|Views: 0
|Likes: 0

Published by novarex123

See more

See less

Detecting Vanishing Dimensions via Primordial Gravitational Wave Astronomy

Jonas Mureika

1

and Dejan Stojkovic

2

1

Department of Physics, Loyola Marymount University, Los Angeles, California 90045, USA

2

Department of Physics, SUNY at Buffalo, Buffalo, New York 14260-1500, USA

(Received 7 November 2010; published 8 March 2011)Lower dimensionality at higher energies has manifold theoretical advantages as recently pointed out byAnchordoqui

et al.

[arXiv:1003.5914]. Moreover, it appears that experimental evidence may already existfor it: A statistically signiﬁcant planar alignment of events with energies higher than TeV has beenobserved in some earlier cosmic ray experiments. We propose a robust and independent test for this newparadigm. Since (

2

þ

1

)-dimensional spacetimes have no gravitational degrees of freedom, gravity wavescannot be produced in that epoch. This places a universal maximum frequency at which primordial wavescan propagate, marked by the transition between dimensions. We show that this cutoff frequency may beaccessible to future gravitational wave detectors such as the Laser Interferometer Space Antenna.

À

w

There is growing theoretical evidence to suggest that theshort-distance spatial dimensionality is

less

than the mac-roscopically observed three. Causal dynamical triangula-tions [1] demonstrate that the four-dimensional spacetimecan emerge from two-dimensional simplicial complexes. Ithas also been shown that a noncommutative quantumspacetime with minimal length scale will exhibit the prop-erties of a two-dimensional manifold [2]. Reducing thenumber of dimensions in the far UV limit offers a com-pletely new approach to gauge couplings uniﬁcation [3]. Ina similar vein, the cascading Dvali-Gabadadze-Porratimodel [4] provides a mechanism for the emergence of anextra spatial dimension only at Hubble scales, in order tosolve the cosmological constant problem. It was evenargued that evidence of higher dimensionality at cosmo-logical scales is already present in the current observatio-nal data [5].Combining the essence of both extremes, a framework was recently proposed in which the structure of spacetimeis a fundamentally (

1

þ

1

)-dimensional universe but is‘‘wrapped up’’ in such a way that it appears higher-dimensional at larger distances [6]. The structure of spacemay be envisioned as an

n

-dimensional ordered lattice onwhich dynamics are conﬁned to (presumably)

n

¼

1

–

4

,deﬁned by fundamental scales

L

1

<L

2

<L

3

<L

4

.Physics with

Ã

<

Ã

3

on length scales

L>L

3

¼

Ã

À

13

will appear three-dimensional. When the energy (length)scale becomes of the order of

Ã

2

>

Ã

3

(

L

2

<L

3

),the manifold transitions from

ð

3

þ

1

Þ

to

ð

2

þ

1

Þ

. If

Ã

2

$

1 TeV

, planar events and other interesting effectscould be observed at the LHC for collisions with

ﬃﬃﬃ

s

p !

Ã

2

[7], in addition to unique signatures of lower-dimensional quantum black hole production [8]. Forrandom orientation of lower-dimensional planes or lines,violations of Lorentz invariance induced by a lattice be-come nonsystematic and thus evade strong limits put ontheorieswithsystematicviolationofLorentzinvariance[9].Beyond this novelty, however, the framework effectivelycures all divergences that plague the (

3

þ

1

)-dimensionalaspects of the current ﬁeld theory. For example, the ﬁne-tuning problem is alleviated. The radiative corrections tothe Higgs boson mass in

d

spacetime dimensions areobtained for some cutoff energy

Ã

from the top,

W

, andHiggs self-coupling loop diagram contributions:

Á

m

2

H

$

X

i

Z

Ã

d

d

k

ð

2

Þ

d

1

k

2

À

m

2

i

¼

F

d

ð

Ã

Þ

;

(1)where the index

i

refers to the diagram and the function

F

d

ð

Ã

Þ

denotes the total divergence behavior after fullevaluation of the contributing Feynman integrals. Whilequadratically divergent for

d

¼

4

[i.e.,

F

4

ð

Ã

Þ$

Ã

2

], theone-loop corrections to the Higgs boson mass in (

2

þ

1

)dimensions are linearly divergent in the cutoff scale, whilein the (

1

þ

1

)-dimensional case they are only logarithmi-cally divergent.Furthermore, the problems plaguing (

