Skip to content

Bondi–Sachs Coordinates and Expansion

Scrollable formula. Use Left and Right arrow keys to reveal hidden content.

Scrollable table. Use Left and Right arrow keys to reveal additional columns.

Scrollable figure. Use Left and Right arrow keys to inspect hidden content.

Bondi–Sachs coordinates turn the vacuum Einstein equation near future null infinity into a radial hierarchy. The angular metric contains two freely varying radiative components; the remaining leading coefficients are obtained from hypersurface equations or propagated by asymptotic constraints. This division is more informative than a list of powers of 1/r1/r: it says what must be supplied, what the field equations determine, and which coordinate conditions make those statements true. We assume vacuum outside a compact source and a smooth power expansion in an asymptotic Bondi frame; logarithmic asymptotics and detector-specific reconstruction are not treated here.

Helpful background. Einstein Equations supplies the vacuum equations placed into characteristic form; Waves, Green Functions, and Geometric Optics supplies outgoing null characteristics; Causality and Global Structure supplies future null infinity; Gravitational Radiation and Waveforms supplies the physical interpretation of radiative coefficients; and Formulations of General Relativity supplies the distinction among gauge, constraints, and freely specified characteristic data.

Use coordinates (u,r,xA)(u,r,x^A) adapted to outgoing null hypersurfaces. The level sets u=constantu=\text{constant} are null, the angles xAx^A remain constant along their generators, and rr is an areal radius: a surface of constant (u,r)(u,r) has area 4πr24\pi r^2. These choices imply

guu=0,guA=0,det⁡hAB=det⁡qAB,g^{uu}=0, \qquad g^{uA}=0, \qquad \det h_{AB}=\det q_{AB},

where qABq_{AB} is the unit-sphere metric and gAB=r2hABg_{AB}=r^2h_{AB}. The most general metric in this gauge is

ds2=−(Vre2β−r2hABUAUB)du2−2e2β du dr−2r2hABUB du dxA+r2hABdxAdxB.\begin{aligned} ds^2={}&-\left(\frac Vr e^{2\beta}-r^2h_{AB}U^AU^B\right)du^2 -2e^{2\beta}\,du\,dr \\ &-2r^2h_{AB}U^B\,du\,dx^A+r^2h_{AB}dx^A dx^B. \end{aligned}

The determinant condition removes the scalar area degree of freedom from hABh_{AB}, leaving precisely two independent angular components. It is a coordinate condition, not an extra field equation. The areal radius can fail where the outgoing null congruence has zero expansion, so this chart is not promised to cover the entire spacetime.

An inertial Bondi frame approaches outgoing Minkowski coordinates as r→∞r\to\infty:

β→0,UA→0,Vr→1,hAB→qAB.\beta\to0, \qquad U^A\to0, \qquad \frac Vr\to1, \qquad h_{AB}\to q_{AB}.

Expand the conformal angular metric as

hAB=qAB+CABr+EABr2+O(r−3).h_{AB} =q_{AB}+\frac{C_{AB}}r+\frac{E_{AB}}{r^2}+O(r^{-3}).

Taking the determinant order by order gives

qABCAB=0,qABEAB=12CABCAB.q^{AB}C_{AB}=0, \qquad q^{AB}E_{AB}=\frac12 C_{AB}C^{AB}.

The second relation follows without using dynamics. For a small matrix perturbation, det⁡(I+ϵA+ϵ2B)\det(I+\epsilon A+\epsilon^2B) has second-order coefficient

tr⁡B+12[(tr⁡A)2−tr⁡(A2)].\operatorname{tr}B +\frac12\left[(\operatorname{tr}A)^2-\operatorname{tr}(A^2)\right].

Apply this formula to q−1hq^{-1}h with ϵ=1/r\epsilon=1/r. The first-order coefficient sets tr⁡C=0\operatorname{tr}C=0; the second then sets tr⁡E=12tr⁡(C2)\operatorname{tr}E=\tfrac12\operatorname{tr}(C^2). This is an independent algebraic check that no sign from the spacetime metric enters the two-dimensional determinant condition.

Thus CABC_{AB} is trace-free and has two components. The determinant alone does not eliminate the trace-free part of EABE_{AB}. The additional vacuum no-logarithm condition makes that part divergence-free; smoothness on the complete two-sphere then forces it to vanish. Consequently EAB=14qABCCDCCDE_{AB}=\tfrac14q_{AB}C_{CD}C^{CD}. Allowing logarithms or angular singularities changes this conclusion; their absence is part of the present regime Mädler and Winicour 2018, §3, eqs. (48)–(54), pp. 13–14, PDF.

