Supersymmetry in a classical world: new insights …aws/_Garching_2015_07_17_2.pdfSupersymmetry in a...
Transcript of Supersymmetry in a classical world: new insights …aws/_Garching_2015_07_17_2.pdfSupersymmetry in a...
Supersymmetry in a classical world: new insights on stochastic dynamics from
topological field theory
1
Igor Ovchinnikov
Excellence Cluster Universe, Garching2015/7/17
University of California at Los Angeles
• Ubiquitous Chaotic Long-Range Order and State-of-Art Theory of Stochastic Differential Equations (SDEs) and Dynamical Systems (DS) Theory
• The Cohomological Theory of SDEs (ChT-SDE): Extended Hilbert Space; Topological Supersymmetry and Its Spontaneous Breaking (Chaos); Ergodicity and Time Reversal Symmetry Breaking; Phase Diagram and Noise-Induced Chaos.
• Example: Healthy Brain is at the Phase of Noise-Induced Chaos• Conclusion
Plan of the Talk
2Excellence Cluster Universe, Garching2015/7/17
3
Power-law (or scale free) statistics (e.g, Richter Scale) of highly nonlinear processes or events such as earthquakes
1/f, pink, or flicker noise (long-term memory effect)- algebraic power-spectra
Butterfly Effect, i.e., infinite memory of perturbations/init.cond.
Mysterious Chaotic Long-Range Order
Excellence Cluster Universe, Garching2015/7/17
4
Darwin Theory of Evolution (1859)
Punctuated Equilibria(Eldredge, Gauld, 1972)
Biological evolution
Mysterious Chaotic Long-Range Order
Excellence Cluster Universe, Garching2015/7/17
5
Econodynamics
Biological evolution
Mysterious Chaotic Long-Range Order
Excellence Cluster Universe, Garching2015/7/17
6
Neurophysics Econodynamics
Biological evolution
Mysterious Chaotic Long-Range Order
Excellence Cluster Universe, Garching2015/7/17
Geophysics
Neurophysics Econodynamics
Magneto-hydrodynamics
Aerodynamics
Traffic
Biological evolution
Soft condensed matter
Hydrodynamics
Flocking
Internet
7
Mysterious Chaotic Long-Range Order
Excellence Cluster Universe, Garching2015/7/17
Geophysics
Astrophysics
Neurophysics Econodynamics
Magneto-hydrodynamics
Aerodynamics
Traffic
Biological evolution
Soft condensed matter
Hydrodynamics
Flocking
Internet
8
Mysterious Chaotic Long-Range Order
Excellence Cluster Universe, Garching2015/7/17
9
Astrophysics:… a wide range of phenomena in astrophysics, such as planetary magnetospheres, solar flares, cataclysmic variable stars, accretion disks, black holes and gamma-ray bursts, and also to phenomena in galactic physics and cosmology…
Mysterious Chaotic Long-Range Order: Astrophysics
Excellence Cluster Universe, Garching2015/7/17
Hamilton dynamics
Least Action Principle,
probabilisticdeterministic
Quantum Theory
General Form O(partial)DEs
10Excellence Cluster Universe, Garching2015/7/17
Euler Equations
- Hard Condensed Matter- Atomic and Molecular Physics- Elementary Particle Physics
Low-Energy Effective Equations
of Motion
General Form S(partial)DEs
Quantization
- Sort Condensed Matter- Hydro-,Aero- … dynamics- Everything “non-quantum”,
i.e., above the scale of quantum degeneracy/coherence
- All Sciences other than Physics
Generality of S(partial)DEs in Physics
))(()( txFtx
NoisetxFtx ))(()(
0)(/ txS
minS
/ˆ iSeU
Hamilton dynamics
Least Action Principle,
probabilisticdeterministic
Quantum Theory
General Form O(partial)DEs
11Excellence Cluster Universe, Garching2015/7/17
Euler Equations
- Hard Condensed Matter- Atomic and Molecular Physics- Elementary Particle Physics
Low-Energy Effective Equations
of Motion
General Form S(partial)DEs
Quantization
- Sort Condensed Matter- Hydro-,Aero- … dynamics- Everything “non-quantum”,
i.e., above the scale of quantum degeneracy/coherence
- All Sciences other than Physics
Generality of S(partial)DEs in Physics
Domain of observation of CLRO
))(()( txFtx
NoisetxFtx ))(()(
0)(/ txS
minS
/ˆ iSeU
Theory of SDEs is older than the quantum theory and general relativity.Theories of Brownian motion: Smoluchowskii(1906), Einstein (1905), even earlier works
12Excellence Cluster Universe, Garching2015/7/17
State-of-Art theory of S(partial)DEs
Complexity
13
Modern Picture: Open Questions
Excellence Cluster Universe, Garching2015/7/17
Turbulence
Non-equilibrium
dynamics
Chaotic
dynamics
ErgodicThermodynamic
equilibrium
Ergodic theory of
chaos
?