3

þ

1

)-dime-nsional quantum gravity quantization programs are solvedby virtue of the fact that spacetime is dimensionally re-duced. Indeed, effective models of quantum gravity areplentiful in (

2

þ

1

) and even (

1

þ

1

) dimensions [10–12].
Similarly, the cosmological constant problem may be ex-plained as a Casimir-type energy between two adjacent‘‘foliations’’ of three-dimensional space as the scale size

L>L

4

opens up a fourth space dimension.What makes this proposal of evolving dimensions veryattractive is that some evidence of the lower-dimensionalstructureof our spacetime ata TeV scale may already exist.Namely, alignment of the main energy ﬂuxes in a target(transverse) plane has been observed in families of cosmicray particles [13–15]. The fraction of events with align-
ment is statistically signiﬁcant for families with energieshigher than TeV and a large number of hadrons. This canbe interpreted as evidence for coplanar scattering of PRL

106,

101101 (2011)PHYSICAL REVIEW LETTERS

week ending11 MARCH 2011

0031-9007

=

11

=

106(10)

=

101101(4) 101101-1

Ó

2011 American Physical Society

secondary hadrons produced in the early stages of theatmospheric cascade development.An interesting side effect of such a dimensional reduc-tion scheme is the distinct nature of gravity in lowerdimensions. It is well known that, in a (

2

þ

1

)-dimensionaluniverse, there are no local gravitational degrees of free-dom, and hence there are no gravitational waves (or grav-itons). If the Universe was indeed three-dimensional atsome earlier epoch, it is reasonable to deduce that noprimordial gravitational waves (PGWs) of this era existtoday. There is thus a maximum frequency for PGWs,implicitly related to the dimensional transition scale

Ã

2

,beyond which no waves can exist. This indicates thatgravitational wave astronomy can be used as a tool forprobing the novel ‘‘vanishing dimensions’’ framework.We note that the idea of using PGWs and their frequencyspectrum to determine dimensional characteristics of spacetime is not new. It has been shown, for example,that phase shifts can be introduced from PGW interactionswith extra dimensions [16]. Similarly, Ref. [17] demon-
strates the thermalization of PGWs via propagationthrough extra dimensions. Alternatively, PGWs can revealthe existence of topological Chern-Simons terms in themodiﬁed Einstein-Hilbert action [18].In order to determinean approximatevalue for the cutoff frequency, we revisit the current state of PGW detection.The standard cosmological theory predicts that gravita-tional waves will be generated in the pre- or posti-nﬂationary regime due to quantum ﬂuctuations of thespacetime manifold. A standard Friedmann-Robertson-Walker cosmology is assumed, with the usual radiation-and matter-dominatederas. Gravitywavescanbe producedat different times

t

Ã

<t

0

¼

H

À

1

, when the temperature of the Universe was

T

Ã

. The comoving entropy per volume of the Universe at temperature

T

Ã

can be expressed as afunction of the scale factor

a

ð

t

Þ

as

S

$

g

S

ð

T

Þ

a

3

ð

t

Þ

T

3

;

(2)where the factor

g

S

represents the effective number of degrees of freedom at temperature

T

in terms of entropy,

g

S

ð

T

Þ¼

X

i

¼

bosons

g

i

T

i

T

3

þ

78

X

j

¼

fermions

g

j

T

j

T

3

:

(3)The parameters

i

and

j

run over all particle species. Inthe standard model, this assumes a constant value for

T

’

300 GeV

, with

g

S

ð

T

Þ¼

106

:

75

due to the fact thatall species were thermalized to a common temperature.Assuming that entropy is generally conserved over theevolution of the Universe, one can write

g

S

ð

T

Ã

Þ

a

3

ð

t

Ã

Þ

T

3

Ã

¼

g

S

ð

T

0

Þ

a

3

ð

t

0

Þ

T

30

(4)with

T

0

¼

2

:

