Orthonormal frame in spacetime

An orthonormal frame $\{e_a\}$ satisfies $g(e_a,e_b)={\eta }_{ab}$. Any two such frames differ by a local Lorentz transformation.

Along a curve with tangent $T$, a frame is parallel propagated when ${\nabla }_T{e}_a=0$. Metric compatibility makes every inner product $g(e_a,e_b)$ constant along the curve, so an initially orthonormal frame remains orthonormal.

The dual coframe obeys ${\theta }^a(e_b)={\delta }_b^a$ and reconstructs the metric as $g={\eta }_{ab}{\theta }^a\otimes {\theta }^b$.

An orthonormal coframe converts fixed flat gamma matrices to coordinate-index Clifford multiplication. Its metric identity gives the curved anticommutation relation. The spin connection is chosen so that these matrices are covariantly constant under the combined spin and tensor connection, which is the tetrad postulate in Clifford form.

A zweibein is a two-dimensional orthonormal coframe in spacetime: its components obey $h_{\mu \nu }={\eta }_{AB}{e}^A{}_{\mu }{e}^B{}_{\nu }$. It converts worldsheet coordinate indices into local Lorentz-frame indices and defines curved gamma matrices ${\gamma }^{\mu }=e_A{}^{\mu }{\gamma }^A$. The associated spin connection supplies the covariant derivative of worldsheet spinors. In flat conformal gauge the coframe can be constant.

Connection one-forms are defined by $\nabla e_b={\omega }^a{}_b\otimes e_a$. For the Levi-Civita connection they obey Cartan's torsion-free equation $d{\theta }^a+{\omega }^a{}_b\wedge {\theta }^b=0$ and metric compatibility ${\omega }_{ab}=-{\omega }_{ba}$.

In an orthonormal coframe in spacetime, antisymmetry applies after lowering the first index with the Lorentzian frame metric. Thus mixed time–space connection 1-forms are equal under index interchange, while mixed spatial forms are opposite. The same rule holds for curvature 2-forms. Confusing lowered antisymmetry with ordinary mixed-index antisymmetry changes the curvature signs.

In an orthonormal frame, the spin connection is the Lorentz connection acting on spinors. With ${\gamma }^{AB}=\frac{1}{2}[{\gamma }^A,{\gamma }^B]$, its covariant derivative is ${\nabla }_{\mu }\psi ={\partial }_{\mu }\psi +\frac{1}{4}{\omega }_{\mu AB}{\gamma }^{AB}\psi$. The metric connection determines $\omega$ when torsion conditions are specified. A zweibein gives its two-dimensional form; omitting it on a curved worldsheet generally destroys covariance.

For $g_{ab}={\Omega }^2{\eta }_{ab}$, use the vierbein $e^{\mu }{}_a=\Omega {\delta }^{\mu }{}_a$ and constant flat gamma matrices. The tetrad postulate and ${\Gamma }^c{}_{ab}={\delta }_a^c{\partial }_b\log \Omega +{\delta }_b^c{\partial }_a\log \Omega -{\eta }_{ab}{\eta }^{cd}{\partial }_d\log \Omega$ give the displayed lowered spin connection. With $L={\hat{\gamma }}^c{\partial }_c\log \Omega$, its spinor action is ${\nabla }_a={\partial }_a+[{\hat{\gamma }}_a,L]/4$. The Lorentz index lowering matters in indefinite signature.

The tetrad postulate expresses compatibility of the spacetime and Lorentz-frame connections on a vierbein. For the inverse coframe $e_{\nu }{}^b$, solving gives ${\omega }_a{}^{\mu }{}_{\nu }=e_{\nu }{}^b({\Gamma }^c{}_{ab}{e}^{\mu }{}_c-{\partial }_a{e}^{\mu }{}_b)$. Metric compatibility makes the lowered connection antisymmetric in its Lorentz indices. Lifting this connection through the Spinor representation of the Lorentz group defines the spinor covariant derivative.