Theory of SDEs is older than the quantum theory and general relativity.Theories of Brownian motion: Smoluchowskii(1906), Einstein (1905), even earlier works
14
Accidental pictureCLRO is a “critical” phenomenon - some excitation has zero gap because the DS is at a “phase transition”Contradiction with ubiquity of CLROSelf-Organized Criticality: postulation of existence of mysterious tendency of self-fine-tuning into the phase transition into ordinary chaos
Symmetry breaking picture CLRO is a symmetry breaking phenomenon, and CLRO is a result of the Goldstone theorem. Requirement from ubiquity of CLROAll stochastic systems must possess such a symmetry and it must be a supersymmetry
Excellence Cluster Universe, Garching2015/7/17
Chaotic Long-Range Order
Chaotic Long-Range Order: Potential Origin
)())(()2())(()( 2/1 ttxetxFtx a
a
DS’s variables from
some topological
manifold, X, called
phase space
15Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Stochastic Evolution
)())(()2())(()( 2/1 ttxetxFtx a
a
Temperature
A set of vector fields,
vielbeins, If (in)dependent
on x(t), it is said that noise
is (additive)multiplicative
Flow vector field
DS’s variables from
some topological
manifold, X, called
phase space
16Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Stochastic Evolution
)())(()2())(()( 2/1 ttxetxFtx a
a
TemperatureGuassian white
Noise variables
A set of vector fields,
vielbeins, If (in)dependent
on x(t), it is said that noise
is (additive)multiplicative
Flow vector field
DS’s variables from
some topological
manifold, X, called
phase space
Possible generalizations: any noise; partial differential equations; flow
vector field and vielbeins with explicit time/space dependence and
“integral” or temporary (and spatially) nonlocal dependence on x…
17Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Stochastic Evolution
( ')x t ' 2( ) ( ( '), )ttx t M x t
' 3( ) ( ( '), )ttx t M x t
' 1( ) ( ( '), )ttx t M x t
18Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Stochastic Evolution
)())(()2())(()( 2/1 ttxetxFtx a
a
19Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Stochastic Evolution
)())(()2())(()( 2/1 ttxetxFtx a
a
))'(,()( ),()'(:)(
maps offamily - estrajectoridependent Noise
'' txMtxtxtxM tttt
( ')x t ' 2( ) ( ( '), )ttx t M x t
' 3( ) ( ( '), )ttx t M x t
' 1( ) ( ( '), )ttx t M x t
20Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Stochastic Evolution
)())(()2())(()( 2/1 ttxetxFtx a
a
))'(,()( ),()'(:)(
maps offamily - estrajectoridependent Noise
'' txMtxtxtxM tttt
( ')t
t'moment Timet't moment Time
( , )t
Temporal evolution
( ')t 2( , )t
3( , )t
1( , )t
- Hilbert space is the exterior algebra
- Consideration of generalized (not only total) probability
distributions is a mathematical necessity
1
1
( ) 1 ( ) ( )...( !) ( ) ... ( )k
k
k k i i ki ik x dx dx X
- GPDs in the coordinate-free setting = differential or k-forms:
(1) 1 2 1 2 1 2( ) ( | ) ( | )p x p x x dx p x x dx
( )
0
( ) ( )D
k
k
X X
A k-form is naturally coupled to k-
dimensional submanifolds (k-chains)
Examples: conditional and total probability
distributions
(2) 1 2( ) ( )p x p x dx dx
Meaning: in local coordinates
where k-chain belongs the
hyperplane 1,..,k Dx x consts
is the probability to find
within the chain, given
the other variables are known
( )
( ) ( ) 1
k
k k
cp
1,.., kx x
(2)c
(1)c
21Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Hilbert Space
1
1
( ) 1 ( ) ( )...( !) ( ) ... ( )k
k
k k i i ki ik x dx dx X
- GPDs in the coordinate-free setting = differential or k-forms:
22Excellence Cluster Universe, Garching2015/7/17
Standard Conditional Probability Distribution
11 1arg
arg arg
1
1
1 1
1
( ... ) ( .
