---
title: Affine-Null Characteristic Formulation
url: https://www.emergentmind.com/topics/affine-null-characteristic-formulation
type: topic
---

# Affine-Null Characteristic Formulation

Affine-null characteristic formulation is a characteristic formulation of Einstein’s equations in which null hypersurfaces are labeled by a null coordinate and the generators of those hypersurfaces are parametrized by an affine parameter rather than by the Bondi-Sachs areal radius. In the general-relativistic literature, this choice appears in several closely related settings: outgoing or ingoing single-null evolution for gravitational radiation and black-hole spacetimes [1303.6969], characteristic initial-boundary value problems with two intersecting null hypersurfaces [1810.04743], spherically symmetric collapse and black-hole interior evolution [1910.03439], hyperbolicity analyses of Bondi-like systems including affine gauge [2404.16720], and conformal or compactified formulations reaching null infinity [2504.18973]. The central geometric point is that the affine parameter remains regular where the null expansion vanishes, so the areal radius becomes a metric function rather than a coordinate; the central analytic point is that the Einstein equations typically lose the naive Bondi-Sachs hierarchy and must be reorganized through auxiliary variables to recover a sequential radial integration scheme [1303.6969].

## 1. Geometric definition and coordinate gauges

In the outgoing single-null affine-null gauge, the metric form used in characteristic relativity is
\[
ds^2 = V\,du^2 + 2\,du\,dr - h_{ab}\left(d\theta^a+W^a du\right)\left(d\theta^b+W^b du\right),
\]
with coordinate conditions
\[
g_{rr}=0,\qquad g_{ra}=0,\qquad g_{ru}=1.
\]
This makes \(u=\mathrm{const}\) a null hypersurface and \(r\) an affine parameter along the outgoing null geodesics generated by \(\partial_r\) [2104.09972]. In the metric formulation designed for characteristic numerical evolution, the affine-null coordinates are instead written as \(x^\alpha=(u,\lambda,x^A)\), with metric
\[
ds^2 = -(V - g_{AB} W^A W^B)\,du^2 - 2\,du\,d\lambda - 2 g_{AB} W^B\,du\,dx^A + g_{AB}\,dx^A dx^B,
\]
and angular decomposition
\[
g_{AB} = r^2 h_{AB}, \qquad \det(h_{AB}) = \det(q_{AB}),
\]
so that \(r\) is no longer a coordinate but a metric function \(r(u,\lambda,x^A)\) [1303.6969].

A closely related double-affine characteristic gauge is used for two intersecting null hypersurfaces. With coordinates
\[
x^a=(w,\lambda,x^A),
\]
the affine-null metric ansatz is
\[
ds^2 = -W\,dw^2 + 2\epsilon\,dw\,d\lambda + r^2 h_{AB}\bigl(dx^A-W^A dw\bigr)\bigl(dx^B-W^B dw\bigr),
\]
with
\[
W|_{\lambda=0}=0,\qquad W_{,\lambda}|_{\lambda=0}=0,\qquad W^A|_{\lambda=0}=0,
\]
and
\[
h^{AC}h_{CB}=\delta^A{}_B,\qquad \det(h_{AB})=h(x^A)
\]
[1810.04743]. In this setting, \(w\) labels the family of null hypersurfaces \(N_w\), \(\lambda\) is an affine parameter along the null generators of \(N_w\), and the null boundary \({B}\) sits at \(\lambda=0\) [1810.04743].

In spherical symmetry, the affine-null metric takes the particularly simple form
\[
g_{ab}dx^a dx^b = -\mathcal{V}\,du^2 - 2\,du\,d\lambda + r^2(d\theta^2+\sin^2\theta\,d\phi^2)
\]
[1910.03439], or equivalently
\[
ds^2 = -V(u,\lambda)\,du^2 - 2\,du\,d\lambda + r^2(u,\lambda)\,q_{AB}(x^C)\,dx^A dx^B
\]
[2405.19122]. In \(n\)-dimensional spherical symmetry, the corresponding Einstein-Maxwell-scalar formulation uses
\[
ds^2=-V(w,\lambda)\,dw^2+2\epsilon\,dw\,d\lambda+r^2(w,\lambda)q_{AB}dx^A dx^B,
\]
with \(\epsilon^2=1\) [2606.13636].

