Skip to content

Diagnostics

This page lists dynamo output diagnostics common in the literature, how to calculate them from leeds dynamo code outputs, and references to alternative definitions.

On this page, a quantity with a superscript \(^*\) indicates it has been non-dimensionalised, so \(\chi^*\) is the non-dimensionalised version of \(\chi\).

Magnetic Energy

Dimensional magnetic energy is defined as \(E_\text{mag} = \frac{1}{2\mu_0}\int{\boldsymbol{B^2}}dV\), where \(\mu_0\) is the magnetic permeability. In the dynamo code, energy is non-dimensionalised by \(\rho D \eta^2\), where \(\rho\) is the density, \(D\) is the shell thickness, and \(\eta\) is magnetic diffusivity.

Magnetic energy is calculated in the code as

\[ E_\text{mag}^* = \frac{\text{Pm}}{\text{Ek}}\int{(\boldsymbol{B}^*)^2}dV^*, \]

where Pm is the magnetic Prandtl number and Ek is the Ekman number. Can be found in column 2 of mag_energy.dat.

Magnetic Energy Density

Magnetic Energy divided by the shell volume.

\[ E_\text{mag}^{*\text{density}} = \frac{\text{Pm}}{\text{Ek}V^*}\int{(\boldsymbol{B}^*)^2}dV^*. \]

For Earth-like simulations, \(E_\text{mag}^{*\text{density}}\approx E^*_\text{mag} / 14.59\).

Kinetic Energy

Dimensional kinetic energy defined as \(E_\text{kin} = \frac{1}{2}\rho\int\boldsymbol{u}^2 dV\), where \(V\) is the shell volume, and \(\rho\) is the fluid density. Energy is non-dimensionalised by \(\rho D \eta^2\), where \(D\) is the shell thickness, and \(\eta\) is magnetic diffusivity. Non-dimensional kinetic energy is calculated in the dynamo code as

\[ E_\text{kin}^* = \frac{1}{2}\int(\boldsymbol{u}^*)^2 dV. \]

It can be found in column 2 of vel_energy.dat.

Kinetic Energy Density

Kinetic Energy divided by the shell volume.

\[ E_\text{kin}^{*\text{density}} = \frac{1}{2V^*}\int (\boldsymbol{u}^*)^2 dV^*. \]

For Earth-like geometry, \(E_\text{kin}^{*\text{density}}\approx E^*_\text{kin}/14.59\).

Ohmic Dissipation

Dimensional ohmic dissipation defined as \(D_\Omega = \eta/\mu_0\int(\nabla\times\boldsymbol{B})^2dV\), non dimensionalised with \(\rho\eta^3/D\) where \(\eta\) is magnetic diffusivity, \(\mu_0\) is magnetic permeability, \(\rho\) is density.

Calculated in code as

\[ D_\Omega^* = \frac{2 \text{Pm}}{\text{Ek}}\int\left(\nabla^* \times \boldsymbol{B}^*\right)^2 dV^*, \]

where Pm is the magnetic Prandtl number and Ek is the Ekman number.

Can be found in column 7 of mag_energy.dat.

Viscous Dissipation

Dimensional viscous dissipation defined as \(D_\nu = \rho\nu\int\left(\nabla \times \boldsymbol{u}\right)^2 dV\), non dimensionalised with \(\rho\eta^3/D\), where \(\rho\) is the density, \(\nu\) is the kinematic viscosity, and \(\eta\) is the magnetic diffusivity.

Calculated in code as

\[ D_\nu^* = \text{Pm}\int\left(\nabla^* \times \boldsymbol{u^*}\right)^2 dV^*, \]

where Pm is the magnetic Prandtl number.

Found in column 7 of vel_energy.dat.

Magnetic Reynolds Number

Defined as $$ \text{Rm} = \frac{U D}{\eta}, $$ where \(U\) is a characteristic velocity, \(D\), the characteristic lengthscale, is taken as the shell thickness, and \(\eta\) is the magnetic diffusivity.

In the code, Rm is calculated as

\[ \text{Rm} = \sqrt{\frac{2 E^*_\text{kin}}{V^*}}, \]

where \(E^*_\text{kin}\) is the non-dimensional kinetic energy and \(V^*\) is the non-dimensional shell volume.

For Earth like geometry, can be plotted from gnuplot using vel_energy.dat as

plot "vel_energy.dat" u 1:(sqrt(2*$2/14.59)) w l
or directly from column 2 of dynamo_outputs.dat.

Reynolds number

Defined as

\[ \text{Re} = \frac{U D}{\nu}, \]

where \(U\) is a characteristic velocity, \(D\), the characteristic lengthscale, is taken as the shell thickness, and \(\nu\) is the kinematic viscosity.

