This question arose upon reading a New Scientist article, sometime last year, about Hawking radiation at the event horizons of black holes. The article also discussed the analogous generation of Hawking radiation at the de Sitter horizon of the observable universe. My initial question to the author of the article concerned the observer-dependent status of the de Sitter horizon with the consequence that two virtual particles which for observer A existed at the de Sitter horizon could lie well within the de Sitter horizon of observer B. It seemed to me to demand a paradox that a particle, which for A became a real particle with an effectively unlimited lifetime, would remain virtual (and therefore be almost instantaneously destroyed) for B.
The author (whose name I cannot remember) mentioned that other readers pointed out the same apparent paradox and he suggested that Time Dilation might come to the rescue in the sense that, for A, the virtual particles at the de Sitter horizon would appear to take an infinitely long time to disappear over the horizon. My immediate problem with this is that it would seem to preclude the very existence of Hawking radiation. I shared this concern with the author in a second e-mail, after which he admitted that the problem was too hard and advised that I contact “an expert in the field”! However, subsequent letters to other people have not fared any better.
Another problem in my mind is that if Time Dilation is indeed relevant, would it not be relevant also at the event horizon of a black hole and would this mean that black hole evaporation should take an infinite time?

David A. J. Seargent

• Saying... - bruce - May 23, 2003 - 04:54 UTC
  • Correction - bruce - May 24, 2003 - 21:57 UTC
• de Sitter type radiation fills the universe - livarot - June 30, 2003 - 18:23 UTC