Substitute the series into the radial Einstein equations in their hierarchical order. The rrrr equation first determines β\beta; the rArA equation then determines UAU^A; the angular trace determines VV. At the first consequential orders,

β=−CABCAB32r2+O(r−3),\beta=-\frac{C_{AB}C^{AB}}{32r^2}+O(r^{-3}), UA=−DBCAB2r2+O(r−3),V=r−2m+O(r−1).U^A=-\frac{D_BC^{AB}}{2r^2}+O(r^{-3}), \qquad V=r-2m+O(r^{-1}).

Here DAD_A is the covariant derivative of qABq_{AB} and m(u,xA)m(u,x^A) is the Bondi mass aspect, with dimensions of length because GG remains explicit. This chain, including the determinant restrictions and the next-order angular-momentum coefficient, is derived in Mädler and Winicour 2018, §3, pp. 10–16, PDF.

Notice the different radial orders in physical components: gAB=r2qAB+rCAB+O(1)g_{AB}=r^2q_{AB}+rC_{AB}+O(1), while guAg_{uA} has a finite leading angular-gradient term and guu=−1+2m/r+O(r−2)g_{uu}=-1+2m/r+O(r^{-2}). Radiation, longitudinal frame structure, and Coulombic mass therefore occupy distinct coefficients even though they belong to one metric.

Free radiative data and constrained coefficients

Section titled “Free radiative data and constrained coefficients”

The hierarchy separates three roles:

CoefficientHow it entersStatus in the formal asymptotic solution
CAB(u,xA)C_{AB}(u,x^A)1/r1/r coefficient of hABh_{AB}Freely prescribed radiative data over the retarded-time interval
m(u,xA)m(u,x^A)O(1)O(1) correction in V=r−2m+⋯V=r-2m+\cdotsInitial value supplied; retarded-time evolution fixed by a supplementary constraint
LA(u,xA)L_A(u,x^A)Next independent coefficient in UAU^AInitial value supplied; evolution fixed by the angular supplementary constraint
β\beta, leading UAU^A, and higher radial coefficientsRadial metric functionsDetermined sequentially on each outgoing null hypersurface

Calling CABC_{AB} “free” does not mean that arbitrary data automatically produce a complete global solution. In this formal expansion, CABC_{AB} is supplied for all uu in the interval; it is not predicted from one initial cut. A well-posed characteristic evolution also needs suitable data on an initial null hypersurface or worldtube and compatibility at their intersection. That continuum formulation belongs to the characteristic-evolution treatment; the present result identifies the asymptotic coefficients.

The news convention used throughout this volume is NAB=∂uCABN_{AB}=\partial_uC_{AB}. Mädler and Winicour define NABMW=12∂uCABN^{\rm MW}_{AB}=\tfrac12\partial_uC_{AB}, so their constraint formulas must be translated using NABMW=NAB/2N^{\rm MW}_{AB}=N_{AB}/2 before comparison.

Minkowski space is recovered with

CAB=0,m=0,LA=0,C_{AB}=0, \qquad m=0, \qquad L_A=0,

which sets β=UA=0\beta=U^A=0 and V=rV=r. Outgoing Eddington–Finkelstein Schwarzschild is recovered with CAB=0C_{AB}=0, UA=β=0U^A=\beta=0, and m=GMm=GM, giving V=r−2GMV=r-2GM. These limits check the sign and dimension of the mass term. They also show that a nonzero mass aspect is Coulombic data, not by itself radiation: both stationary metrics have vanishing shear and news.

Assume EAB=a qABCCDCCDE_{AB}=a\,q_{AB}C_{CD}C^{CD} after its trace-free part has been excluded. Use the determinant condition to determine aa.

Solution

Contract the ansatz with qABq^{AB}. Because the unit sphere is two-dimensional, qABqAB=2q^{AB}q_{AB}=2, so qABEAB=2a CCDCCDq^{AB}E_{AB}=2a\,C_{CD}C^{CD}. The determinant expansion requires this trace to equal 12CCDCCD\tfrac12C_{CD}C^{CD}. For nonzero shear the coefficients therefore obey 2a=1/22a=1/2, hence a=1/4a=1/4; continuity gives the same result when the shear vanishes.

  • Thomas Mädler and Jeffrey Winicour, “Bondi–Sachs Formalism,” Scholarpedia 11 (2016) 33528, updated arXiv version (2018), arXiv:1609.01731v3, Open PDF.