Lift the Levi-Civita connection to the spinor bundle using the spin representation of the Lorentz algebra. Under a local Lorentz transformation, its matrix ${\Omega }_{\mu }=\frac{1}{4}{\omega }_{\mu AB}{\gamma }^{AB}$ obeys ${\Omega }_{\mu }^{\prime }=S{\Omega }_{\mu }{S}^{-1}-({\partial }_{\mu }S)S^{-1}$, so ${\nabla }_{\mu }^{\prime }{\psi }^{\prime }=S{\nabla }_{\mu }\psi$. The Clifford algebra identity $[{\gamma }^C,{\gamma }^{AB}]=2({\eta }^{CA}{\gamma }^B-{\eta }^{CB}{\gamma }^A)$ fixes the coefficient. The tetrad postulate then gives ${\partial }_{\mu }{\gamma }^{\nu }+{\Gamma }_{\mu \rho }^{\nu }{\gamma }^{\rho }+[{\Omega }_{\mu },{\gamma }^{\nu }]=0$, establishing compatibility with curved Clifford multiplication.

A twistor spinor has vanishing gamma-trace-free part of its spinor covariant derivative. In $n$ dimensions the displayed equation uses gamma matrices satisfying $\{{\gamma }_a,{\gamma }_b\}=2g_{ab}$. It is weaker than being a parallel spinor. In four dimensions it implies $[{\gamma }_a{\nabla }_b+{\gamma }_b{\nabla }_a-g_{ab}{\gamma }^c{\nabla }_c/2]ϵ=0$.

In the conformal frame for $g={\Omega }^2\eta$, identify the spinor fibres using the chosen vierbein. If ${ϵ}_0$ is a flat constant spinor field, the spin connection of a conformally flat metric gives ${\nabla }_aϵ={\hat{\gamma }}_a{\hat{\gamma }}^c({\partial }_c\log \Omega )ϵ/2$. Contracting with curved ${\gamma }^a={\Omega }^{-1}{\hat{\gamma }}^a$ yields ${\gamma }^a{\nabla }_aϵ=n{\Omega }^{-1}{\hat{\gamma }}^c({\partial }_c\log \Omega )ϵ/2$. Thus ${\nabla }_aϵ={\gamma }_a{\gamma }^b{\nabla }_bϵ/n$: the rescaled parallel spinor is a twistor spinor.

The complexified tangent space of a four-dimensional Lorentzian manifold factors locally into an unprimed and a primed two-dimensional spinor representation. Their alternating forms ${ϵ}_{AB}$ and ${ϵ}_{A^{\prime }{B}^{\prime }}$ encode the metric tensor and raise or lower spinor indices. A convention must specify the order of contraction with these alternating forms. An antisymmetric spacetime tensor decomposes into symmetric unprimed and primed spinors, $F_{A{A}^{\prime }B{B}^{\prime }}={\varphi }_{AB}{ϵ}_{A^{\prime }{B}^{\prime }}+{\tilde{\varphi }}_{A^{\prime }{B}^{\prime }}{ϵ}_{AB}$. This splits the six-dimensional space of differential two-forms into the two three-dimensional eigenspaces of the complexified Hodge star operator.

A two-component two-component spinor is an element of a two-dimensional complex spin space with invariant antisymmetric form ${ϵ}_{AB}$. In complexified four-dimensional geometry there are independent unprimed and primed spin spaces, and vectors are represented by $x^{A{A}^{\prime }}$. Raising an index uses the antisymmetric form, so ${\pi }^{A^{\prime }}{\pi }_{A^{\prime }}=0$. This identity makes the incidence-plane directions ${\pi }^{A^{\prime }}{\partial }_{A{A}^{\prime }}$ totally null.

The totally symmetric unprimed four-index spinor represents one chiral part of the Weyl tensor. The other part is represented by a primed spinor, conjugate to the unprimed one on a Lorentzian real slice. The trace-free symmetry separates conformal curvature from the Ricci tensor and scalar-curvature terms.