728 K

.The characteristic frequency of a gravitational waveproduced at some time

t

Ã

in the past is thus redshifted toits present-day value

f

0

¼

f

Ã

a

ð

t

Ã

Þ

a

ð

t

0

Þ

f

0

¼’

9

:

37

Â

10

À

5

Hz

ð

H

Â

1 mm

Þ

1

=

2

g

À

1

=

12

Ã

ð

g

Ã

=g

Ã

S

Þ

1

=

3

T

2

:

728

;

(5)where the original production frequency

f

Ã

is boundedby the horizon size of the Universe at time

t

Ã

, i.e.,

f

Ã

$

À

1

Ã

$

H

À

1

Ã

. Note that this is an upper bound, andthe actual value may be smaller by a factor

Ã

$

H

À

1

Ã

,although the ﬁnal result is weakly sensitive to the value

1

T

Ã

by noting that, during the radiation-dominated phase,the scale is

H

Ã

¼

8

3

g

Ã

T

4

Ã

90

M

2Pl

:

f

Ã

¼

7

:

655

Â

10

À

5

ð

g

Ã

Þ

1

=

6

T

Ã

TeV

Hz

%

1

:

67

Â

10

À

4

T

Ã

TeV

Hz

;

(7)where the latter equality holds for

g

Ã

$

10

2

. When

T

Ã

¼

1 TeV

, the frequency is

f

Ã

$

10

À

4

Hz

. This iswell below the seismic limit of

f

$

40 Hz

on ground-based gravity wave interferometer experiments likeLIGO or VIRGO [20] but sits precisely at the thresholdof the Laser Interferometer Space Antenna (LISA)’s sen-sitivity range. Indeed, the latter observatory is expected toprobe a variety of early-Universe phenomenology from the100 GeV–1000 TeV period [21]. Figure1demonstrates the
threshold PGW frequency as a function of transitionenergy

Ã

2

¼

T

2D

.At this point, it is instructive to study the physics of expanding (

1

þ

1

) and (

2

þ

1

)-dimensional universes inorder to check if some unexpected dynamical features can

FIG. 1. Frequency threshold for primordial gravitationalwaves produced when the Universe was at temperature

T

Ã

.The hatched region is outside the sensitivity cutoff of LISA.

PRL

106,

101101 (2011)PHYSICAL REVIEW LETTERS

week ending11 MARCH 2011

101101-2

change the conclusions derived so far. We will show belowthat this does not happen.In any spacetime, the curvature tensor

R

may bedecomposed into a Ricci scalar

R

, Ricci tensor

R

, andconformally invariant Weyl tensor

C

. In three dimen-sions the Weyl tensor vanishes, and

R

can be ex-pressed solely through

R

and

R

. Explicitly,

R

¼

G

:

(8)This in turn implies that any solution of the vacuumEinstein’s equations is locally ﬂat. Thus, (

2

þ

1

)-dime-nsional spacetime has no local gravitational degrees of freedom, i.e., no gravitational waves in the classical theoryand no gravitons in the quantum theory. Gravity is thenuniquely determined by a local distribution of matter.The number of degrees of freedom in such a theory isﬁnite, the quantum ﬁeld theory reduces to quantummechanics, and the problem of nonrenormalizabilitydisappears. A (

2

þ

1

)-dimensional Friedmann-Robertson-Walker metric is

ds

2

¼

dt

2

À

a

ð

t

Þ

2

dr

2

1

À

kr

2

þ

r

2

d

2

;

(9)where

a

ð

t

Þ

is the scale factor and

k

¼À

1

;

0

;

þ

1

. Einstein’sequations for this metric are

_

aa

2

¼

2

G

À

ka

2

;ddt

ð

a

2

Þþ

pddta

2

¼

0

;

(10)where

G

is the (

2

þ

1

)-dimensional gravitational constant,

p

is thepressure,and

is theenergydensity.Inaradiation-dominated universe,

p

¼

12

and

a

3

¼

0

a

30

¼

const

,which gives

_

a

¼Æ

ﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃﬃ

2

G

0

a

30

a

À

k

s

;

€

a

¼

G

0

a

30

a

2

:

(11)For

k

¼

0

the solution to these equations is

a

ð

t

Þ¼

92

G

0

a

30

1

=

3

t

2

=

3

:

