Towards operational 3D-VAR assimilation of chlorophyll biogeochemical-Argo float data into a Biogeochemical Model of the Mediterranean Sea (2019)

by G. Cossarini 1*, L. Mariotti 1, L. Feudale 1, A.Mignot 2,3, S.Salon 1, V.Taillandier 3, A.Teruzzi 1  and F. D’ortenzio 3

         

1 Istituto Nazionale di Oceanografia e di Geofisica Sperimentale (OGS), Borgo Grotta Gigante 42/c, Sgonico, Italy
2 Mercator Ocean International, Ramonville-Saint-Agne, 31520, France
3 CNRS, Laboratoire d’Océanographie de Villefranche, LOV, Sorbonne Universités, Villefranche-sur-Mer, France
 
 
* Corresponding author : gcossarini@inogs.it 
Full paperCossarini, G., Mariotti, L., Feudale, L., Mignot, A., Salon, S., Taillandier, V., Teruzzi, A., D'Ortenzio, F. (2019). “Towards operational 3D-Var assimilation of chlorophyll Biogeochemical-Argo float data into a biogeochemical model of the Mediterranean Sea” -  Ocean Modelling vol. 133, pp 112-128, https://doi.org/10.1016/j.ocemod.2018.11.005
Context

Today, our ability to forecast the state of the ocean, as we can predict the weather tomorrow, is an essential response to some real economic and societal challenges (Le Traon et al., 2017). The recent performance of digital computers, combined with increasingly numerous and accurate observations (e.g. satellite data), allows us the modelling and prediction of the physical and biogeochemical state of the ocean. The in-situ data, and especially those recently collected thanks to autonomous platforms such as Argo floats, refine the fit of numerical models and play a key role in forecast performance.

Recently, the new biogeochemical (BGC) Argo network is a new challenge for operational oceanography (Johnson and Claustre, 2016). Thirty-nine BGC-Argo floats equipped with biogeochemical sensors were deployed from 2012 to 2017 in the Mediterranean Sea, thus making this network one of the densest of the global ocean. 

The Mediterranean Sea is characterized by peculiar vertical phytoplankton characteristics, with a marked seasonal cycle of surface and deep chlorophyll maximum (DCM) along with a noticeable west-east deepening of the DCM and trophic gradient (Lavigne et al., 2015; Mignot et al., 2014). That makes BGC-Argo vertical profiles very helpful for tracking those significant biogeochemical processes in the basin. 

 In this study, chlorophyll data collected from BGC-Argo floats during year 2015 in the Mediterranean Sea are assimilated into an operational ocean forecasting system in order to investigate the interim structures and dynamics of marine ecosystems at a regional scale. The BGC-Argo float assimilation is particularly discussed with the aim of showing its capability to correct the vertical structure of modelled chlorophyll. 

 Preliminary results show that the assimilation of observed chlorophyll vertical profiles from BGC-Argo floats in an operational modeling system is feasible and provides significant corrections to chlorophyll concentrations and vertical features.

 

This study has been leaded by OGS and LOV teams in the framework of Copernicus Marine Environment Monitoring Service Evolution (CMEMS-SE 2015-2018) MASSIMILI project. The assimilation of BGC-Argo data (i.e. chlorophyll and nitrate) will be operational in the CMEMS Mediterranean biogeochemical forecast system starting from 2020. CMEMS is implemented by Mercator Ocean in the framework of a delegation agreement with the European Union.

Data & Method

The operational Mediterranean biogeochemistry model (Bolzon et al., 2017) of the CMEMS system (Le Traon et al., 2017) is used in this study for the assimilation of observed chlorophyll vertical profiles from BGC-Argo floats.  

The assimilation scheme adopted is a variational method (3DVarBio), which is the current assimilation scheme in use in the CMEMS MED-MFC Biogeochemistry component. Some key operators of the 3DVarBio code have been re-designed and optimized for the assimilation of the chlorophyll vertical profiles. 

Figure 1: Positions of the BGC-Argo float profiles in 2015 (point colors indicate months); identification numbers are shown by the last 3 digits (e.g., 649 instead of 6901649) and show floats with chlorophyll data (black) and with chlorophyll and nitrate data (blue). The division of the Mediterranean Sea into sub-basins (names in italics) is indicated by the grey lines. From Cossarini et al., 2019.

In model experiments that cover year 2015, 25 floats with chlorophyll data are available and located mainly in the western Mediterranean, the northern Ionian Sea and the central part of the Levantine basin (figure 1). The measured chlorophyll profiles are processed with quality control (QC) methods in real-time and delayed mode with specific algorithms. In addition, new quality control methods have been developed for this study and are presently endorsed at the international level (Schmechtig et al., 2016).

Main Results 

Chlorophyll data from four specific biogeochemical-Argo floats are studied (figure 2) in order to analyse in details the effects of BGC-Argo floats assimilation on chlorophyll vertical patterns, and the capability of assimilation to correct the vertical modeled chlorophyll structures. Two floats (wmo’s numbers 491 and 600, figure 1) drifted in the western part of the basin, while the two others drifted in the eastern part (wmo’s numbers 528 and 773, figure 1). 

