Are geomagnetic data consistent with stably stratified flow at the core–mantle boundary?

Abstract

S U M M A R YRecent first principles calculations of the Earth's outer core thermal and electrical conductivities have raised their values by a factor of three. This has significant implications for geodynamo operation, in particular, forcing the development of a stably stratified layer at the core-mantle boundary (CMB). This study seeks to test the hypothesis of a stably stratified layer in the uppermost core by analysing geomagnetic observations made by the CHAMP satellite. An inversion method is utilized that jointly solves for the time-dependent main field and the core surface flow, where we assume the temporal variability of the main field, its secular variation (SV), to be entirely due to advective motion within the liquid outer core. The results show that a large-scale pure toroidal flow, consistent with a stably stratified layer atop the outer core, is not compatible with the observed magnetic field during the CHAMP era. However, allowing just a small amount of poloidal flow leads to a model fitting the observations satisfactorily. As this poloidal flow component is large scale, within a predominantly toroidal, essentially tangentially geostrophic flow, it is compatible with a stably stratified upper outer core. Further, our assumption of little or no diffusive SV may not hold, and a small amount of SV generated locally by diffusion might lead to a large-scale pure toroidal flow providing an acceptable fit to the data.Key words: Magnetic field; Rapid time variations; Satellite magnetics. I N T RO D U C T I O NMagnetic data are one of the rare sources of information about the motion of the iron-rich liquid forming the outer core of the Earth. It is generally accepted that the Earth's magnetic field is generated mainly in the outer core by convective flows associated with heat and light element release due to the slow freezing of the Earth's inner core. A better understanding of the evolution of the Earth's deep interior is therefore closely linked to our ability to describe this outer core flow. The magnetic field and core flow are related through the induction equation, derived directly from Maxwell's equations.Most approaches begin by deriving magnetic field models from measurements at the Earth's surface and satellite altitude. These models are then used to deduce the flow in the core, even if the modelling step somehow filters out part of the information available in the data. Several difficulties limit further our ability to describe the core flow.First, the magnetic field is measured at the Earth's surface and, after neglecting mantle electrical conductivity to allow the field to be downward continued to the core-mantle boundary (CMB), only the radial component of the poloidal part of the magnetic field is continuous across the conductivity jump there. This component can reasonably be estimated at the top of the free stream, just under the liquid viscous boundary layer (Roberts & Scott 1965;Jault & Le Mouël 1991). However, it is available only for the longest wavelengths, because the field generated...

Decadal variability in core surface flows deduced from geomagnetic observatory monthly means

“…• E, westward flow beneath the Atlantic Ocean, then turning south, is seen in models derived from satellite data (e.g Holme and Silva and Hulot 2012;Lesur et al 2015;Baerenzung et al 2016). The strong southward flow beneath the western Indian Ocean visible in Fig.…”

“…Similarly, the small clockwise eddy beneath the south-western Pacific Ocean is not generally a feature of models derived from satellite data -in fact, the tangentially geostrophic flow of Holme and Olsen (2006) has a weak clockwise eddy, but their toroidal flow has an anti-clockwise eddy in the same location (and, as noted above, features of this size are not well resolved by observatory data). Besides generally decreasing power spectra for both toroidal and poloidal flow components, and an approximately order-of-magnitude difference between the toroidal and poloidal power, the dominant feature of the spectra is the loss of power at toroidal degree 3 (e.g Holme and Olsen 2006;Lesur et al 2010;Lesur et al 2015;Baerenzung et al 2014Baerenzung et al , 2016, regardless of whether or not the flow is assumed tangentially geostrophic. All these features are present in our spectra, for both snapshot and moderately TO-like flows (Fig.…”

Constraints from material properties on the dynamics and evolution of Earth’s core

“…Geomagnetic data are presently unable to unambiguously identify a stable layer 98,99 , although a recent constraint on core electrical conductivity from long-term dipole field variations is consistent with the high conductivity estimates that argue in favour of stratification 100 . In isolation, both a stable layer and lateral heat flow variations can explain prominent features of the present geomagnetic field: wave motions in a ∼100 km-thick stable layer can account for short-period fluctuations in the dipole field 78 ∆ρ (gm/cc) 0.24 [17] 0.6 [40] 0.8 [33] 1.0 [33] c S O -0.0002 [14] 0.0004 [14] 0.0006 [81] c S Si -0.0554 [14] 0.0430 [14] 0.0096 [81] c L O -0.0256 [14] 0.0428 [14] 0.0559 [81] c L Si -0.0560 [14] 0.0461 [14] 0.0115 [81] Cp (J/kg/K) 715 [56] -800 [53] --- [25] , 1.4 [23] , 1.86 [26, * ] 1.12 [25] 1.11 [25] 1.18 [25] k (W/m/K) 159 [25] , 150 [23] , 170 [26] 107 [25] 99 [25] 101 [25] D O (×10 −8 m 2 /s) [ Table 1.…”

Magnetic observatories: biases over CHAMP satellite mission

“…2 describes the theoretical framework of this study and the considered approach to analyze different signals contained in the monthly observatory biases over the 9 years of the CHAMP (CHAllenging Minisatellite Payload) satellite mission (June 2000-August 2009). These monthly observatory biases are calculated by comparing the observatory data to the core field predictions given by the G model (Lesur et al, 2015). In Sect.…”

“…Taking advantage of 9 years of the CHAMP (CHAllenging Minisatellite Payload) satellite mission (June 2000-August 2009), we investigate the temporal evolution of the observatory monthly magnetic biases. To determine these biases we compute X (northward), Y (eastward) and Z (vertically downward) monthly means from 42 observatory 1 min values or hourly values, and compare them to synthetic monthly means obtained from a G field model (Lesur et al, 2015). Afterwards, the average of biases at all observatories over 9 years is calculated and analyzed.…”