The geometrical motivation is stated consistently across these formulations: \(r\) is tied to surface area and can fail at vanishing expansion, whereas the affine parameter is tied to the geodesic parametrization of the null generators and remains regular until caustics form [1303.6969]. In cosmological past-null-cone applications, the diameter distance reaches a maximum at the apparent horizon and then decreases, so the same value of \(r\) can correspond to multiple points on the cone, while the affine parameter remains a unique label along each null geodesic all the way through the refocussed region [1111.6025].

## 2. Relation to Bondi-Sachs and the loss of the naive hierarchy

The Bondi-Sachs formulation uses outgoing null hypersurfaces \(u=\text{const}\) and areal radius \(r\), with metric
\[
ds^2 = -\left( \frac{V}{r} e^{2\beta} - r^2 h_{AB} U^A U^B \right) du^2 - 2 e^{2\beta} \, du\,dr - 2 r^2 h_{AB} U^B \, du\,dx^A + r^2 h_{AB} dx^A dx^B,
\]
and determinant condition
\[
\det(h_{AB}) = \det(q_{AB}) = q(x^C)
\]
[1303.6969]. Its main equations take the hierarchical radial form
\[
\beta_{,r} = N_\beta[h_{CD}],
\]
\[
(r^4 e^{-2\beta} U_{,r})_{,r} = N_U[h_{CD},\beta],
\]
\[
V_{,r} = N_V[h_{CD},\beta,U],
\]
\[
M^A M^B (r h_{AB,u})_{,r} = N_h[h_{CD},\beta,U,V],
\]
so one can integrate radially in the sequence
\[
[h_{AB}] \rightarrow \beta \rightarrow U \rightarrow V \rightarrow h_{AB,u}
\]
[1303.6969].

The affine-null system is related to Bondi-Sachs by
\[
\partial_r \lambda(u,r,x^A) = e^{2\beta},
\]
but the paper emphasizes that this is misleading at the PDE level because the hypersurfaces \(r=\) const and \(\lambda=\) const are not the same [1303.6969]. In affine-null coordinates, the corresponding equations initially look like
\[
r^{-1} r_{,\lambda\lambda} = H_r[h_{CD}],
\]
\[
(r h_{AB} W^B)_{,\lambda\lambda} = H_W[h_{CD},r],
\]
\[
2(r^2)_{,u\lambda} = H_V[h_{CD},r,W],
\]
\[
M^A M^B (r h_{AB})_{,u\lambda} = H_h[h_{CD},r,W,V].
\]
The key difficulty is the third equation: it contains \(r_{,u\lambda}\) on the left, so the system is not a pure radial hypersurface hierarchy in the same clean sense as Bondi-Sachs [1303.6969].

This same obstruction reappears in later spherical and conformal formulations. In the affine-null spherical Einstein-scalar system, the raw field equations are
\[
r_{,\lambda\lambda} = -\frac{r}{2}\,\Phi_{,\lambda}\Phi_{,\lambda},
\]
\[
(\mathcal V r r_{,\lambda})_{,\lambda} = 1 + 2(r^2)_{,\lambda u},
\]
\[
(r^2\Phi_{,u})_{,\lambda} + (r^2\Phi_{,\lambda})_{,u} - (r^2\mathcal V \Phi_{,\lambda})_{,\lambda} = 0
\]
[1910.03439]. In the conformal affine-null system, the specialized PDEs for \(\mathcal R,\mathcal W,\Phi\) still contain a derivative \(\mathcal R_{,u}\), which destroys the straightforward Bondi-Sachs-like hierarchy [2504.18973].

A common misconception is that replacing \(r\) by an affine parameter is merely a coordinate relabeling. The cited formulations argue otherwise: the change is geometrically natural, but analytically it alters the structure of the reduced Einstein system and requires a different hierarchy [1303.6969].

