Eulerian comparison between lagrangian drifters velocities and sea surface currents reconstructed from the SWOT swath in the Mediterranean Sea

IGE - MEOM team internship (June-July 2024)

Leo Boux de Casson leo.boux_de_casson@ens-lyon.fr

Internship supervisors: Julien Le Sommer and Vadim Bertrand

---

• Dataproducts used:

  • Mapped SSH field in the mediterranean sea

    • Mihost allnadirs SSH

    • Mihost allnadirs+swot SSH

  • SWOT swath SSH product

    • SWOT_L3_LR_SSH_Expert v0.3

  • SVP Drifters data from the BioSWOT-Med and C-SWOT-2023 missions

    • L2 processed data

• SSC are calculated thanks to the jaxparrow module https://github.com/meom-group/jaxparrow (Vadim Bertrand)

---

• Repo Structure:

  • load_field_dataset.py : load the field dataproduct in a standardize xarray dataset format

  • Drifters_filters/ : notebooks with tests of several filters, the final decision is to filter only the velocities with the scipy low_pass filter

    • wavelet filter for positions and velocities
    • gaussian filter for positions, best is gaussian_std = 7hours
    • scipy low_pass filter, best is cutoff = 48h

  • Field_processing_and_plots/ :

    • notebooks to see the spatial/temporal field filtering and to have plots and animations of the geostrophic/cyclogeostrophic obtained thanks to jaxparrow
    • the best filter is gaussian_std = 3.5km

  • L4_eulerian_comparison/ : notebooks with the eulerian comparison of drifters velocities and Mihost fields to see the amelioration of adding swot data in the Mihost mapped SSH

  • L3_eulerian_comparison/ :

    • MAIN_save_filtered_field_and_Generate_scenes.py : file to generate a "scenes dataset" with a given field and drifters filtering. Scenes are 3-days long drifters/field comparison to see what is happening at small scales (~10km)
    • notebooks to see some GOOD and BAD scenes, based on the metric $\frac{||U_{field} - U_{drifter}||}{||U_{drifter}||}$ and also see the overall eulerian comparison results in the swath

  • save_data/ :

    • input data are in fields_data/ and drifters/
    • Best filtering Scenes data are in drifters_scenes/

  • save_plots/ : Results plots, with movies of the SWOT swath (SSH, Vorticity...), filtering PSD plots and the 20 best and 20 worst scenes in L3_Scenes/L3_Scenes_GOODBAD_best_filtering