10
9/6/11 1 Thanks to the organizers for bringing us together!

GTC119/6/11 4 € Lorentz invariance and energy positivity imply the Minkowsi vacuum is a thermal state when restricted to the wedge: ρ R =Tr L 00∝exp− 2π H Boost & ’ * + Bisognano-Wichmann

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Page 1: GTC119/6/11 4 € Lorentz invariance and energy positivity imply the Minkowsi vacuum is a thermal state when restricted to the wedge: ρ R =Tr L 00∝exp− 2π H Boost & ’ * + Bisognano-Wichmann

9/6/11  

1  

Thanks  to  the  organizers  for  bringing  us  together!    

Page 2: GTC119/6/11 4 € Lorentz invariance and energy positivity imply the Minkowsi vacuum is a thermal state when restricted to the wedge: ρ R =Tr L 00∝exp− 2π H Boost & ’ * + Bisognano-Wichmann

9/6/11  

2  

Horizon  entropy  and  higher  curvature  equa?ons  of  state  

Ted  Jacobson  University  of  Maryland  

Plan  of  talk:  

  Horizon  entropy  &  Einstein  Equa?on  of  State      

  Higher  curvature  

  Noetheresque  approach  

Based  on:      

TJ,  ‘95  paper  

Chris  Eling,    Raf  Guedens,  TJ,    ’06  Non-­‐equilibrium  paper  

Raf  Guedens,  TJ,    Sudipta  Sarkar    ‘11  Noetheresque  paper    

Page 3: GTC119/6/11 4 € Lorentz invariance and energy positivity imply the Minkowsi vacuum is a thermal state when restricted to the wedge: ρ R =Tr L 00∝exp− 2π H Boost & ’ * + Bisognano-Wichmann

9/6/11  

3  

Credo:  

Black  hole  thermodynamics  is  a  special  case  of  causal  horizon  thermodynamics.  

It  originates  from  thermality  of  the  Lorentz-­‐invariant  local  vacuum  in  local  Rindler  wedge,  at  temperature                                                        rela?ve  to  the  local    boost  Hamiltonian.  

This  thermal  state  has  a  huge  entropy,  somehow  rendered  finite  by  quantum  gravity  effects,    that  scales,  for  any  local  causal  horizon  (LCH)  at  leading  order,  like  the  area.  

Horizon  evolu?on  can  be  viewed  as  a  small  perturba?on  of  a  local  equilibrium  state,  so  this  entropy  sa?sfies  the  Clausius  rela?on  

When  applied  to  all  LCH’s,  together  with  conserva?on  of  ma\er  energy-­‐momentum,    this  implies  the  Einstein  equa?on.  

TB = �/2π

δS =δQB

TB

Euclidean  space   Minkowski  space  

ds2 = dx2 + dy2 = dr2 + r2dθ2 ds2 = dt2 − dx2 = l2dη2 − dl2

RotaAon  symmetry   Lorentz  boost  symmetry  

Page 4: GTC119/6/11 4 € Lorentz invariance and energy positivity imply the Minkowsi vacuum is a thermal state when restricted to the wedge: ρ R =Tr L 00∝exp− 2π H Boost & ’ * + Bisognano-Wichmann

9/6/11  

4  

Lorentz invariance and energy positivity imply the Minkowsi vacuum is a thermal state when restricted to the wedge:

ρR = TrL 0 0 ∝ exp − 2πHBoost

⎝ ⎜

⎠ ⎟

Bisognano-Wichmann (1975), Davies (1975), Unruh (1976)

A uniformly accelerated observer a distance l from the horizon sees the temperature, Tlocal = a /2π = /2π l.

Davies-­‐Unruh  effect  

L   R  

Accelera?on  and    Tlocal  diverge  as  l  goes  to  0.  

