12.03.2012: Wavelet analysis in potential density for HOE-DYLAN 5 and 9¶

Below, I show the wavelet analysis of the potential density anomaly for HOE-DYLAN 5 and 9. The first figure shows the temporal average of the wavelet power spectrum (WPS) in order to have an idea of how the WPS varies in the vertical. Then, I show the WPS for three different depths (120 m, 175 m and 250 m).

The semi-diurnal signal is the dominant signal below the surface mixed layer during both cruises (Figs. 1 and 6). Its amplitude peaks between 150 and 300 m but I am not sure this is a real feature or an artifact of the wavelet analysis (note that the results do not change much when we calculate the WPS for the potential density anomaly weighted by the buoyancy frequency, that is, divided by N^(3/2) where N is the local averaged buoyancy frequency).

Interestingly although the signal is there during the whole of HOE-DYLAN 5, it appears mostly in the second part of HOE-DYLAN 9. This increase in the amplitude of the semi-diurnal tide has been visually observed in the time series of the potential density (not shown) and was noticed to coincide with a decrease in the overall level of chlorophyll (not shown). Notice, however, that it is difficult to see the expected spring-neap modulation of the semi-diurnal signal possibly indicating that there is a lot of noise in the wavelet analysis. At this point, we will only mention that rapid (days to weeks) changes in the amplitude (and phase) of internal tides are possible due to their interactions with the submesoscale and mesoscale field (e.g. Chavanne et al. 2010). Further work is needed (I am thinking about the analysis of the horizontal velocity field) to assess how the semi-diurnal tide changes with time during HOE-DYLAN 9.

In both cruises, there is also energy at the near-inertial period (Figs. 2, 3, 4, 7, 8 and 9). During HOE-DYLAN 5, a near-inertial event may have propagated downward from around 100-150 m depth at the beginning of the cruise to around 200-250 m depth at the end (Fig. 5). Unlike during HOE-DYLAN 5, the near-inertial signal during HOE-DYLAN 9 is not in general particularly larger than signals at other periods.

HOE-DYLAN 5¶

../../../../../../_images/WPS_tav_HD5.png

Figure 1: Time-averaged wavelet power spectrum (WPS) of potential density anomaly during HOE-DYLAN 5.

../../../../../../_images/WPS_120m_HD5.png

Figure 2: WPS of potential density anomaly at 120 m during HOE-DYLAN 5. Periods on the vertical axis are in hours.

../../../../../../_images/WPS_175m_HD5.png

Figure 3: WPS of potential density anomaly at 175 m during HOE-DYLAN 5. Periods on the vertical axis are in hours.

../../../../../../_images/WPS_250m_HD5.png

Figure 4: WPS of potential density anomaly at 250 m during HOE-DYLAN 5. Periods on the vertical axis are in hours.

../../../../../../_images/WPS_inertial_HD5.png

Figure 5: WPS near the inertial period during HOE-DYLAN 5. The black arrow indicates a possible downward (energy) propagation of the near-inertial signal.