The spinor covariant derivative commutator is a differential two-form in its derivative indices, so the two-component spinor calculus decomposition defines symmetric curvature operators ${\Delta }_{AB}$ and ${\Delta }_{A^{\prime }{B}^{\prime }}$. With ${ϵ}^{AB}{ϵ}_{AB}=2$, the unprimed projection is ${\Delta }_{AB}={ϵ}^{A^{\prime }{B}^{\prime }}[{\nabla }_{A{A}^{\prime }},{\nabla }_{B{B}^{\prime }}]/2$, and the primed projection is analogous. For a torsion-free connection these are algebraic curvature derivations: they annihilate scalar functions and obey the Leibniz rule on tensor products.

Metric compatibility makes the curvature action on a two-component spinor symplectic, so its lowered curvature spinor is symmetric in each pair. The pair-exchange symmetry of the Riemann curvature tensor makes $X_{ABCD}=X_{CDAB}$. This six-component tensor decomposes into the five-component totally symmetric Weyl curvature spinor ${\Psi }_{ABCD}=X_{(ABCD)}$ and one scalar term. In the displayed normalization, with alternating forms satisfying ${ϵ}^{AB}{ϵ}_{AB}=2$, its scalar is $\Lambda =X_{ABCD}{ϵ}^{AC}{ϵ}^{BD}/3$. This coefficient follows by contracting the displayed scalar term; different normalizations of that term give different scalar conventions.

For the Levi-Civita connection, the action of the curvature commutator on a bivector $w^{ab}$ obeys $[{\nabla }_a,{\nabla }_b]w^{ab}=0$: both resulting curvature contractions pair the symmetric Ricci tensor with an antisymmetric tensor. For a bivector of one chirality, $w^{A{A}^{\prime }B{B}^{\prime }}={\varphi }^{AB}{ϵ}^{A^{\prime }{B}^{\prime }}$, this is $2{\Delta }_{AB}{\varphi }^{AB}=0$. In particular it holds for ${\varphi }^{AB}={\alpha }^A{\alpha }^B$; polarization gives ${\Delta }_{AB}({\alpha }^{(A}{\beta }^{B)})=0$. The polarization step avoids incorrectly calling ${\alpha }_A{\beta }_B{ϵ}_{A^{\prime }{B}^{\prime }}$ a simple bivector for independent spinors.

A parallel spinor field is annihilated by the spinor covariant derivative. Its value at one point determines it on a connected domain by parallel transport. The spinor curvature identity implies $R_{\mu \nu AB}{\gamma }^{AB}ϵ=0$, but this local condition alone need not settle global existence: the spin structure and holonomy also matter. Lorentzian parallel spinors need not force the whole Ricci tensor to vanish, unlike the positive-definite case.

In flat three-dimensional spacetime with polar spatial orthonormal coframe $(dT,d\rho ,\rho d\varphi )$, the spin connection has ${\omega }_{12}=-d\varphi$. Thus every parallel spinor has the displayed components. Since $({\gamma }^1{\gamma }^2{)}^2=-1$, these components change sign under $\varphi \mapsto \varphi +2\pi$. This is the lift of a full rotation to the spin structure extending over the disk; the polar frame itself does not extend over the origin. In a Cartesian frame the same parallel spinors have constant components.

The spinor covariant derivative has curvature $d\Omega +\Omega \wedge \Omega =\frac{1}{4}{R}_{AB}{\gamma }^{AB}$, since the spin representation preserves the Lorentz algebra bracket. In a coordinate basis with the torsion-free connection, antisymmetric second covariant derivatives therefore give the displayed commutator. With unit-weight antisymmetrization, ${\nabla }_{[\mu }{\nabla }_{\nu ]}\psi =\frac{1}{8}{R}_{\mu \nu AB}{\gamma }^{AB}\psi =\frac{1}{16}{R}_{\mu \nu AB}[{\gamma }^A,{\gamma }^B]\psi$. A noncoordinate frame requires its frame-commutator correction when using the operator commutator; the full second covariant derivative includes that correction automatically.

