The geomagnetic main field, generated by geodynamo processes in the Earth's fluid outer core, plays an important role in shielding our habitat against solar wind and cosmic rays. Global reconstructions of geomagnetic field evolution on millennial timescales have broad applications e.g.
Both spatial and temporal resolution of millennial scale models are limited, due to a strongly heterogeneous distribution of available data and large data and dating uncertainties. However, neither model uncertainties, nor the sensitivity of the resulting model to prior assumptions, are well quantified so far. Another question is, wether the dipole contribution is rather independent from the smaller-scale field evolution. The aim of this project is to develop a new, correlation-based Bayesian modeling method for paleo- and archeomagnetic data.
The algorithm will be implemented as a readily adaptable and documented Python library. Furthermore, we will use this method to produce an improved Holocene field model which will, for the first time, include robust and realistic uncertainty estimates depending on data coverage in space and time.
A set of parameters for this model will be provided, as well as software for modeling various field quantities - including Gauss coefficients - together with estimates of uncertainities. The latter are particularly important when the model is used in data assimilation with the aim of predicting the future evolution of Earth's magnetic field.
2018 - 2021