## 3. Restoration of the characteristic hierarchy by auxiliary variables

The standard method for restoring a sequential radial integration scheme is the introduction of auxiliary variables. In the general metric formulation, the auxiliary variables are
\[
Y = V - \frac{2 r_{,u}}{r_{,\lambda}},
\]
\[
\rho = r_{,u},
\]
\[
k_{AB} = h_{AB,u},
\]
together with
\[
\sigma = \frac14 m^A m^B h_{AB,\lambda},
\]
\[
\kappa = \frac14 m^A m^B k_{AB},
\]
\[
J = 4(r_{,\lambda} \kappa - \rho \sigma)
\]
[1303.6969]. The main equations then become
\[
r^{-1} r_{,\lambda\lambda} = -\sigma \bar\sigma,
\]
\[
(r h_{AB} W^B)_{,\lambda\lambda} = H_W[h_{CD},r],
\]
\[
Y_{,\lambda} = H_Y[h_{CD},r,W],
\]
\[
r J_{,\lambda} = H_J[h_{CD},r,W,Y],
\]
\[
\rho_{,\lambda\lambda} = H_\rho[h_{CD},r,J],
\]
so that once \(h_{AB}\) is known on a null hypersurface, one can integrate these equations in order to obtain
\[
[r, W, Y, J, \rho],
\]
and then reconstruct \(V\) from \(Y\) [1303.6969].

In the two-null-boundary formulation, the reorganization uses
\[
Y := W + \epsilon \frac{2r_{,w}}{r_{,\lambda}}, \qquad J := r_{,\lambda}(h_{AB,w}m^A m^B)-r_{,w}(h_{AB,\lambda}m^A m^B),
\]
together with
\[
\rho=r_{,w},\qquad \Theta=r_{,\lambda},\qquad n_{AB}=h_{AB,w},\qquad \sigma_{AB}=h_{AB,\lambda}
\]
[1810.04743]. The hypersurface hierarchy then becomes
\[
h_{AB,\lambda}=\sigma_{AB},
\]
\[
\frac{r_{,\lambda\lambda}}{r}=-\frac14 \sigma\bar\sigma,
\]
\[
\Theta=r_{,\lambda},
\]
\[
\epsilon\bigl(r^4 h_{AB}W^B_{,\lambda}\bigr)_{,\lambda} =2r^2 F_{A[W_{,\lambda\lambda}]},
\]
\[
\bigl[Y(r^2)_{,\lambda}\bigr]_{,\lambda} = \mathcal{R}+F_{[Y_{,\lambda}]},
\]
\[
-r\,\chi_{,\lambda}=F_{[\chi_{,\lambda}]},
\]
\[
\mu_{,\lambda}=F_{[\mu_{,\lambda}]},
\]
\[
\rho_{,\lambda}=\mu,
\]
with
\[
\chi := \frac{rJ}{\Theta}
\]
[1810.04743].

In spherical symmetry, the auxiliary variables simplify further. One set is
\[
\rho \equiv r_{,u}, \qquad \mathcal Y \equiv \mathcal V - 2\frac{\rho}{r_{,\lambda}}, \qquad K \equiv 2(r_{,\lambda}\Phi_{,u} - \rho\Phi_{,\lambda}),
\]
leading to
\[
r_{,\lambda\lambda} = -\frac{r}{2}\Phi_{,\lambda}\Phi_{,\lambda},
\]
\[
\big(\mathcal Y (r^2)_{,\lambda}\big)_{,\lambda} = 2,
\]
\[
r\left(\frac{rK}{r_{,\lambda}}\right)_{,\lambda} - (r^2\mathcal Y\Phi_{,\lambda})_{,\lambda}=0
\]
[1910.03439]. For horizon-penetrating evolution, the regularized variables
\[
L = \frac{rK - \lambda \Phi_{,\lambda}}{r_{,\lambda}},
\qquad
Q = \frac{P-1}{\lambda}
\]
lead to the final regular system
\[
r_{,\lambda\lambda} = -\frac{r}{2}\left(\Phi_{,\lambda}\right)^2,
\]
\[
L_{,\lambda} = \frac{\lambda\Phi_{,\lambda}}{r},
\]
\[
Q_{,\lambda} = \frac{1}{\lambda^2} - \frac{1}{r^2} + \frac{L^2}{2\lambda^2},
\]
\[
\Phi_{,u} = \frac{\lambda Q}{2}\Phi_{,\lambda} + \frac{1}{2}\Phi_{,\lambda} + \frac{L}{2r}
\]
[1910.03439]. The closely related apparent-horizon tracking formulation uses
\[
K := 2r_{,\lambda}\Phi_{,u} - 2r_{,u}\Phi_{,\lambda},
\]
\[
L := rK - \lambda\Phi_{,\lambda}
     = 2r\Phi_{,u} - rV\Phi_{,\lambda},
\]
\[
Q := \frac{V-1}{\lambda},
\]
and yields
\[
r_{,\lambda\lambda} = -\frac{r}{2}\Phi_{,\lambda}^2,
\]
\[
L_{,\lambda} = \frac{\lambda}{r}\Phi_{,\lambda},
\]
\[
Q_{,\lambda} = \frac{1}{\lambda^2}-\frac{1}{r^2}+\frac{1}{2}\left(\frac{L}{\lambda}\right)^2,
\]
\[
\Phi_{,u} = \frac{1+\lambda Q}{2}\Phi_{,\lambda} + \frac{L}{2r}
\]
[2405.19122].

