A summary of methods yielding information about the generation and configuration of the geomagnetic main field is presented with special focus on complications concerning these methods. A global source model constructed with the help of machine learning (and deep learning) is proposed to mitigate these issues, in particular the uncertainties caused by vigorous convection and small scale fields.

No abstract

Two-dimensional numerical calculations in cylindrical shell geometry have been carried out to investigate the effect of the temperature-dependent viscosity on the pattern and the characteristic parameters of the thermal convection occurring in the Earth's mantle. Systematic model runs established that the viscosity decreasing with the temperature is reduced around the hot core-mantle boundary (CMB) which facilitates 'the heat transport' from 'the core to the mantle'. On the other hand, the viscosity increases near the cold surface which hinders the heat outcome and results in higher mantle temperature and lower surface velocity. A power law relation was revealed between the strength of the temperature-dependence and the observed parameters, such as the velocity, surface mobility, heat flow, average temperature and viscosity. Two additional 'mantle-like' models were built up with extra strong temperature-dependent viscosity to imitate the flow in the Earth's mantle. In model 1, in which the viscosity decreases seven orders of magnitude with the temperature increase, a highly viscous stagnant lid evolves along the cold surface which does not participate in the convection. The existence of the stagnant surface lid reduces the surface heat flow and generates a low viscosity asthenosphere beneath the lid with vigorous small-scale convection. In model 2, in which the viscosity decreases only six orders of magnitude with the temperature and the pressure-dependent viscosity is stronger, does not form a surface stagnant lid, highly viscous 'slabs' submerge to the CMB and effectively influence the hot upwelling plumes. Based on our numerical results it is necessary to implicate the yield stress into the simulations in order to obtain a highly viscous, 'rigid' surface lid, the lithosphere which can be broken up and subduct down to the mantle.

<p>As it is well-known, stress fields are responsible for earthquake formation. In order to analyse stress relations in a study area using focal mechanisms’ (FMS) inversions, it is vital to consider three fundamental criteria:</p><p>(1)       The investigated area is characterized by a homogeneous stress field.</p><p>(2)       The earthquakes occur with variable directions on pre-existing faults.</p><p>(3)       The deviation of the fault slip vector from the shear stress vector is minimal (Wallace-Bott hypothesis).</p><p>The authors have attempted to develop a “fully-automated” algorithm to carry out the classification of the earthquakes as a prerequisite of stress estimations. This algorithm does not call for the setting of hyper-parameters, thus subjectivity can be reduced significantly and the running time can also decrease. Nevertheless, there is an optional hyper-parameter that is eligible to filter outliers, isolated points (earthquakes) in the input dataset.</p><p>In this presentation, they show the operation of this algorithm in case of synthetic datasets consisting of different groups of FMS and a real seismic dataset. The latter come from a survey area in the earthquake-prone Vrancea-zone (Romania). This is a relatively small region (around 30*70 km) in the external part of SE-Carpathians where the distribution of the seismic events is quite dense and heterogeneous.</p><p>It shall be noted that though the initial results are promising, further developments are still necessary. The source codes are soon to be uploaded to a public GitHub repository which will be available for the whole scientific community.</p>