Vacuum  entanglement  entropy    

S = −Tr(ρR ln ρR) ≈�

dA dl T 3local ∝

�dA dl l−3

S = α A

Hypothesis:    physics  regulates  the  divergence:  

Page 5: GTC119/6/11 4 € Lorentz invariance and energy positivity imply the Minkowsi vacuum is a thermal state when restricted to the wedge: ρ R =Tr L 00∝exp− 2π H Boost & ’ * + Bisognano-Wichmann

9/6/11  

5  

Horizon  thermodynamics  

S = α A

2.  Boost  energy  flux  across  the  horizon  is    ‘thermalized’  at  the  Unruh  temperature.  

1.  The  horizon  system  is  a  ‘heat  bath’,    with  universal  entropy  area  density.  Postulate  for  all  

such  horizons  

Implies  focusing  of  light  rays  by  space?me  curvature:  the  causal  structure  must  sa?sfy  Einstein  field  equa?on,  with  Newton’s  constant  

G= 14α

3.  Energy  conserva?on  (energy-­‐momentum  tensor  divergence-­‐free)    

Adjust  horizon  so  expansion  [and  shear]*  at  final  point                              vanishes.  Then  null  geodesic  focusing  eqn  implies*    

δQ

T≈ 2π

H

(−λTabkakb) dλ dA

Boost  Killing  vector  that  vanishes  at                                  is                                                                                                                          so  

δS ≈ α

H

(−λRabkakb) dλ dA

Rab =2π

�αTab + Φgab

If  this  holds  for  all  horizons  it  follows  that,  for  some                Φ

λ = 0

λ = 0

*Three  op?ons:  1.    thin  patch  2.  symmetric  patch  3.  Arrange  vanishing    expansion  on  whole    final  surface?  

ξa = −λk

a + O(x3)

*Shear  ok  if  allow  viscous  entropy    produc?on  term  in  Clausius  rela?on  with  viscosity  ra?o  η/s = �/4π

Page 6: GTC119/6/11 4 € Lorentz invariance and energy positivity imply the Minkowsi vacuum is a thermal state when restricted to the wedge: ρ R =Tr L 00∝exp− 2π H Boost & ’ * + Bisognano-Wichmann

9/6/11  

6  

Rab =2π

�αTab + Φgab

Ma\er  energy  conserva?on    ∇aTab = 0 implies  ∇aRab = ∇bΦ

Bianchi  iden?ty  implies    ∇aRab =12∇bR

Φ =12R + ΛHence  

               is    a  cosmological  constant,  and  Newton’s  constant  is  given  by    Λ

G =1

4�αNota  bene:                                                  implies    

I  conjecture  the  converse  holds  as  well.    

α <∞ G �= 0

Where  to  place  the  bifurca?on  point  of  the  Killing  vector?    

Spacelike  outside  the  horizon.  

Timelike  outside  the  horizon.  

Varia?on  away  from  a  sta?onary  state,  like  the  physical  process  1st  law.  Changes  defini?on  of  heat  current,  but  not  the  TOTAL  heat  flux  if  integrate  from  p0  to  p:  � p

p0

TabξbnewdH

b =�

dA

� 0

λ0

dλ(λ− λ0)Tkk =�

dA

� 0

λ0

dλ(−λ)Tkk

ξanew = (λ− λ0)ka

ξa = −λka

Page 7: GTC119/6/11 4 € Lorentz invariance and energy positivity imply the Minkowsi vacuum is a thermal state when restricted to the wedge: ρ R =Tr L 00∝exp− 2π H Boost & ’ * + Bisognano-Wichmann

9/6/11  

7  

Sgrav ∼1l2P

�d4V (Λ + R + l2R2 + l�4R3 + · · · )

Can  higher  curvature  terms  be  captured  in  the  equa?on  of  state  of  vacuum  thermodynamics?  

If  “natural”,  i.e.  only  one  length  scale,  the  answer  is  NO!                  Let  R  ~  lc-­‐2  .  

Rela?ve  ambiguity  of  local  KV  and  therefore  heat  is  O(x2/lc2).    

Smallest  x  we  can  use  is  the  Planck  length  lP,  so  minimum  ambiguity  of  the  heat  is  ~  lP2/lc2.  

But  this  is  the  rela?ve  size  of  the  R2  term!  The  R3  term  is  even  more  suppressed.  

What  if  NOT  natural,  i.e.  if    l  >>lP  ??        Then  I  claim  the  smallest  length  we  can  use  is  l,  so  once  again  the  R2  term  is  lost  in  the  noise.    (See  forthcoming  paper.)    

