FSLE with and without submesoscale flow

Here we test that the submesoscale features are indeed mostly due to the stretching and straning by the mesoscale flow, explaining why FSLE calculated from the mesocale flow (such as deduced from T/P SSH) can estimate the submesoscale features. We use Paulo’s output model of the horizontal velocity that contains submesoscale features. First, we show that the FSLE using the full spatial scale of the flow (Fig. 2) does isolate the submesoscale structure (revealed by the relative vorticity; Fig. 1). We show then that the FSLE result is closely the same if we use the mesoscale (>60km) flow alone (Fig. 3).

Figure 1: Relative vorticity of the horizontal flow from Paulo’s model output day 80.

Figure 2: Backward-in-time FSLE calculated for model day 80 in Paulo’s model (80 days with initial distance of 0.01° and final distance of 0.4°, using the 2-day velocity field at 10(?) m) using the full horizontal velocity field. Computed with RESEARCH/PROJECTS/MARINE_BIOLOGY/SUBMESOSCALE_PROCESSES/FSLE/analysis/corr_vert_vel_in_Calil_model/fsle_for_comp_with_w_Paulo_model.m.

Figure 3: As in Fig. 2 but using only the mesoscale (>60km) flow. Computed with RESEARCH/PROJECTS/MARINE_BIOLOGY/SUBMESOSCALE_PROCESSES/FSLE/analysis/corr_vert_vel_in_Calil_model/fsle_from_mesofield.m.