No Access
Published Online: 04 June 1998
Journal of Mathematical Physics 23, 810 (1982); https://doi.org/10.1063/1.525421
more...View Affiliations
  • Enrico Fermi Institute, University of Chicago, Chicago, Illinois 60637
We prove that given an asymptotically flat (in a very weak sense) initial data set, there always exists a spinor field that satisfies Witten’s equation and that becomes constant at infinity. Thus we fill a gap in Witten’s arguments on the nonnegativity of the total mass of an isolated system, when measured at spatial infinity. We also include a review of Witten’s argument.
  1. 1. H. Bondi, M. G. J. Van Der Berg, and A. W. K. Metzner, Proc. R. Soc. London Ser. A 269, 21 (1962); Google ScholarCrossref
    R. Sachs, Proc. R. Soc. London Ser. A 270, 193 (1962). , Google ScholarCrossref
  2. 2. R. Arnowitt, S. Deser, and C. W. Misner, Phys. Rev. 118, 1100 (1960); Google ScholarCrossref
    R. Arnowitt, S. Deser, and C. W. Misner, 121, 1556 (1961); , Phys. Rev. , Google ScholarCrossref
    R. Arnowitt, S. Deser, and C. W. Misner, 122, 997 (1961); , Phys. Rev. , Google ScholarCrossref
    R. Geroch, J. Math. Phys. 13, 956 (1972). , Google ScholarScitation
  3. 3. D. Brill, Ann. Phys. 7, 466 (1959). Google ScholarCrossref
  4. 4. D. Brill and S. Deser, Am. Phys. 50, 548 (1968). Google Scholar
  5. 5. D. Brill, S. Deser, and L. Fadeev, Phys. Lett. A 26, 538 (1968). Google ScholarCrossref
  6. 6. R. Arnowitt, S. Deser, and C. W. Misner, Ann. Phys. (N. Y.) 11, 116 (1960). Google ScholarCrossref
  7. 7. C. Leibovitz and W. Israel, Phys. Rev. D 1, 3226 (1970). Google ScholarCrossref
  8. 8. C. W. Misner, Astrophysics and General Relativity, edited by M. Chretieu, S. Deser, and J. Goldstein (Gordon and Breach, New York, 1968). Google Scholar
  9. 9. R. Geroch, Ann. N. Y. Acad. Sci. 224, 108 (1973). Google ScholarCrossref
  10. 10. P. S. Jang, J. Math. Phys. 17, 141 (1976). Google ScholarScitation
  11. 11. N. O’Murchadha and J. W. York, Jr., Phys. Rev. D 10, 2345 (1974). Google ScholarCrossref
  12. 12. M. Leite, Ph.D. Dissertation, M.I.T., 1977 (unpublished). Google Scholar
  13. 13. P. S. Jang, J. Math. Phys. 19, 1155 (1978). Google ScholarScitation
  14. 14. P. Schoen and S. T. Yau, Phys. Rev. Lett. 42, 547 (1979). Google ScholarCrossref
  15. 15. E. Witten, Commun. Math. Phys. 80, 381 (1981). Google ScholarCrossref
  16. 16. A. Sen, “A Quantum Theory of Spin 3/2 Field in Einstein Spaces,” to appear in Int. J. Theor. Phys. Google Scholar
  17. 17. A. Sen, J. Math. Phys. 22 (8), 1781 (1981). Google ScholarScitation
  18. 18. With the adjoint given by λ†A = 2̄tABλ̄B, one gets DABλc = 2t(ABB)Bλc, where BB is the spinor representation of the space‐time covariant derivative. Thus DABλC is the neutrino zero mode equation or Witten’s equation. Google Scholar
  19. 19. This identity is the spinorial version of the Green identity for scalars. Google Scholar
  20. 20. It is convenient for simplicity to use the second order equation (1.6) instead of Witten’s equation. DABβB = 0. In fact, they are equivalent, in the sense that every solution of (1.6) found in Sec. II satisfies Witten’s equation.25, Google Scholar
  21. 21. In fact, our definition is so weak that it even admits initial data sets with metrics not “approaching” the metric ηab, of the definition, in all directions. As an example consider the “spiked” space with metric (1+exp[−x2y2z2])ηab. Google Scholar
  22. 22. Lp denotes the Banach space of measurable functions on S such that |f|p. is integrable. Its norm is given by ‖f‖Lp≡(∫s|f|pdV)1/p. Google Scholar
  23. 23. C. M. Morrey, Multiple Integrals in the Calculus of Variations (Springer, New York, 1966), p. 81. Google Scholar
  24. 24. A. Friedman, Partial Differential Equations (Holt, Reinhart, and Winston, New York, 1969), p. 54. Google Scholar
  25. 25. Y. Choquet‐Bruhat and D. Cristodoulou, Acta Math. 146, 129 (1981). Google ScholarCrossref
  26. 26.To see this, note that for any of our solutions of (l.6), βA, λB = DABβA is square integrable. Thus, it suffices to show that the equation DBCλB = 0 has no nontrivial solution in L2∩C. Assume, on the contrary, there exists one, λA. Then, using identity (1.5) for DCADBCλB = 0, and the local mass condition (1.9) we get
    s(r)(DABλC)(DABλC)dV⩽|∫Σ(r)†CDaλC)dΣa|
    Here, Σ(r), with r∈[0,∞], denotes a typical element of a one‐parameter family of nested surfaces (topologically S2), and S(r) denotes the region enclosed by the Σ(r). The surfaces and the parameters are to be chosen so that dV⩾drdΣ. Calling the left‐hand side of th expression above g2(r), integrating it over r∈[0,x], and using the Hölder inequality, we get 0xg2(r)dr⩽cg(x), where c2 = ∫s†CλC)dV. But one can show there exists no function g which is positive, nondecreasing, and defined everywhere on [0,∞] and which satisfies the above inequality.
  27. 27. It is immediate, from the form of βC, that the volume integral in (1.7) is finite for any asymptotically flat initial data set, in the sense of Sec. II. Google Scholar
  28. 28. To see this, first note that T is diffeomorphic to 12R3. The required flat metric on T is then chosen to agree with ηab outside some compact set. Google Scholar
  1. © 1982 American Institute of Physics.