Figure 2. Hovmoller diagrams of the assimilation (upper) and the reference (middle) simulations and their differences (lower) and the timing of the 52 assimilation steps (dashed white vertical lines) for four selected BGC-Argo floats: 6901491 (a), 6901528 (b), 6901600 (c) and 6901773 (d). Hovmoller diagrams are calculated by averaging chlorophyll concentration in an area of radius 30 km surrounding the float trajectories; x-labels report the first day of each month in 2015. From Cossarini et al., 2019.

 

The upper/middle plots exemplify the behavior of phytoplankton dynamics simulated by the Biogeochemical Flux Model (BFM), which are consistent with what described in Lavigne et al., 2015: 

-        a winter surface bloom, during conditions of vertical mixing

-        a summer deep chlorophyll maximum, during summer stratified conditions 

-        and a clear decreasing trophic gradient from the western (a and b) to the eastern (c and d) Mediterranean Sea

 

The major changes introduced by the assimilation during winter and early spring months affect the intensity and timing of the chlorophyll surface bloom and to a lesser degree, the depth of the vertically mixed bloom (lower plots, fig. 2). Further, during summer, the BGC-Argo data assimilation provides fine corrections of the vertical displacement of the DCM induced by mesoscale dynamics. This study is the first step in demonstrating the great value of collecting vertical in-situ biogeochemical profiles to resolve particular vertical phytoplankton dynamics in marine ecosystems. 

References
  • Bolzon, G., Cossarini, G., Lazzari, P., Salon, S., Teruzzi, A., Crise, A., Solidoro, C., 2017. Mediterranean Sea biogeochemical Analysis and Forecast (CMEMS MED AF- Biogeochemistry 2013–2017). [Data Set]. Copernicus Monitoring Environment Marine Service https://doi.org/10.25423/MEDSEA_ANALYSIS_FORECAST_BIO_006_ 006
  • Lavigne, H., D'ortenzio, F., d'Alcalà, M.R., Claustre, H., Sauzède, R., Gacic, M., 2015. On the vertical distribution of the chlorophyll a concentration in the Mediterranean Sea: a basin-scale and seasonal approach. Biogeosciences 12 (16), 5021–5039 2015. 
  • Le Traon, P.Y., Ali, A., Alvarez Fanjul, E., Aouf, L., Axell, L., Aznar, R., Ballarotta, M., Behrens, A., Benkiran, M., Bentamy, A., Bertino, L., Bowyer, P., Brando, V., Breivik, L.A., Buongiorno Nardelli, B., Cailleau, S., Ciliberti, S.A., Clementi, E., Colella, S., Mc Connell, N., Coppini, G., Cossarini, G., Dabrowski, T., de Alfonso Alonso-Muñoyerro, M., O'Dea, E., Desportes, C., Dinessen, F., Drevillon, M., Drillet, Y., Drudi, M., Dussurget, R., Faugère, Y., Forneris, V., Fratianni, C., Le Galloudec, O., García- Hermosa, M.I., García Sotillo, M., Garnesson, P., Garric, G., Golbeck, I., Gourrion, J., Grégoire, M.L., Guinehut, S., Gutknecht, E., Harris, C., Hernandez, F., Huess, V., Johannessen, J.A., Kay, S., Killick, R., King, R., de Kloe, J., Korres, G., Lagemaa, P., Lecci, R., Legeais, J.F., Lellouche, J.M., Levier, B., Lorente, P., Mangin, A., Martin, M., Melet, A., Murawski, J., Özsoy, E., Palazov, A., Pardo, S., Parent, L., Pascual, A., Pascual, A., Paul, J., Peneva, E., Perruche, C., Peterson, D., Petit de la Villeon, L., Pinardi, N., Pouliquen, S., Pujol, M.I., Rainaud, R., Rampal, P., Reffray, G., Regnier, C., Reppucci, A., Ryan, A., Salon, S., Samuelsen, A., Santoleri, R., Saulter, A., She, J., Solidoro, C., Stanev, E., Staneva, J., Stoffelen, A., Storto, A., Sykes, P., Szekely, T., Taburet, G., Taylor, B., Tintore, J., Toledano, C., Tonani, M., Tuomi, L., Volpe, G., Wedhe, H., Williams, T., Vandendbulcke, L., van Zanten, D., von Schuckmann, K., Xie, J., Zacharioudaki, A., Zuo, H., 2017. The Copernicus marine environmental monitoring service: main scientific achievements and future prospects. Spec. Issue Mercat. Océan J. #56. https://doi.org/10.25575/56.
  • Mignot, A., Claustre, H., Uitz, J., Poteau, A., D'Ortenzio, F., Xing, X., 2014. Understanding the seasonal dynamics of phytoplankton biomass and the deep chlorophyll maximum in oligotrophic environments: a Bio-Argo float investigation, Glob. Biogeochem. Cycl. 28 (8), 856–876. https://doi.org/10.1002/2013GB004781
  • Schmechtig, C., Thierry, V., The Bio Argo Team, 2016. Argo Quality Control Manual for Biogeochemical Data. The Bio Argo https://doi.org/10.13155/40879