In the conformal compactified Einstein-scalar system, the auxiliary fields are \(\mathcal Y\) and \(\mathcal L\), with
\[
\mathcal Y = \frac{x+(1-x)\mathcal Z_0(u)}{\mathcal R},
\]
\[
\Phi_{,u} = \frac12\left\{\frac{\mathcal L}{\mathcal R} + \mathcal W[(1-x)\Phi]_{,x}\right\},
\]
\[
\mathcal L_{,x} = \mathcal Y\,[(1-x)\Phi]_{,x},
\]
and the final hierarchy
\[
0=\mathcal R_{,xx} + \frac{\kappa}{2}\mathcal R\big([(1-x)\Phi]_{,x}\big)^2,
\]
\[
\mathcal L_{,x}=\frac{[x+(1-x)\mathcal Z_0][(1-x)\Phi]_{,x}}{\mathcal R},
\]
\[
\mathcal W_{,xx}= \kappa\,\frac{[(1-x)\Phi]_{,x}\mathcal L}{\mathcal R} +\frac{2[\mathcal R+(1-x)\mathcal R_{,x}]\mathcal Z_0}{\mathcal R^3} -\frac1x\left(\frac{x^2}{\mathcal R^2}\right)_{,x},
\]
\[
\Phi_{,u}=\frac12\left\{\frac{\mathcal L}{\mathcal R} +\mathcal W[(1-x)\Phi]_{,x}\right\}
\]
[2504.18973].

A plausible implication is that the “affine-null characteristic formulation” is best understood not as a single fixed reduced system but as a family of null-coordinate reductions that share one invariant feature: the affine gauge is geometrically simple, whereas the PDE hierarchy must be reconstructed.

## 4. Hyperbolicity, well-posedness, and tetrad-based reformulations

The hyperbolicity question is a central technical issue. One line of work shows that the second-order Einstein equations in Bondi-like coordinates, including affine gauge, are not strongly hyperbolic [2404.16720]. In the Bondi-like metric
\[
ds^2=-2G\,du\,dx-H\,du^2+R^2\left[e^{2Sf}S^{-1}(dy+Sb\,du)^2+e^{-2Sf}S\,d\varphi^2\right],
\]
affine gauge corresponds to prescribing \(G\), so the null generator
\[
U:=G^{-1}\partial_x
\]
is affinely parametrized [2404.16720]. The paper states that in the frozen-coefficient Minkowski limit, the problematic mode is
\[
{\bf k}_0=(-\xi,0,k),
\]
whose multiplicity exceeds the dimension of its eigenspace, so the second-order Bondi-like formulations are only weakly hyperbolic, not strongly hyperbolic [2404.16720]. Giannakopoulos, Hilditch, and Zilhão are likewise cited as having shown that the standard metric-based Bondi / affine-null evolution systems are only weakly hyperbolic, not strongly hyperbolic [2104.09972].