Example: standard definition of Cond.Prob.Density on
In coordinate-free
.. | ) ( )
setti
... .
ng
..
... .( ) .
( . . |
.
..
k D kD ktot cond m
D km m
kc
D
k D k D
ond cond
D
k
P x x P x x P
P P
P P
x x x x
x x
x
x x
xx
d d
1
1 1arg
) ...
( ... ) .
.
.
.
.
k k
D Dtot cond m tot
D
D
dx dx
P
x
dxP P P x x dx
Quantum theory
Pr
( ) ( ) ( )Total
obabilityFunction
x x TPF x
ChT-SDE
Unthermalized, Thermalized, All vars.unstable vars. stable vars.
1 1 1( ) ... ( ) ... ( ) ...k k D Dx dx dx x dx dx TPF x dx dx
Bra-ket “factorization” of total probability density
23Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Hilbert Space
24Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Stochastic Evolution
)())(()2())(()( 2/1 ttxetxFtx a
a
))'(,()( ),()'(:)(
maps offamily - estrajectoridependent Noise
'' txMtxtxtxM tttt
( ')t
t'moment Timet't moment Time
( , )t
Temporal evolution
( ')t 2( , )t
3( , )t
1( , )t
25Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Stochastic Evolution
)())(()2())(()( 2/1 ttxetxFtx a
a
))'(,()( ),()'(:)(
maps offamily - estrajectoridependent Noise
'' txMtxtxtxM tttt
1
1
( )
...( ') ( ( ')) ( ') ... ( ')k
k
iik
i it x t dx t dx t
t'moment Timet't moment Time
)(...)()))'(((),( 1
1
)(
...
*
' tdxtdxtxMt k
k
iik
iitt
))(,()'( ' txMtx tt )by Pullback (M t't
Temporal evolution
Variable substitution
( ')t 2( , )t
3( , )t
1( , )t
26Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Stochastic Evolution
)())(()2())(()( 2/1 ttxetxFtx a
a
))'(,()( ),()'(:)(
maps offamily - estrajectoridependent Noise
'' txMtxtxtxM tttt
)by inducedpullback )(
)(ˆ ),'(ˆ)(
is),( GPDs,for evolution Stochastic
*
'
*
'''
(MM
MMtMt
XΩψ
t'ttt
Noisetttttt
( ')t 2( , )t
3( , )t
1( , )t
27Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Stochastic Evolution
)())(()2())(()( 2/1 ttxetxFtx a
a
))'(,()( ),()'(:)(
maps offamily - estrajectoridependent Noise
'' txMtxtxtxM tttt
)by inducedpullback )(
)(ˆ ),'(ˆ)(
is),( GPDs,for evolution Stochastic
*
'
*
'''
(MM
MMtMt
XΩψ
t'ttt
Noisetttttt
( ')t ( )t
28Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Stochastic Evolution
)())(()2())(()( 2/1 ttxetxFtx a
a
))'(,()( ),()'(:)(
maps offamily - estrajectoridependent Noise
'' txMtxtxtxM tttt
)by inducedpullback )(
)(ˆ ),'(ˆ)(
is),( GPDs,for evolution Stochastic
*
'
*
'''
(MM
MMtMt
XΩψ
t'ttt
Noisetttttt
H
LLLHeMaa eeF
Htt
tt
ˆ - mally,infiniteseor
ˆˆˆˆ ,ˆ
t
ˆ)'(
'
( ')t ( )t
)())(()2())(()( 2/1 ttxetxFtx a
a
H
LLLHeMaa eeF
Htt
tt
ˆ - mally,infiniteseor
ˆˆˆˆ ,ˆ
t
ˆ)'(
'
2ijˆ ˆ g
noise induced metr
Noise indiced diffusion
.
c
..
i
a ae e i j
ij i ja a
L Lx x
g e e
"Average" flow, Lie or
physical derivative along F
29Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Stochastic Evolution
))'(,()( ),()'(:)(
maps offamily - estrajectoridependent Noise
'' txMtxtxtxM tttt
)by inducedpullback )(
)(ˆ ),'(ˆ)(
is),( GPDs,for evolution Stochastic
*
'
*
'''
(MM
MMtMt
XΩψ
t'ttt
Noisetttttt
( ')t ( )t
No Approximations!
30Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Stochastic Evolution
)())(()2())(()( 2/1 ttxetxFtx a
a
))'(,()( ),()'(:)(
maps offamily - estrajectoridependent Noise
'' txMtxtxtxM tttt
)by inducedpullback )(
)(ˆ ),'(ˆ)(
is),( GPDs,for evolution Stochastic
*
'
*
'''
(MM
MMtMt
XΩψ
t'ttt
Noisetttttt
H
LLLHeMaa eeF
Htt
tt
ˆ - mally,infiniteseor
ˆˆˆˆ ,ˆ
t
ˆ)'(
'
( ) ( ) ( )
Stratonovich Interp. Weyl symmetrization
D i i j D
t i i a j aP x d x F e e P x d x
( ')t ( )t
31Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Supersymmetry
t
Stochastic evolution on exterior algebra
The Fokker-Planck operator can be given explicitely supersymmetric form
ˆ ˆ ˆ ˆ ˆ, .
ˆˆˆ
ˆ ˆ
where ˆ ˆ ˆ,
a a
a a
F e e
F e e
H H L L L
H [d,d],
d i Θi L
interior multiplication, and the use of
Cartan formula, has been made with being exterior derivative.
Also,
ˆ ˆˆ ˆ
ˆ ˆ
is a sup ersymmetry of the modelˆ
0,
F
F F
i
L [d,i ] d
[d,H]
d
,
1 1
1 1
1 1
( ) ( )... ...
( ) ( )
Exterior differentive:
is very fundamental to albegraic topology
-) It is the matter of Stokes' theorem,
ˆ ( ) ..
,
. ( )
...
ˆ
k k
k k
k k
k i i k j i ii i i ij
k k
c c
d x dx dx x dx dx dxx
d
2
where is the boundary operator.
-) Its cohomology is De Rahm cohomology
is the algebraic representative of "boundary" operator
It is nilpotent, , "boundary of a boundary" is empty
ˆ
ˆ 0
d
d
*'
commutes with any pullback , and thus with the evolution o
Interpretation: continuous-time dynamics conserv
ˆ ˆ ˆ ˆ[ ,
es the "concept of boundary
( )] 0 [ , ]pe to 0r
"
ra ttd d M d M
*'( )ttM
d̂
d̂A Poincare dual of a chain
*'( )ttM
32Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Meaning of Supersymmetry Operator
1 2 2 1
1 2 2 1
1
1
( )...
Properties of wedge product
are those for anticommuting or fermionic fields
Differential forms can be given as functions
of bosonic and fermionic field
.
s
k
i i i i
i i i i
k ii i
dx dx dx dx
dx
1
1
1
( )...
In these terms, exterior derivative has the form
.. ...
ˆ
k k
k
i k i ii i
i
i
dx
dx
33Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Meaning of Supersymmetry Operator
34Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Supersymmetry
t
Stochastic evolution on exterior algebra
The Fokker-Planck operator can be given explicitely supersymmetric form
ˆ ˆ ˆ ˆ ˆ, .
ˆˆˆ
ˆ ˆ
where ˆ ˆ ˆ,
a a
a a
F e e
F e e
H H L L L
H [d,d],
d i Θi L
interior multiplication, and the use of
Cartan formula, has been made with being exterior derivative.
Also,
ˆ ˆˆ ˆ
ˆ ˆ
is a sup ersymmetry of the modelˆ
0,
F
F F
i
L [d,i ] d
[d,H]
d
,
t
Stochastic evolution on exterior algebra
The Fokker-Planck operator can be given explicitely supersymmetric form
ˆ ˆ ˆ ˆ ˆ, .
ˆˆˆ
ˆ ˆ
where ˆ ˆ ˆ,
a a
a a
F e e
F e e
H H L L L
H [d,d],
d i Θi L
interior multiplication, and the use of
Cartan formula, has been made with being exterior derivative.
