irrespective of the precise
mechanism, complex physics is generating a texture
on the wave field, and our models need to recognize
this.
One of the many advantages of the Hamiltonian formulation is that the numerical evaluation of
these solutions, while still challenging, is far easier
than the corresponding task for the standard NavierStokes equation, being directly aided by the fact
that, at any order, only a minority of terms contribute significantly, though even this statement needs
to be read in the context of modelling of the role of
nonresonant wave interactions, as explored by Annenkov and Shrira [18].
The discussion above deals with the case of a homogeneous fluid. In reality the ocean is a stratified
medium with pronounced vertical profiles of temperature and salinity. If we adopt the simple model
of a two-layer fluid, the possibility of waves along
the density interface or pycnocline emerges, and such
waves, commonly referred to as internal waves, are
prevalent in certain parts of the world's oceans. Once
again, a Hamiltonian formulation describes the situation elegantly, and once again the dynamics can be
established. HF radar signals do not penetrate to the
pycnocline, but the coupling between the two layers,
expressed in the Hamiltonian, predicts that the surface
gravity waves sensed by HF radars will be modulated
by the internal waves. And, sure enough, that is the
case, as reported in [19] and illustrated in Figure 17.

TARGET CLASSIFICATION

Figure 16.

Apparent radar signature of Fermi-Pasta-Ulam recurrence, observed in Gulf St Vincent
when the wave field was highly directional. The solid line is a measure of the power spectral density of a long gravity wave, based on the amplitude of its bound second harmonic.
The dashed line indicates total received power, with the expected fall-off with range.

Figure 17.

Apparent signatures of internal wave packets generated by the strong M2 tidal flow onto
the continental shelf in North West Australia. The colour scale indicates significant wave
height. Both the spacing and the 2 dB increase in scattering coefficient of the striations are
consistent with modelling results.

We have seen how a more general model of the
phenomenology reveals that the target return is
distributed over the multidimensional analysis domain. For this
reason there is some merit in adopting a more pragmatic definition.
Accordingly, we can define the generalised radar signature
(GRS) of an object x with respect to a radar Q as [20]:
GRS ( x ) = response of Q when x is present
− response of Q when x is absent

(9)

with the implied dependence on parameters representing the
geometry and the propagation environment. When it comes to
attempting to discriminate between targets on the basis of their
GRS, the characteristics and configuration of the radar system
inevitably enter the picture, along with the properties of the
propagation paths. As the value of OTH information is greatly
reduced in the absence of a target classification capability, these
considerations can be of paramount importance, especially in
the design of signal analysis and interpretation systems and in

IEEE A&E SYSTEMS MAGAZINE

DECEMBER 2017

Table of Contents for the Digital Edition of Aerospace and Electronic Systems Magazine December 2017