14 Internal Solitary Waves System in the Mozambique Channel
277
e.g. Kundu and Cohen 2004). For the MC case, we have assumed a semi-diurnal
frequency for the IT with a M 2 tidal constituent (with σ = 1.4 × 10
−4 rad s
−1 ), and
a mean density profile typical for the southern winter stratification (namely October,
see Fig. 2a of da Silva et al. 2009). Based on Fig. 14.7, we also considered an average
ocean depth of 300 m to the west of the ISWs along their apparent propagation
path (region between the two consecutive packets that are observed in the SAR).
Additionally, we assumed that the two packets in Fig. 14.7 were generated at the
same phase of the tide and at the same position—that is near the impact point of the
IT beams with the pycnocline. Under those assumptions, and considering that the
two ISW packets in the SAR are separated along their apparent propagation path by
an average distance of 13 km (see black arrow on top of left panel in Fig. 14.7), the
estimated phase speed of the mode 2 ISWs in the SAR is 0.3 ms
−1 . On the other hand,
the TG model assuming an average depth of 300 m predicts linear phase speeds of
0.4 and 1.0 ms
−1 for mode 2 and mode 1 linear IWs of tidal frequency, respectively.
This further suggests that the waves in Fig. 14.7 are indeed mode 2 waves, since both
signatures (discussed in Sect. 14.3.2) and phase speeds are consistent with a mode 2
vertical structure.
14.4 Eddy-Like Structures and IW Refraction
Da Silva et al. (2009) reported that, on some occasions, ISWs in MC (off the Sofala
Bank; see Fig. 14.1 for location) were observed farther south than the average pattern.
In their study, the seasonal variability of currents during the southern winter was the
plausible explanation for this ISWs propagation anomaly. To justify this seasonal
variability they referred to the work done by Maltrud et al. (1998), who believed that
the Mozambique Current (which flows southward along the western slopes of the
basin) was much stronger during the austral winter. Based on the knowledge of the
Mozambique Current available at that time, da Silva et al. (2009) then proposed that
the increased southward flow in the southern winter months could explain why some
ISWs were refracted (or advected) towards more southerly pathways at that time of
year. However, more recent research indicates that the MC currents are in fact highly
variable, which is mainly due to the presence of eddy-like features that originate in
the Indian Ocean and travel southwards throughout MC (Ridderinkhof et al. 2010;
Ullgren et al. 2010). In this Section, we will show an example of ISWs propagating
through eddy-like structures (which are observed in altimetry data) and undergoing
dramatic refraction patterns.
Figure 14.8 shows a map of ISWs based on our interpretation of an Envisat
ASAR image (WS mode), acquired on 4 December 2009 at 20:07 UTC, that is
centred approximately at 21.0
◦ S and 36.3
◦ E. For clarity, two types of wave packets
(discussed earlier in Sect. 14.2) were marked with ‘SG’ for shelf-generation and ‘LG’
for local generation), until they reach 21.5
◦ S.
This is because further south of 21.5
◦ S along their trajectory, the interpretation
of the sea surface signatures starts to become ambiguous, mainly due to dissipation,
277
e.g. Kundu and Cohen 2004). For the MC case, we have assumed a semi-diurnal
frequency for the IT with a M 2 tidal constituent (with σ = 1.4 × 10
−4 rad s
−1 ), and
a mean density profile typical for the southern winter stratification (namely October,
see Fig. 2a of da Silva et al. 2009). Based on Fig. 14.7, we also considered an average
ocean depth of 300 m to the west of the ISWs along their apparent propagation
path (region between the two consecutive packets that are observed in the SAR).
Additionally, we assumed that the two packets in Fig. 14.7 were generated at the
same phase of the tide and at the same position—that is near the impact point of the
IT beams with the pycnocline. Under those assumptions, and considering that the
two ISW packets in the SAR are separated along their apparent propagation path by
an average distance of 13 km (see black arrow on top of left panel in Fig. 14.7), the
estimated phase speed of the mode 2 ISWs in the SAR is 0.3 ms
−1 . On the other hand,
the TG model assuming an average depth of 300 m predicts linear phase speeds of
0.4 and 1.0 ms
−1 for mode 2 and mode 1 linear IWs of tidal frequency, respectively.
This further suggests that the waves in Fig. 14.7 are indeed mode 2 waves, since both
signatures (discussed in Sect. 14.3.2) and phase speeds are consistent with a mode 2
vertical structure.
14.4 Eddy-Like Structures and IW Refraction
Da Silva et al. (2009) reported that, on some occasions, ISWs in MC (off the Sofala
Bank; see Fig. 14.1 for location) were observed farther south than the average pattern.
In their study, the seasonal variability of currents during the southern winter was the
plausible explanation for this ISWs propagation anomaly. To justify this seasonal
variability they referred to the work done by Maltrud et al. (1998), who believed that
the Mozambique Current (which flows southward along the western slopes of the
basin) was much stronger during the austral winter. Based on the knowledge of the
Mozambique Current available at that time, da Silva et al. (2009) then proposed that
the increased southward flow in the southern winter months could explain why some
ISWs were refracted (or advected) towards more southerly pathways at that time of
year. However, more recent research indicates that the MC currents are in fact highly
variable, which is mainly due to the presence of eddy-like features that originate in
the Indian Ocean and travel southwards throughout MC (Ridderinkhof et al. 2010;
Ullgren et al. 2010). In this Section, we will show an example of ISWs propagating
through eddy-like structures (which are observed in altimetry data) and undergoing
dramatic refraction patterns.
Figure 14.8 shows a map of ISWs based on our interpretation of an Envisat
ASAR image (WS mode), acquired on 4 December 2009 at 20:07 UTC, that is
centred approximately at 21.0
◦ S and 36.3
◦ E. For clarity, two types of wave packets
(discussed earlier in Sect. 14.2) were marked with ‘SG’ for shelf-generation and ‘LG’
for local generation), until they reach 21.5
◦ S.
This is because further south of 21.5
◦ S along their trajectory, the interpretation
of the sea surface signatures starts to become ambiguous, mainly due to dissipation,