A distinct response is the tetrad-based Newman-Penrose reformulation in affine-null coordinates. For the outgoing case, the tetrad is chosen as
\[
l^\mu \partial_\mu = \partial_r, \qquad n^\mu \partial_\mu = \partial_u + P\,\partial_r + R^a \partial_a, \qquad m^\mu \partial_\mu = Q\,\partial_r + S^a \partial_a,
\]
with gauge fixing
\[
\kappa=\pi=\epsilon=0
\]
[2104.09972]. The evolved variables are
\[
\vec{\Psi}^T=(\Psi_0,\Psi_1,\Psi_2,\Psi_3,\Psi_4),
\]
\[
\vec{\Gamma}^T=(\rho,\sigma,\tau,\alpha,\beta,\gamma,\lambda,\mu,\nu),
\]
\[
\vec g^T=(n^v,n^r,n^a,m^v,m^r,m^a)
\]
[2104.09972]. The principal part is block diagonal, with
\[
\mathcal D_2^\mu = \mathrm{id}_9\times l^\mu, \qquad \mathcal D_3^\mu = \mathrm{id}_8\times l^\mu,
\]
and a \(5\times5\) Weyl block \(\mathcal D_1^\mu\) explicitly given in the paper [2104.09972]. The principal symbol is Hermitian, and for
\[
t^\mu = l^\mu+n^\mu,
\]
the contraction with the principal matrix is positive definite, so the system is symmetric hyperbolic [2104.09972].

The contrast is sharp. The second-order metric formulations in Bondi-like coordinates, including affine gauge, fail strong hyperbolicity [2404.16720]. The tetrad-based affine-null Newman-Penrose/GHP system is symmetric hyperbolic [2104.09972]. This suggests that the hyperbolicity problem is not a generic obstruction to affine-null coordinates as such, but to particular variable choices and reductions. That suggestion is reinforced by the first-order symmetric hyperbolic formulation of the linearized vacuum Einstein equations in Bondi gauge about Schwarzschild, which establishes a well-posed characteristic initial-boundary value problem in an \(L^2\)-type energy norm for data on an outgoing null cone and a timelike cylinder or an ingoing null cone [2404.16720].

## 5. Spherical symmetry, black holes, and matter-coupled extensions

Spherical symmetry has provided the main testing ground for the affine-null characteristic formulation. In vacuum, the two-intersecting-null-hypersurface framework yields a black-hole branch
\[
g^{(BH)}_{ab}dx^adx^b = \left(\frac{K}{2M}\right)\frac{\lambda^2}{r}\,dw^2 +2\epsilon\,dw\,d\lambda +r^2 f_{AB}dx^Adx^B, \qquad r = K\left(2M+\frac{\epsilon w\lambda}{4M}\right),
\]
identified as the double-null Israel black hole solution [1810.04743]. The horizon is at
\[
\lambda=0,
\]
and the singularity occurs where
\[
w\lambda = -8\epsilon M^2
\]
[1810.04743].

The same framework is used in spherical symmetry to recover Reissner-Nordström from the Einstein-Maxwell hierarchy with affine-null metric
\[
ds^2 = W(w,y)\,dw^2 + 2\,dw\,dy - r(w,y)^2\,q_{AB}dx^A dx^B
\]
and Maxwell gauge
\[
A_a = \alpha(w,y)\,dw
\]
[2107.10120]. In the non-extremal case, the direct coordinate-construction approach gives
\[
r = r_H - k_H w y,
\]
with
\[
r_H = m+\sqrt{m^2-Q^2},
\]
and
\[
ds^2 = \frac{(1-2k_H(r+r_H))y^2}{r^2}\,dw^2 + 2\,dw\,dy - r^2 d\Omega^2
\]
[2107.10120].

For dynamical scalar collapse, the affine-null system is especially well adapted to horizon penetration. The outgoing expansion is
\[
\Theta^+ = \frac{2r_{,\lambda}}{r},
\]
so an apparent horizon occurs where
\[
r_{,\lambda}(u,\lambda_A)=0
\]
[2405.19122]. The paper derives analytically that the apparent horizon is spacelike:
\[
g^{ab}\nabla_a r_{,\lambda}\nabla_b r_{,\lambda}\big|_{\lambda=\lambda_A} = -\frac{\Phi_{,\lambda}^2}{2}\bigg|_{\lambda_A}<0
\]
[2405.19122]. It also shows that the final singularity traced by the caustics is spacelike [2405.19122]. Earlier numerical work had already emphasized that the affine parameter remains well behaved across the horizon and inside black-hole regions up to the final singularity [1910.03439].

