05.07.11: Animations of SST gradient, FSLE and stratification for WHOTS site (Nov. 1, 2008-Mar. 31, 2010)¶
In the two animations below are shown the sea surface temperature (SST) gradient or SST filaments only together with the finite-size Lyapunov exponents (FSLE; black contours) and the Brunt-Vaisala frequency (BVF; lower panel) at WHOTS site from Nov. 1 2008 to Mar. 31, 2010 (indicated by a red and blue dots in the first and second animation, respectively).
- With SST gradient
- With SST filaments only (the SST filaments are defined as locations where SST gradient is larger than 1.2e-4°C/m)
In both animations, the SST gradient has been smoothed by a Gaussian filter of 0.2°width. the vertical dashed white line in the lower panel indicates the day for which the SST gradient or filament and the FSLE are shown in the upper panel.
From the first animation, we see that the SST gradient has the same scale and amplitude than those simulated in the HYCOM regional numerical simulation. There is also some rapid (days) and large-scale change of the SST gradient that may be due to either 1) oceanic barotropic waves passing through, 2) atmospheric perturbations or 3) unrealistic effect in the SST product. More importantly, visual inspection of the upper panel suggests that more often than not high SST gradient are located within a FSLE filament. This seems particularly true in Jan.-Mar. and in particular in 2010. An example is given in Fig. 1 for Feb. 8, 2010:
Figure 1: (upper) SST gradient and FSLE on Feb. 8, 2010 and (lower) WHOTS BVF from Nov. 1 2008 to Mar. 31, 2010. The vertical dashed white line in the lower panel indicates Feb. 8, 2010.
Visual inspection can be misleading; both SST and FSLE filaments have a high density and our eye is trained to see pattern –even when there is none. The second animation only shows the SST filaments (defined as locations where SST gradient is larger than 1.2e-4°C/m) with the FSLE filaments and, indeed, suggests it is much more difficult to see now a pattern in SST gradient that is similar to a pattern in FSLE. There are some days where there may be some correlation (as on Feb. 8, 2010; Fig. 2) but a more objective method is needed to assess the correlation.
Figure 2: As in Fig. 1 except that only SST gradient larger than 1.2e-4°C/m are plotted in the upper panel.
For the first animation, I used anim_FSLE_SSTgrad_Nov08_Mar10.py (to use only from ulua in order to reach the MUR SST product) in RESEARCH/PROJECTS/MARINE_BIOLOGY/SUBMESOSCALE_PROCESSES/MUR_SST/analysis/anim_FSLE_SSTgrad_Nov08_Mar10 on the central disk. For the second animation, I used anim_FSLE_SSTgrad_Nov08_Mar10_2.py (also to use inly from ulua) in RESEARCH/PROJECTS/MARINE_BIOLOGY/SUBMESOSCALE_PROCESSES/MUR_SST/analysis/anim_FSLE_SSTgrad_Nov08_Mar10_2 on the central disk.