Logarithmic matching between past infinity and future infinity: The massless scalar field

Oscar Fuentealba, and Marc Henneaux · 2024

Read Logarithmic matching between past infinity and future infinity: The massless scalar field by Oscar Fuentealba, and Marc Henneaux (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-05088.txt.

Introduction

In flat spacetime, matching what a field does at past null infinity to what it does at future null infinity underpins the study of symmetries and conservation laws. Oscar Fuentealba and Marc Henneaux point out that the standard matching rules assume the field's expansion near infinity has no leading logarithmic term, an assumption that often fails. Their paper, the first in a series, works out the corrected rules for a massless scalar field.

When dominant logarithms are present, the matching conditions come out with the opposite sign to the usual ones. The authors also check the angle-dependent conserved charges built from the field's decay and Lorentz symmetry. These stay well defined and finite under the right definitions, even with leading logarithms. They extend the analysis to higher dimensions, where fractional powers or weaker logarithms appear, and plan to treat light and gravity next.

Catalog record

CREATOROscar Fuentealba, and Marc Henneaux
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE98 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-05088

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Logarithmic matching between past infinity and future infinity: The massless scalar field Oscar Fuentealba, and Marc Henneaux Abstract Matching conditions relating the fields at the future of past null infinity with the fields at the past of future null infinity play a central role in the analysis of asymptotic symmetries and conservation laws in asymptotically flat spacetimes. These matching conditions can be derived from initial data given on a Cauchy hypersurface by integrating forward and backward in time the field equations to leading order in an asymptotic expansion, all the way to future and past null infinities. The standard matching conditions considered in the literature are valid only in the case when the expansion near null infinity (which is generically polylogarithmic) has no dominant logarithmic term. The absence of dominant logarithmic term, in turn, holds only when the leading order of the initial conditions on a Cauchy hypersurface (which contains no logarithm) fulfills definite parity conditions under the antipodal map of the sphere at infinity. One can consistently consider opposite parity conditions. While these do not conflict with the asymptotic symmetry group, they lead to a very different asymptotic behaviour near null infinity, where the expansion starts now with logarithmic terms that are no longer subdominant (even though such logarithmic terms are absent in the initial data), which implies different matching conditions. It turns out that many of the analytic features relevant to gravity are already present for massless spin zero and spin one fields. This paper is the first in a series in which we derive the matching conditions for a massless scalar field with initial conditions leading to logarithms at null infinity. We prove that these involve the opposite sign with respect to the usual matching conditions. We also analyse the matching of the angle-dependent conserved charges that follow from the asymptotic decay and Lorentz invariance. We show in particular that these are well defined and finite at null infinity even in the presence of leading logarithmic terms provided one uses the correct definitions. The free massless scalar field has the virtue of presenting the polylogarithmic features in a particularly clear setting that shows their inevitability, since there is no subtle gauge fixing issue or nonlinear intrincacies involved in the problem. We also consider the case of higher spacetime dimensions where fractional powers of r r (odd spacetime dimensions) or subdominant logarithmic terms (even spacetime dimensions) are present. Mixed matching conditions are then relevant. In subsequent papers, we will extend the analysis to the electromagnetic and the gravitational fields. 1 Introduction The question of the behaviour of matter and gravitational fields in the vicinity of null infinity is a non-trivial dynamical question. From the perspective of the Cauchy problem, this question can be phrased as follows: given “reasonable” initial conditions on a Cauchy hypersurface, what type of expansions do the resulting fields admit near future null infinity, in particular, close to the “critical sets" where spatial infinity and null infinity “meet" [ 1 , 2 , 3 , 4 , 5 , 6 ] 1 1 1 These critical sets correspond respectively to the past of future null infinity and the future of past null infinity and are very clearly described in [ 7 , 8 , 9 , 10 ] . ? The reason that the question of the behaviour near null infinity is non-trivial is that spatial infinity and the critical sets define points with very different properties from the point of view of the relevant linear second order differential equations that govern the asymptotic dynamical evolution. While spatial infinity corresponds to a regular point of these differential equations, the critical sets correspond by contrast to Fuchsian singularities. For D=4 and the leading order in the 1/r expansion, the roots of the indicial equation coincide. One independent solution is then logarithm-free - actually, polynomial - and defines what is called the “ P -branch". The other independent solution involves logarithms and defines the “ Q -branch". General solutions are linear combinations of both branches. As a result, even if the initial data obey standard expansions in 1/r at spatial infinity, the fields will generically develop polylogarithmic terms of the form r^{-k}\ln^{m}r near null infinity, as it is well documented (see [ 1 , 2 , 3 , 4 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 , 30 ] and references therein). These polylogarithmic terms are not related to nonlinearities or to the presence of sources not vanishing at null infinity, or linked to a bad choice of gauge. This is particularly clear for a free massless scalar field \phi in Minkowski space - a linear system with no gauge invariance. One finds that if the scalar field behaves asymptotically at spatial infinity as \phi\sim\frac{1}{r} , its behaviour at null infinity (in the vicinity of the critical sets) is \phi\sim\frac{\log r}{r} (rather than the expected \phi\sim\frac{1}{r} ), unless the leading \frac{1}{r} -order of the initial data at spatial infinity is even under the antipodal map on the sphere. In that case, the coefficient of the leading \frac{\log r}{r} -term at null infinity is zero, and the expansion starts with \frac{1}{r} , with subdominant terms of order \frac{\log r}{r^{2}} and lower [ 23 ] . But if no parity condition is imposed on the leading \frac{1}{r} -order at spatial infinity, dominant terms of order \frac{\log r}{r} are unavoidable. The reason that parity conditions control the absence or presence of logarithms is due to the fact that the P -branch and the Q -branch obey opposite parity properties under the antipodal map of the hyperboloid at infinity, i.e., the combined antipodal map of the sphere at infinity with hyperbolic time reversal, which exchanges the past of future null infinity with the future of past null infinity. The P -branch is even while the Q -branch is odd. Imposing even or odd parity conditions on the leading order of the initial data eliminates the dominant term of either the Q -branch or of the P -branch. Similar features hold for electromagnetism or gravity where the leading terms in the expansion near null infinity involve (in four spacetime dimensions) a dominant logarithmic term (e.g. of order \frac{\log r}{r^{2}} for the radial electric field rather than the expected order \frac{1}{r^{2}} ) unless the initial conditions are appropriately constrained by (twisted) parity conditions [ 20 , 21 , 22 ] . Even if one imposes the appropriate parity conditions that eliminate the leading logarithmic term, logarithms will appear in the expansion at subdominant orders. The presence of polylogarithmic terms at null infinity is sometimes feared to conflict with the asymptotic symmetries. We recall that the asymptotic symmetry groups of electromagnetism and gravity are infinite-dimensional and equal either to the group of angle-dependent U(1) transformations for electromagnetism [ 31 , 32 , 33 ] or the BMS group for gravity [ 34 , 35 , 36 , 37 , 38 , 39 , 40 ] . These groups, obtained first at

Read the complete text