Séminaire ISTerre


Recent advances in geomagnetic data assimilation: a story of scientific and technical development

jeudi 6 février 2020 - 09h30
Loïc Huder - ISTerre (séminaire post-doc)
--- 
Data assimilation algorithms use numerical models to forecast a given quantity in time and physical measurements to correct this forecast (analysis step). Such algorithms are already commonplace in oceanology and meteorology but were considered rather recently in the study of geomagnetism.
The generation of Earth's magnetic field by flows in the outer core (geodynamo) is indeed a challenging problem that even state-of-the-art numerical simulations are not able to model fully. As numerical simulations become more and more Earth-like and as recent satellite missions (CHAMP, SWARM) give us unprecedented coverage and quality of geomagnetic data, the developement of data assimilation algorithms for the geodynamo is extermely timely and can give new insights in the evolution of the Earth magnetic field.
In this presentation, I will present our recent work on the scientific and technical improvement of geomagnetic data assimilation algorithms. I will review our recent scientific results in building magnetic models and reanalysed core flows, using our Python package pygeodyn. I will also show how good practices in technical development can be invaluable to ensure the scientific accuracy and perenability of scientific software in general.

Equipe organisatrice : Organisation labo

Salle Dolomieu, Maison des Géosciences, 38400 Saint Martin d'Hères