Skip to content
DraconiticResearch
Search Sources API
203,203 studies31 sourcesUpdated 10 min ago
Peer-reviewedCC BY 4.0Biogeosciences

Solving calibration and reanalysis challenges of ocean biogeochemical dynamics with neural schemes: a 1D vertical model case-study

  • Jean Littaye1,2,3,4
  • Laurent Memery
  • Ronan Fablet
4 affiliations
  1. 1IMT Atlantique (Nantes, France)
  2. 2Laboratoire des Sciences de l'Environnement Marin (Plouzané, France)
  3. 3Centre National de la Recherche Scientifique (Paris, France)
  4. 4Ifremer (Brest, France)
Open at sourcePDFDOICite

Abstract

Numerous studies in climate and ocean sciences have highlighted the crucial role of ocean biogeochemical (BGC) models in studying and monitoring the global carbon cycle. Despite major advances due to both modelling and observation efforts, the quantification and reduction of the uncertainties in ocean BGC processes remain a key challenge. These difficulties arise primarily from the scarcity of observational datasets and the considerable uncertainties in ocean physics. Current ocean physics reanalyses still struggle to accurately represent the ocean's complex dynamics, particularly at small scales, which play a critical role in driving biogeochemical cycles. Consequently, the performance of operational ocean Data Assimilation (DA) systems remains limited when applied to BGC dynamics, using both BGC observations and physical reanalyses. This stands for model calibration and reanalysis applications.

Here, we explore machine learning approaches to address these challenges. To this end, we develop an Observing System Simulation Experiment (OSSE) framework for 1D ocean BGC dynamics, designed for both training and benchmarking purposes. We rely on a differentiable programming code of a 1D Nitrate-Ammonium-Phytoplankton-Zooplankton-Detritus (NNPZD) ocean BGC model forced by solar irradiance and vertical mixing. The proposed OSSE incorporates location-dependent uncertainties in physical forcings and considers realistic configurations of in situ observing systems. Based on these OSSEs, we design numerical experiments addressing both the calibration of BGC model parameters, the reconstruction of 1D ocean BGC state variables from sparse observations and the reduction of the uncertainties in the physical forcings. For calibration and inversion, we investigate a model-based variational DA scheme, an end-toend deep learning scheme and their hybrid combination. Our results demonstrate the potential of learning-based schemes to substantially reduce calibration uncertainties and improve physical forcing estimates. When coupled with a variational DA scheme, the learning-based approach yields enhanced reconstructions of ocean BGC state variables. Sensitivity analyses with respect to forcing uncertainties and observing system configurations provide insights into how these findings could be extended to real-world ocean BGC modelling and monitoring.

Versions

Related

  1. Satellites and Reanalysis Reveal the Dominance of Dynamics over Precipitation in Marine Cold-Air Outbreak Cloud Break-Up
    Filip Severin von der Lippe, Tim Carlsen, Trude Storelvmo +1
    EGUspherePreprintCC BY 4.0General earth scienceEGUsphere
  2. Linking Horizontal and Vertical Submesoscale Lagrangian Dynamics South of the Ibiza Channel
    Sara Cloux, Ismael Hernandez-Carrasco, Alejandro Orfila +2
    arXivPreprintCC BY-NC-ND 4.0Full textAtmospheric science & meteorologyFluid dynamicsarXiv
  3. Aerosol vertical distributions shaped by boundary layer dynamics in a coastal urban environment: insights from the TRACER campaign
    Bo Chen, Seth A. Thompson, Brianna H. Matthews +5
    Atmospheric Chemistry and PhysicsPeer-reviewedCC BY 4.0Atmospheric science & meteorologyCopernicusCrossref
  4. Current dynamics, drivers, and vertical thermal structure around Sir Bu Nair Island (Arabian/Persian Gulf)
    Geórgenes H. Cavalcante, Filipe Vieira, Rita Bento +2
    Ocean SciencePeer-reviewedCC BY 4.0Full textOceanography & cryosphereCrossref
  5. Editorial: Microorganisms for enhanced carbon sequestration and biogeochemical cycling during soil remediation
    Shu Wang, Yanling Wang, Yuxuan Wan +1
    Frontiers in MicrobiologyPeer-reviewedCC BY 4.0Full textHydrology & soilsCrossref

Cite

BibTeXRIS
BibTeX
@article{littaye2026solving,
  title = {{Solving calibration and reanalysis challenges of ocean biogeochemical dynamics with neural schemes: a 1D vertical model case-study}},
  author = {Jean Littaye and Laurent Memery and Ronan Fablet},
  journal = {Biogeosciences},
  year = {2026},
  month = oct,
  doi = {10.5194/bg-23-6879-2026},
  url = {https://hal.science/hal-05802951v1},
  note = {CC BY 4.0},
}
RIS
TY  - JOUR
TI  - Solving calibration and reanalysis challenges of ocean biogeochemical dynamics with neural schemes: a 1D vertical model case-study
AU  - Jean Littaye
AU  - Laurent Memery
AU  - Ronan Fablet
PY  - 2026
DA  - 2026/10/06
JO  - Biogeosciences
DO  - 10.5194/bg-23-6879-2026
UR  - https://hal.science/hal-05802951v1
AB  - Numerous studies in climate and ocean sciences have highlighted the crucial role of ocean biogeochemical (BGC) models in studying and monitoring the global carbon cycle. Despite major advances due to both modelling and observation efforts, the quantification and reduction of the uncertainties in ocean BGC processes remain a key challenge. These difficulties arise primarily from the scarcity of observational datasets and the considerable uncertainties in ocean physics. Current ocean physics reanalyses still struggle to accurately represent the ocean's complex dynamics, particularly at small scales, which play a critical role in driving biogeochemical cycles. Consequently, the performance of operational ocean Data Assimilation (DA) systems remains limited when applied to BGC dynamics, using both BGC observations and physical reanalyses. This stands for model calibration and reanalysis applications.  Here, we explore machine learning approaches to address these challenges. To this end, we develop an Observing System Simulation Experiment (OSSE) framework for 1D ocean BGC dynamics, designed for both training and benchmarking purposes. We rely on a differentiable programming code of a 1D Nitrate-Ammonium-Phytoplankton-Zooplankton-Detritus (NNPZD) ocean BGC model forced by solar irradiance and vertical mixing. The proposed OSSE incorporates location-dependent uncertainties in physical forcings and considers realistic configurations of in situ observing systems. Based on these OSSEs, we design numerical experiments addressing both the calibration of BGC model parameters, the reconstruction of 1D ocean BGC state variables from sparse observations and the reduction of the uncertainties in the physical forcings. For calibration and inversion, we investigate a model-based variational DA scheme, an end-toend deep learning scheme and their hybrid combination. Our results demonstrate the potential of learning-based schemes to substantially reduce calibration uncertainties and improve physical forcing estimates. When coupled with a variational DA scheme, the learning-based approach yields enhanced reconstructions of ocean BGC state variables. Sensitivity analyses with respect to forcing uncertainties and observing system configurations provide insights into how these findings could be extended to real-world ocean BGC modelling and monitoring.
SN  - 1726-4170
SN  - 1726-4189
ER  -