The spherically symmetric Einstein-massless-scalar system also yields asymptotic quantities in affine-null form. The Misner-Sharp mass is
\[
m(u,\lambda)=\frac{1}{2}(r-\lambda r_{,\lambda}),
\]
and the Bondi mass is
\[
M_B(u)=\lim_{\lambda\to\infty}m(u,\lambda)=\frac{R(u)}{2}
\]
[1910.03439]. The mass-loss law becomes
\[
M_{B,u} = -\frac{1}{2}H\big[(H\alpha)_{,u}\big]^2
\]
[1910.03439].

Matter-coupled generalizations preserve the same hierarchical logic. In \(n\)-dimensional spherical symmetry for the Einstein-Maxwell system with a charged massless complex scalar field, the Maxwell potential is chosen in radial gauge,
\[
A_\lambda=0, \qquad A_a dx^a=\alpha(w,\lambda)\,dw,
\]
with enclosed charge
\[
Q=-\epsilon r^{n-2}\alpha_{,\lambda}
\]
[2606.13636]. The auxiliary variables are
\[
Z = r^{n-3}\left(r_{,\lambda}V+2\epsilon r_{,w}\right),
\]
\[
\mathcal{L}= \left[2r\left(\Phi_{,w}r_{,\lambda}-r_{,w}\Phi_{,\lambda}\right) +\frac{\epsilon Z\Phi_{,\lambda}}{r^{n-4}}\right]\frac{1}{r_{,\lambda}}
\]
[2606.13636]. The hierarchical equations include
\[
r_{,\lambda\lambda} = -\frac{\kappa r}{n-2}|\Phi_{,\lambda}|^2,
\]
\[
Q_{,\lambda} = \Omega_{n-2}q\,r^{n-2}\operatorname{Im}\!\left(\overline{\Phi}\,\Phi_{,\lambda}\right),
\]
\[
\alpha_{,\lambda} = -\frac{\epsilon Q}{r^{n-2}},
\]
\[
Z_{,\lambda} = (n-3)r^{n-4} -\frac{\kappa}{n-2}\frac{Q^2}{\Omega_{n-2}r^{n-2}},
\]
and scalar transport
\[
\Phi_{,w}=\frac{\mathcal{L}}{2r}-\frac{\epsilon V\Phi_{,\lambda}}{2}
\]
[2606.13636]. As a consistency check, the scalar-free limit recovers both non-extremal and extremal Reissner-Nordström-Tangherlini branches directly from the hierarchy [2606.13636].

## 6. Asymptotics, conformal compactification, and applications

Affine-null formulations are closely tied to null infinity because outgoing affine-null coordinates are natural for radiation reaching future null infinity and are useful in Cauchy-characteristic extraction [2104.09972]. The original motivation for the metric formulation was precisely that, in Cauchy-characteristic extraction, the inner worldtube supplied by a Cauchy code is generally not located on the \(r=\) const grid of a Bondi-Sachs code, so one must first interpolate Cauchy data into an auxiliary affine-null system anyway [1303.6969]. Using affine-null directly can eliminate that extra step and reduce interpolation error [1303.6969].