CAVEAT:  The  rela?ve  ambiguity  of  the  KV  on  the  horizon  can  be  one  order  smaller  if  one    Imposes  a  further  condi?on  that  it  vanish  on  a  geodesic  D-­‐2  surface  through  p…  

Nevertheless,  we  and  many  others  tried  to  include  the  higher  derivaAve  terms…  

Special  case:  s(R).  Then  since  dR  is  nonzero  in  general,  the  entropy  rate  of  change    has  an  order  unity  part,  which  cannot  match  the  heat.  This  can  be  dealt  with  in    many  ways:  

1.  Conformal  transforma?on  (field  redefini?on)  to  “Einstein  frame”:  s(R)  √h  =  √h’  

2.      Cancel  with  nonzero  expansion  at  p,  allow  for  bulk  viscous  entropy  produc?on.                (Eling,  Guedens,  TJ)  3.  Auxiliary  scalar  field  (Chirco,  Eling,  Libera?)          

Might  declare  victory  in  this  special  case,    but  what  about  the  general  case?  

Page 8: GTC119/6/11 4 € Lorentz invariance and energy positivity imply the Minkowsi vacuum is a thermal state when restricted to the wedge: ρ R =Tr L 00∝exp− 2π H Boost & ’ * + Bisognano-Wichmann

9/6/11  

8  

General  higher  curvature  entropies?  

The  s(R)  approaches  don’t  work.  Need  to  account  for  the  change  of  the    Different  projec?ons  of  curvature  that  occur  in  the  entropy.  Lovelock  case  looks  sweet:  e.g.  Gauss-­‐Bonnet  gives  integral  of  intrinsic  Ricci  scalar  of  the  horizon  cut.  The  problem:  how  to  evaluate  the  evolu?on  of  this  quan?ty??  

Idea:  try  to  mimic  the  way  entropy  density  of  sta?onary  black  holes  occurs    in  Wald’s  Noether  charge  approach.    (Brustein  &  Hadad;  Parikh  &  Sarkar;  Padmanabhan;  Guedens,  TJ  &  Sarkar)  

S =�

sabNabdA, sab = P abmn∇mξn + W abnξn

There  are  various  problems  and  issues  with  the  previous  work.  Awer  two    years  of  struggle  we  think  we’ve  sorted  it  all  out,  and  the  paper  will  appear  shortly  –  perhaps  even  during  this  conference!    

Some  aspects  of  this  Noetheresque  approach:  

1.  Depends  on  KV,  which  is  not  unique.  Sta?s?cal  interpreta?on?  

2.  Evaluate  change  of  entropy  using  Stoke’s  theorem:  two  horizon  slices  must  share  a              common  boundary  (see  Fig.).      

3.    LKV  must  sa?sfy  Killing  equa?on  and  Killing  iden?ty  to  required  order.  Using  a  null            geodesic  normal  coordinate  system  adapted  to  the  horizon  we  showed  this  can  be  done            but  only  if  the  horizon  patch  is  parametrically  narrower  than  it  is  long.    

4.  We  show  that  the  symmetric  part                                              does  not  contribute  to  the  entropy  change.  

5.  Clausius  rela?on  implies:    (The  W  term  cancels  the  contribu?on  from  the  deriva?ve  of  s(R)  in  that  case.)    

6.  The  equa?on  of  state  is  integrable  for              if    

in  which  case  it  is  the  field  equa?on  for  the  Lagrangian  L.  We  don’t  know  if  this  is  the  only  way  the  integrability  condi?on  can  be  sa?sfied,  but  it  seems  it  might  be.  

7.  We  must  take  the  bifurca?on  surface  as  the  earlier  surface,  not  the  later  one,    If  the  entropy    on  a  constant  V  slice  is  to  be  the  area  and  not  minus  the  area  (in  the  GR  case).  

W a[de] +∇cPac[de] = 0

P ab(mn)

P ab[mn] = −∂L[g, Rpqrs]∂Rabmn

Φ

Page 9: GTC119/6/11 4 € Lorentz invariance and energy positivity imply the Minkowsi vacuum is a thermal state when restricted to the wedge: ρ R =Tr L 00∝exp− 2π H Boost & ’ * + Bisognano-Wichmann

9/6/11  

9  

p

!

!0

ka

V = V0

V = 0

!

P0 !a = 0( )

! p

Conclusion:    

Let’s  talk  about  it…  

Page 10: GTC119/6/11 4 € Lorentz invariance and energy positivity imply the Minkowsi vacuum is a thermal state when restricted to the wedge: ρ R =Tr L 00∝exp− 2π H Boost & ’ * + Bisognano-Wichmann

9/6/11  

10