Table Of Contents

This Page

Comparison between a 1/30 and a 1/10th-of-a-degree OFES snapshot

We compare in this note a snapshot from the 1/30th-of-a-degree OFES model to one from the 1/10th-of-a-degree OFES model. The two snapshots are from Jan. 3, 2001 but note that, as we are focusing on turbulent features such as eddies, we compare only the snapshots qualitatively. We also focus on a small domain, the same domain that has been used recently to survey shallow/deep events of isopycnals from ARGO floats and in the 1/10th-of-a-degree OFES model.

We start first with some quantities at the surface (Figs. 1 to 4): the sea surface height anomaly (SSHA; defined as SSH minus the spatial averaged SSH over the shown domain), the relative vorticity ζ, Rossby number R and Okubo-Weiss parameter (OW).

A first difference is that, as expected, submesoscale features appear to be more energetic in the 1/30th than in the 1/10th-of-a-degree OFES model (Figs. 2 and 3). In the snapshot from the 1/30th-of-a-degree OFES model, there are several energetic submesoscale features: one in the southwest corner of the domain, another in the south-southeast corner, both where the gradient of SSHA is relatively large. In the snapshot from the 1/10th-of-a-degree OFES model, there is one energetic submesoscale feature but this one does not seem to related to a strong gradient of SSHA.

Surface OW is plotted in Fig. 4. An interesting feature is the pattern of OW along the energetic submesoscale filament in the southwest corner of the domain: OW alternates between small strain-dominated domain and small vorticity-dominated domain. This is also true for the energetic submesoscale flow in the 1/10th-of-a-degree OFES model. How does this pattern happen? In Fig. 5 is shown a focus on the flow field of the energetic submesoscale of the 1/30th-of-a-degree OFES model, together with OW, while Fig. 6 shows the vorticity and strain component of OW, together with OW. The submesoscale feature is a southeastward-flowing jet that has both strong strain and vorticity regions on each side. Strain and vorticity vary in the direction of the flow and they reach maximum at different positions of the jet: when one side of the jet is dominated by vorticity, the other side is dominated by strain. These alternations result in a checkerboard pattern in OW.

Another interesting feature is that OW does not, here, separate between inside, at the edge of or outside eddies. This suggests that this qualitative separation used in previous work might work only when it is computed from the mesoscale SSH, excluding in particular the submesoscale. Fig. 7 shows SSHA and low-pass filtered SSH for wavelengths longer than 150 km in the 1/30th-of-a-degree OFES snapshot. Fig. 8 compares surface ζ with the contour of SSH (the filtered one is shown but it does not matter for the argument) in the upper panel, while in the lower panel is shown OW computed from filtered SSH, together with filtered SSH. Using filtered SSH, OW now succeeds to separate between the inside of eddies with the outside but, given the example of the energetic submesoscale, we see that this OW fails to locate where the submesoscale is. This suggests that OW, from either the full SSH or from a filtered version of it, might not be the appropriate parameter to distinguish the inside of eddies with locations of submesoscale processes.

../../../../../../_images/SSHA_30_10_OFES_Jan12001_Jan32001.png

Figure 1: Sea surface height anomaly (SSHA; defined as SSH minus the spatial averaged SSH over the shown domain) in (a) the 1/30th-of-a-degree OFES model and (b) the 1/10th-of-a-degree OFES model on Jan. 3, 2001. See read_hr_OFES.m in RESEARCH/PROJECTS/MARINE_BIOLOGY/SUBMESOSCALE_PROCESSES/OFES/hr_OFES.

../../../../../../_images/ZETA_30_10_OFES_Jan12001_Jan32001_surf_2.png

Figure 2: Surface (2.5 m depth) relative vorticity ζ in (a) the 1/30th-of-a-degree OFES model and (b) the 1/10th-of-a-degree OFES model on Jan. 3, 2001. SSHA is replotted in black contours: positive SSHA (plain) and negative SSHA (dots). Contours are every 1 cm, between -10 and +10 cm.