In the conformal compactification of the affine-null Einstein-scalar system, the physical coordinates are
\[
\tilde x^a=(\tilde u,\tilde\lambda,\tilde x^A),
\]
with line element
\[
d\tilde s^2_{AN} = -W\, d\tilde u^2 -2\, d\tilde u\, d\tilde\lambda + r^2 q_{AB} d\tilde x^A d\tilde x^B
\]
[2504.18973]. The conformal factor and compactified coordinate are chosen as
\[
\Omega(\tilde\lambda)=\frac{1}{1+\tilde\lambda},
\qquad
x=\frac{\tilde\lambda}{1+\tilde\lambda}\in[0,1],
\qquad
\Omega=1-x,
\]
so that \(\mathcal I^+\) sits at \(x=1\) [2504.18973]. With
\[
\mathcal R := (1-x)r,\qquad \mathcal W := (1-x)W,\qquad \Phi=\frac{\tilde\Phi}{\Omega},
\]
the unphysical line element becomes
\[
ds^2 = -(1-x)\mathcal W\,du^2 -2\,du\,dx + \mathcal R^2 q_{AB} dx^A dx^B
\]
[2504.18973]. The asymptotic expansion yields the Bondi mass
\[
m_B=\frac12\big[H(1-\mathcal Z_0)+\mathcal R_{[1]}\big]
\]
and the balance law
\[
H\frac{dm_B}{du} =-\frac12\kappa\left(H\frac{dC}{du}\right)^2, \qquad C(u)=\lim_{x\to1}\mathcal R\Phi=H\Phi_{[0]}
\]
[2504.18973]. In an inertial conformal frame on \(\mathcal I^+\),
\[
0=\frac{dm_B}{du_b}+\frac12\kappa N^2(u_b), \qquad N(u_b)=\frac{dC}{du_b}
\]
[2504.18973].

The same paper shows that the conformal hierarchy is exactly equivalent to the system obtained by compactifying the affine coordinate in the physical spacetime and introducing regularized variables
\[
R=(1-\hat x)r,\qquad W=(1-\hat x)V,\qquad \Phi=\frac{\tilde\Phi}{1-\hat x}
\]
[2504.18973]. This equivalence is presented as an identity of PDE systems, differing only in interpretation: in the conformal picture the boundary is geometrically built in, whereas in the physical compactified picture it is represented by a finite grid point after renormalization [2504.18973].

Cosmological applications exploit the same geometric advantage of affine parametrization. The affine-null cosmology metric is
\[
ds^2 = -\left(1+\frac{\hat{W}}{\hat{r}}\right)du^2 - 2\,du\,d\lambda + \hat{r}^2\left(d\theta^2+\sin^2\theta\,d\varphi^2\right),
\]
with regularity conditions at the vertex
\[
\hat r(0)=0, \qquad \hat W(0)=0, \qquad \hat W_{,\lambda}(0)=0, \qquad \hat r_{,u}(0)=0, \qquad \hat r_{,\lambda}(0)=1
\]
[1111.6025]. The radial geodesic relation
\[
\frac{dr}{d\lambda}=e^{-2\beta}
\]
shows how the affine parameter is tied to the null generator [1111.6025]. The numerical calculations are reported to have the same stability and convergence properties as the standard characteristic formalism [1111.6025].

Finally, affine-null coordinates have also been used to derive stationary rotating solutions directly from the characteristic hierarchy. In the slowly rotating Kerr construction, the affine-null metric ansatz is
\[
g_{ab}dx^adx^b = -W\,du^2 -2\,du\,d\lambda +R^2 h_{AB}(dx^A-W^A du)(dx^B-W^B du),
\]
with \(\det(h_{AB})=\det(q_{AB})=\sin^2\theta\) [2301.05092]. By solving the vacuum Einstein equations order by order in slow rotation, the final metric is shown to be equivalent to Kerr in the slowly rotating approximation after an explicit transformation from Boyer-Lindquist coordinates [2301.05092].

Taken together, these results support a broad characterization. The affine-null characteristic formulation is a null-coordinate framework in which the affine parameter regularizes horizon and refocussing behavior, restores direct contact with null infinity and null boundaries, and admits hierarchical evolution only after suitable auxiliary-field reformulation [1303.6969]. The main controversy is not over its geometric value, which is used repeatedly across vacuum, scalar, Maxwell, cosmological, and black-hole problems [1111.6025; 1810.04743; 1910.03439; 2606.13636], but over the PDE structure of specific reductions: raw second-order metric systems in Bondi-like gauges are only weakly hyperbolic [2404.16720], whereas tetrad-based or more elaborate first-order reductions can recover symmetric hyperbolicity or a controlled characteristic initial-boundary value problem [2104.09972; 2404.16720].

Source: https://www.emergentmind.com/topics/affine-null-characteristic-formulation