For $\{{\gamma }_a,{\gamma }_b\}=2s{g}_{ab}$, the compatible spinor covariant derivative has curvature $R_{abcd}{\gamma }^{cd}/(4s)$. The torsion-free connection preserves Clifford multiplication, so the mixed derivative terms cancel for $D_a={\nabla }_a+c{\gamma }_a$. Its integrability is therefore $(R_{abcd}{\gamma }^{cd}+8s{c}^2{\gamma }_{ab})ϵ=0$. Changing antisymmetrization weight cannot change the relative coefficient. The common normalization uses $s=1$; $s=2$ gives the coefficient $16c^2$.

Four independent complex Dirac solutions in four dimensions span each spinor fibre, because a connection-parallel solution is determined by its value at one point. Their integrability operator must therefore vanish on the whole fibre. Independence of the six Clifford bivectors then gives the displayed constant-curvature tensor and $R_{ab}=-12s{c}^2{g}_{ab}$. When $c>0$, the Ricci tensor is automatically nondegenerate.

For the torsion-free spin connection with $[{\nabla }_{\nu },{\nabla }_{\rho }]=R_{\nu \rho ab}{\gamma }^{ab}/4$, Clifford expansion and the first Bianchi identity give ${\gamma }^{\mu \nu \rho }{R}_{\nu \rho ab}{\gamma }^{ab}=4R^{\mu }{}_{\lambda }{\gamma }^{\lambda }-2R{\gamma }^{\mu }$. Thus the antisymmetric double derivative equals $G^{\mu }{}_{\lambda }{\gamma }^{\lambda }/2$. This is the shared algebra behind quadratic-order minimal four-dimensional supergravity invariance and the Nester two-form divergence.

Cartan's first structure equation isFor the torsion-free Levi-Civita connection, $T^a=0$.

Cartan's second structure equation defines the curvature forms by

The curvature two-forms are

In the exterior orthonormal coframe $E^0=\sqrt{V}dt$, $E^1=dr/\sqrt{V}$, $E^2=rd\theta$, $E^3=r\sin \theta d\varphi$, use ${\Omega }^a{}_b=\frac{1}{2}{R}^a{}_{bcd}{E}^c\wedge E^d$. The independent mixed-index curvature 2-forms are ${\Omega }^0{}_1=-V^{\prime \prime }{E}^0\wedge E^1/2$, ${\Omega }^0{}_A=-V^{\prime }{E}^0\wedge E^A/(2r)$, ${\Omega }^1{}_A=-V^{\prime }{E}^1\wedge E^A/(2r)$ for $A=2,3$, and ${\Omega }^2{}_3=(1-V)E^2\wedge E^3/r^2$. The orthonormal Ricci tensor is diagonal with entries $V^{\prime \prime }/2+V^{\prime }/r$, $-V^{\prime \prime }/2-V^{\prime }/r$, and twice $(1-V-r{V}^{\prime })/r^2$. The Einstein field equations with negative cosmological constant then give $V=1-2M/r+r^2/l^2$.

// Target: general-relativity.bigb

The curvature operator is diagonal in an orthonormal coframe $\{e^a\}$ when every curvature 2-form satisfies ${\mathcal{R}}^{ab}=K_{ab}{e}^a\wedge e^b$. Its matrix on the corresponding basis of 2-forms is then diagonal.

If ${\mathcal{R}}^{ab}$ is proportional to $e^a\wedge e^b$ for every pair, then a curvature component can be nonzero only when its two index pairs describe the same coordinate two-plane. The contraction $R_{bd}=R^a{}_{bad}$ therefore vanishes for $b\ne d$, so the Ricci tensor is diagonal in the same frame.