Table Of Contents

This Page

12.15.10: Attempt to isolate submesoscale density anomalies

I show here an attempt to isolate submesoscale potential density anomalies σduring the second deployment of WHOTS (wh02a; the first deployment is not used as the APDRC-ARGO product is not available for its entire period). I first remove the seasonal cycle using the 3-month running-mean σ from the monthly-mean APDRC-ARGO product. I then extract AVISO SSH and find the time lag between it and σ (the vertical average between 100 and 150 m of σ is used hereafter). A lag of 11 days is found 9Fig. 1a). The lagged AVISO SSH is then used to find the linear regression with it and σ (Fig. 1b). This can be seen as the mesoscale σ and when we remove it, we define a submesoscale σ (Fig. 1c). Unfortunately, there is still some mesoscale structure left in that submesoscale signal which may make that quantity useless.

../../../../../../_images/meso_submeso_PD_wh02a.png

Figure 1: (a) Original AVISO SSH (black) at the location of the WHOTS mooring and its 11-day lagged version (red). (b) The potential density anomaly σ (defined after removing the 3-month running-mean σ from the monthly-mean APDRC-ARGO product) vertically averaged between 100 and 150 m (black) and its “mesoscale” component deduced from linear regression of the lagged AVISO SSH product. (c) The “submesoscale” component of σ deduced by removing its “mesoscale” component: (green) deduced from the 1-day averaged σ WHOTS time series, (blue) deduced from re-sampling the 1-day averaged time series every 5 days to mimic ARGO time series.


WHOTS_isolate_submesoscale.m in RESEARCH/PROJECTS/MARINE_BIOLOGY/SUBMESOSCALE_PROCESSES/WHOTS/analysis on ipu1.