(12)One can note that three-dimensional solution

a

ð

t

Þ/

t

2

=

3

is different from the usual four-dimensional behavior

a

ð

t

Þ/

t

1

=

2

in the radiation-dominated era.We have to note here that Einstein’s equations, i.e.,

G

¼

T

, are not the only valid option in (

2

þ

1

)-dimensional space. For example, it has been known thattheories with

R

¼

T

, where

R

is the Ricci scalar and

T

isthe trace of

T

, are not good (

3

þ

1

)-dimensional theoriesof gravity since they do not have a good Newtonian limit.However, in the context of evolving dimensions, a goodNewtonian limit is not a requirement since the spacetimebecomes (

3

þ

1

)-dimensional at distances larger than

TeV

À

1

. The solutions of the

R

¼

T

theory were dis-cussed, for example, in Ref. [22]. The solution for aradiation-dominated universe for the metric given in (9)is

a

ð

t

Þ¼

t

.The crossover from a (

2

þ

1

)- to a (

3

þ

1

)-dimensionaluniverse happened when the temperature of the universewas

T

2D

3D

¼

Ã

2

$

1 TeV

. Working backwards, wecan estimate the size of the Universe at the transitionfrom the ratio of scale sizes at various epochs, speciﬁcallybetween the present day (

t

today

$

10

17

s

), the radiation- ormatter-dominated era (

t

RM

$

10

10

s

), and the TeV era(

t

TeV

$

10

À

12

s

). The scale factor at the latter epoch is thus

a

TeV

¼

a

today

t

TeV

t

RM

1

=

2

t

RM

t

today

2

=

3

¼

10

À

15

a

today

:

(13)This value may also be obtained by noting that conser-vation of entropy requires the product

aT

to be constant,and so

a

TeV

¼

10

À

15

a

today

(since

T

today

$

10

À

3

eV

10

28

cm

) at

T

$

1 TeV

was

10

13

cm

. This dis-tance is macroscopic, but it is not in contrast with ourassumption that the crossover from a (

2

þ

1

)- to a(

3

þ

1

)-dimensional universe happened when the tempera-ture of the Universe was

T

$

1 TeV

, since the causallyconnected Universe today contains many causally con-nected regions of some earlier time. Finding the exactsize of the causally connected Universe at the dimensionalcrossover is not a unique task, since it would stronglydepend on an underlying cosmological model. In particu-lar, it would depend on which scale inﬂation and reheatinghappened [23] (if at all). The absolute lower limit in theenergy scale of inﬂation is about 10 MeV (in order not toaffect the earliest landmark of the standard cosmology—nucleosynthesis), but inﬂation may as well happen at anyenergy above the dimensional crossover scale. Further, thedimensional crossover may perhaps be a violent highlynonadiabatic process with huge entropy production. Inthat case the standard relation

aT

¼

constant would notbe valid anymore at temperatures above TeV [but itwould still be valid from

T

$

TeV

until today, as assumedin Eq. (13)].Fortunately, our limits on PGW are robust with respectto the underlying cosmological model. The only explicitinput we used was that the dimensional crossover scale is

T

$

1 TeV

, which is the value strongly favored for theo-retical reasons [6] and perhaps also indicated by the planarevents in cosmic ray experiments [13–15].
Going towards even higher temperatures, the spacetimebecomes (

1

þ

1

)-dimensional. To avoid large hierarchy inthe standard model, the crossover from an (

1

þ

1

)-dimensional to a (

2

þ

1

)-dimensional universe needs tohappen when the temperature of the universe was

T

1D

2D

¼

Ã

1

100 TeV

. Conservation of entropy (if be-tween

T

$

1 TeV

and

T

$

100 TeV

nothing nonadiabatichappened) requires

aT

¼

const

. This implies

a

2D

3D

a

1D

2D

¼

T

1D

2D

T

2D

3D

$

100

:

(14)Similarly,

a

1D

2D

a

0

¼

T

1D

2D

T

0

;

(15)PRL

106,

101101 (2011)PHYSICAL REVIEW LETTERS

week ending11 MARCH 2011

101101-3