../../../../../../_images/R_30_10_OFES_Jan12001_Jan32001_surf.png

Figure 3: Surface (2.5 m depth) Rossby number R in (a) the 1/30th-of-a-degree OFES model and (b) the 1/10th-of-a-degree OFES model on Jan. 3, 2001.

../../../../../../_images/OW_30_10_OFES_Jan12001_Jan32001_surf.png

Figure 4: Surface (2.5 m depth) Okubo-Weiss parameter (OW) in (a) the 1/30th-of-a-degree OFES model and (b) the 1/10th-of-a-degree OFES model on Jan. 3, 2001.

../../../../../../_images/flow_OW_submesoscale_1_30th_OFES.png

Figure 5: Flow (arrows) and OW (contours every 10e-11 1/s2; positive in plain, negative in dashed lines) for the energetic submesoscale feature revealed in Fig. 3.

../../../../../../_images/ZETA2_S2_OW_submesoscale_1_30th_OFES.png

Figure 6: Squared relative vorticity (a) and strain (b) in colors, and OW (contours every 10e-11 1/s2; positive in plain, negative in dashed lines) for the energetic submesoscale feature revealed in Fig. 3.

../../../../../../_images/SSHA_filtSSH_1_30_OFES.png

Figure 7: (a) SSHA (as defined in Fig. 1a) and (b) SSH filtered to exclude wavelengths longer than 150 km in the 1/30th-of-a-degree OFES snapshot.

../../../../../../_images/ZETA_filtSSH_gOWfromfiltSSH_1_30_OFES.png

Figure 8: (a) Surface ζ and (b) OW computed from the filtered SSH of Fig. 7b. The filtered SSH is shown in plain (positive) and dotted (negative) contours, with contours every 2 cm.

We then have a look at a zonal section, one along 21°N in the 1/30th-of-a-degree OFES model and one along 22.5°N in the 1/10th-of-a-degree OFES model. In both models, the energetic submesoscale features are trapped in the surface layer (SML). Strong vertical velocities are associated with these submesoscale features in the SML (Fig. 12). Below it, vertical velocities are still large but are mesoscale and it is not clear what causes these. As described by Lapeyre and Klein (2006), there is little signature of submesoscale features in the density field (Fig. 13).

../../../../../../_images/ZETA_30_10_OFES_Jan12001_Jan32001_21N_22_5N.png

Figure 9: ζ along (a) 21°N in the 1/30th-of-a-degree OFES model and (b) 22.5°N in the 1/10th-of-a-degree OFES model on Jan. 3, 2001.

../../../../../../_images/R_30_10_OFES_Jan12001_Jan32001_21N_22_5N.png

Figure 10: R along (a) 21°N in the 1/30th-of-a-degree OFES model and (b) 22.5°N in the 1/10th-of-a-degree OFES model on Jan. 3, 2001.

../../../../../../_images/OW_30_10_OFES_Jan12001_Jan32001_21N_22_5N.png

Figure 11: OW along (a) 21°N in the 1/30th-of-a-degree OFES model and (b) 22.5°N in the 1/10th-of-a-degree OFES model on Jan. 3, 2001.

../../../../../../_images/w_30_10_OFES_Jan12001_Jan32001_21N_22_5N.png

Figure 12: Vertical velocity (w) along (a) 21°N in the 1/30th-of-a-degree OFES model and (b) 22.5°N in the 1/10th-of-a-degree OFES model on Jan. 3, 2001.

../../../../../../_images/sigma_30_10_OFES_Jan12001_Jan32001_21N_22_5N.png

Figure 13: Potential density σ along (a) 21°N in the 1/30th-of-a-degree OFES model and (b) 22.5°N in the 1/10th-of-a-degree OFES model on Jan. 3, 2001.

Questions:

  • Which vertical velocities matter for the nitrate, the one associated with the submesoscale features in the SML or the one below? In the second case, what causes these large vertical velocities at depth?
  • Which index of submesoscale features indicate where vertical velocities are large?
  • Is there an index using AVISO-like (mesoscale) SSH that can detect the presence of submesoscale features?