In the code, Re calculated as

\[ Re = \frac{1}{\text{Pm}}\sqrt{\frac{2 E^*_\text{kin}}{V^*}}, \]

where Pm is the magnetic Prandtl number, \(E^*_\text{kin}\) is the non-dimensional kinetic energy, and \(V^*\) is the non-dimensional shell volume.

For Earth like geometry, can be plotted from gnuplot using vel_energy.dat as

plot "vel_energy.dat" u 1:((1/Pm)*sqrt(2*$2/14.59)) w l
where Pm should be replaced by its numerical value, or directly from column 3 of dynamo_outputs.dat.

Peclet Number

Defined as

\[ \text{Pe} = \frac{U D}{\kappa} \]

where \(U\) is a characteristic velocity, \(D\), the characteristic lengthscale, is taken as the shell thickness, and \(\kappa\) is the thermal diffusivity.

In the code, Pe calculated as

\[ Pe = \frac{Pr}{\text{Pm}}\sqrt{\frac{2 E^*_\text{kin}}{V^*}}, \]

where Pr is the Prandtl number, Pm is the magnetic Prandtl number, \(E^*_\text{kin}\) is the non-dimensional kinetic energy, and \(V^*\) is the non-dimensional shell volume.

For Earth like geometry, can be plotted from gnuplot using vel_energy.dat as

plot "vel_energy.dat" u 1:((Pr/Pm)*sqrt(2*$2/14.59)) w l
where Pr and Pm should be replaced by their numerical values, or directly from column 4 of dynamo_outputs.dat.

Compositional Peclet Number

Rossby Number

Defined as

\[ \text{Ro} = \frac{U}{\Omega D}, \]

where \(U\) is a characteristic velocity, \(D\), the characteristic lengthscale, is taken as the shell thickness, and \(\Omega\) is the rotation rate.

In the code, Ro calculated as

\[ \text{Ro} = \frac{Ek}{\text{Pm}}\sqrt{\frac{2 E^*_\text{kin}}{V^*}}, \]

where Ek is the Ekman number, Pm is the magnetic Prandtl number, \(E^*_\text{kin}\) is the non-dimensional kinetic energy, and \(V^*\) is the non-dimensional shell volume.

Can be plotted from gnuplot using vel_energy.dat as

plot "vel_energy.dat" u 1:((Ek/Pm)*sqrt(2*$2/V*)) w l
where Ek, Pm, and \(V^*\) should be replaced by their numerical values, or directly from column 6 of dynamo_outputs.dat.

M_Ratio

Defined as

\[ M = \frac{E_\text{mag}}{E_\text{kin}}, \]

where \(E_\text{mag}\) is the dimensional magnetic Energy and \(E_\text{kin}\) is the dimensional kinetic Energy.

Calculated in the code as

\[ M = \frac{E^*_\text{mag}}{E^*_\text{kin}}, \]

where \(E_\text{mag}^*\) and \(E_\text{kin}^*\) are the non-dimensional magnetic and kinetic energies.

Can be plotted from gnuplot using vel_energy.dat and mag_energy.dat as

plot '< paste mag_energy.dat vel_energy.dat' using 1:($2/$8) w l
or from column 7 of dynamo_outputs.dat.

A related qunatitiy is the Squared Alfvén number

Elsasser Number

There are two definitions of the Elsasser number common in the literature. Soderlund et al. (2012) and others define

\[ \Lambda = \frac{\boldsymbol{B}^2}{2\Omega\rho\mu_0\eta}, \]

where \(\boldsymbol{B}\) is the dimesional magnetic field, \(\Omega\) is the rotation rate, \(\rho\) is the fluid density, \(\mu_0\) is the magnetic permeability, and \(\eta\) is the magnetic diffusivity.

Meanwhile, Aubert (2017, 2019) etc. define

\[ \Lambda_\text{A17} = \frac{\boldsymbol{B}^2}{\Omega\rho\mu_0\eta}. \]

\(\Lambda\) as in Soderlund et al. (2012) is calculated in the dynamo code as

\[ \Lambda = \frac{\text{Ek}E^*_\text{mag}}{\text{Pm} V^*}, \]

Where Ek is the Ekman number, \(E^*_\text{mag}\) is the non-dimensional magnetic energy, Pm is the magnetic Prandtl number, and \(V^*\) is the non-dimensional shell volume.

\(\Lambda\) is written out to column 8 of dynamo_outputs.dat. To convert to \(\Lambda_\text{A17}\), multiply by 2.

\(\Lambda\) can also be plotted in gnuplot using mag_energy.dat as

p "mag_energy.dat" using 1:($2*Ek/(V* *Pm)) w l,
where Ek, \(V^*\) and Pm should be replaced by their numerical values.

Lehnert Number

We define the Lehnert number as in Aubert (2017, 2019) etc.

\[ \lambda = \frac{B}{\left(\rho\mu_0\right)^{1/2}\Omega D}, \]

where \(B\) the magnetic field magnitude, \(\rho\) is the fluid density, \(\mu_0\) is the magnetic permeability, \(\Omega\) is the rotation rate, and \(D\) is the shell thickness.

Calculated in the dynamo code as

\[ \lambda = \frac{\text{Ek}}{\text{Pm}}\sqrt{\frac{2 E_\text{mag}^*}{V^*}}, \]

where Ek is the Ekman number, Pm is the magnetic Prandtl number, \(E_\text{mag}^*\) is the non-dimensional magnetic energy, and \(V^*\) is the non-dimensional shell volume.

Written out to column 9 of dynamo_outputs.dat.

Can also be calculated using mag_energy.dat, for example in gnuplot

p "mag_energy.dat" using 1:((Ek/Pm)*sqrt(2*$2/V*)) w l

where Ek, Pm, and \(V^*\) should be replaced by their numerical values.

fdip

Fraction of dipolarity of CMB magnetic field, defined as in Christensen and Aubert (2006)

\[ \mathcal{f}_\text{dip} = \left( \frac {\int{ \boldsymbol{B}_{l=1}(r=r_o) \cdot \boldsymbol{B}_{l=1}(r=r_0) }dS } {\int{ \boldsymbol{B}_{l\leq 12}(r=r_o) \cdot \boldsymbol{B}_{l\leq 12}(r=r_0) }dS} \right)^{1/2}, \]

where \(\boldsymbol{B}_{l}(r)\) is the magnetic field at spherical harmonic degrees \(l\) and radial position \(r_o\), and \(S\) is the spherical surface. Due to this being a ratio of field strengths, no conversions are required due to choice of non-dimensionalisation.

\(\mathcal{f}_\text{dip}\) is written out to column 10 of dynamo_outputs.dat.

Can also be calculated using mag_cmb.dat, for example in gnuplot

p "mag_cmb.dat" u 1:(sqrt($4/$8)),
where column 4 is the surface magnetic energy at \(l=1\) and column 8 is the surface magnetic energy for \(l\leq 12\).

The comparison between the dipole and the \(l\leq 12\) surface magnetic field is made because of comparisons with observational data of the Earth's magnetic field, which is limited in resolution to \(l\approx 13\) due to the crustal magnetic field.

Some literature uses the comparison between the dipole and the whole of the rest of the field

\[ \left( \frac {\int{ \boldsymbol{B}_{l=1}(r=r_o) \cdot \boldsymbol{B}_{l=1}(r=r_0) }dS } {\int{ \boldsymbol{B}(r=r_o) \cdot \boldsymbol{B}(r=r_0) }dS} \right)^{1/2}, \]

which can be calculated in the dynamo code using columns 4 and 2 of mag_cmb.dat.

fohm

Fraction of ohmic dissipation. Defined as

\[ f_\text{ohm} = \frac{D_\Omega}{D_\Omega + D_\nu} \]

where \(D_\Omega\) is the ohmic dissipation and \(D_\nu\) is the viscous dissipation.

Calculated in the code as

\[ f_\text{ohm} = \frac{D^*_\Omega}{D^*_\Omega + D^*_\nu}, \]

and written out to column 11 of dynamo_outputs.dat.

Can also be calculated using vel_energy.dat and mag_energy.dat, for example in gnuplot

p "< paste mag_energy.dat vel_energy.dat" u 1:($6/($6+$13)) w l

Theta_dip

The dipole tilt angle, defined as the angle between the magnetic field dipole and the rotation axis at the north pole. Defined as

\[ \theta = \arccos\left(-\frac{M_z}{|\boldsymbol{M}|} \right) \]

where \(\theta\) is the dipole tilt angle, \(M_z\) is the \(\hat{\boldsymbol{z}}\) component of the magnetic dipole moment, \(|\boldsymbol{M}|\) is the dipole magnitude.

Calculated as $$ \theta = \arccos\left(\frac{g_1^0}{g_1}\right) $$

where \(g_1=\sqrt{\left(g_1^1\right)^2 + \left(h_1^1\right)^2 + \left(g_1^0\right)^2}\) is the non-dimensional dipole magnitude, and \(g_l^m\), \(h_l^m\) are the gauss coefficients of degree \(l\) and order \(m\).

Due to the poloidal-toridal decomposition, the dynamo code takes \(g_1^0 = B^\text{pol}_{1,0}\), \(~g_1^1=-2\Re(B^\text{pol}_{1,1})\), and \(h_1^1=2\Im(B^\text{pol}_{1,1})\), where \(B^\text{pol}_{l,m}\) is the poloidal magnetic scaler of degree \(l\) and order \(m\), and \(\Re\) and \(\Im\) are the real and imaginary coefficients, resepectively.

Written out to column 12 of dynamo_outputs.dat, and column 5 of mag_cmb.dat.

Nusselt

written out to column 13 of dynamo_outputs.dat.

Sherwood

written out to column 14 of dynamo_outputs.dat.

Squared Alfvén