Also,
ˆ ˆˆ ˆ
ˆ ˆ
is a sup ersymmetry of the modelˆ
0,
F
F F
i
L [d,i ] d
[d,H]
d
,
n n
n n
Eigenstates of are either
-) supersymmetric singlets: but
All have zero eigenvalues !
-) or non-supersymmetric double
ˆ
ˆ ˆ0 somethin
ts: and
g
ˆ 0
H
d d
d
35Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Supersymmetry
- Eigenvalues are either real or complex conjugate pairs (Ruelle-Pollicott resonances of DS theory)
- Eigensystem
FP operator is real and thus p
is complete, bi-orthogonal:
suedo-Hermitian
ˆ ˆ,n nH n E n n H E
- Real parts of eigenvalues are bounded from be
, 1,
For physical models with positive definite noise-metric
From
low
- All eigenstates are either supersymmetr
sup
ic
ersymm
singlets or non-
etry
s
knnn n n k n
upersymmetric doublets
- All non-zero eigenvalues correspond to non-supersymmetric pairs
- There always exist a supersymmetric state of the steady-state total probability distribution,
. ., the state ofi e "thermodynamic equili brium".
nIm E
nRe E
nIm E
nRe E
nIm E
nRe E
d- and PT reversal
symmetries broken
d-symmetry broken
36Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Possible FP Spectra
Quantum theory
ˆ( )t iH
nIm E
nRe E
nIm E
nRe E
nIm E
nRe E
37Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Possible FP Spectra
Quantum theory
ˆ( )t iH
Pathintegralrepresentation
Contribution only from
Physical meaning Value
Partition Function Ground states (in large time limit)
Stochastic number of periodic solutions (for some models)
For spectrum bChaotic behavior
Witten Index Supersymmetric states only
Partition function ofnoise (up to a topological factor)
Euler characteristic of
X
,Q
APBCD e
ˆ tHZ Tre
ˆ ˆ
W ( 1)F tHTr e
,Q
PBCD e
| |
2 gt EtZ e
Both have physical meaning only if the entire exterior algebra is the Hilbert space !
d- and PT reversal
symmetries broken
d-symmetry broken
Exp. decayExp. growth
38Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Ground States and Ergodicity
nIm E
nRe E
nIm E
nRe E
nIm E
nRe E
d- and PT reversal
symmetries broken
d-symmetry broken Quantum theory
ˆ( )t iH
1 1 2 1ˆ ˆ ˆ ˆ1 ( ) ( ) ( ) ( )
1 1 1
11
ˆ ˆ ˆ ˆ( ).
Spectra a and b: correlators in the long t
.. ( ) ...
ime limit reduce to those of the ground states nly
ˆ
o
k k kt t H t t H t t H t t Hk k t t k
tg g
O t O t Z Tre O e e O e
N g O
1 1
1
ˆ ˆ( ) ( )1 1 1
Interpretation: ergodi
where in Heisenberg picture
For spectra c: one can use standard trick of
city property
quantum t
ˆ( )... ( ) ,
ˆ ˆ(
heory
(a little Wick rotation o
)
) t
k k
t t H t t H
t O t g
O t e O e
make theory "ergodic"
In order to study response of the system, introduce probing fields
T
( ( )) ( ( )) ( ) ( ( ))
ˆˆ ˆ ˆ ˆ ˆ
he evolution operator tran
( ) ( ) ( )
sforms as
b
i i b ib
b bf
F x t F x t t f x t
H H t H t L H t d
T
T
0
1 1
Chron.Ordering
ground
ˆ- ( )
1 1
statesI
1
1 0n
The response can be charactized by response correlators ( )
ˆ, ,
( ) ˆ ˆˆ ˆ, ( ) ... , ( )( ).... ( )
b
t
b bk k
f
H d
t
tt g f f kb bt
gk
Z Tr e
ZLim Z N g d t d t
t t
T
1
Heisenberg picture
ground state
11
s
ˆ ˆˆ ˆ ˆ ˆ[ , ] (, where )... , ( )b bk
g f f kg
g
N g d A g A t d t
All vanish by the definition of supersymmetric
(ground) states
Interpretation: forgets perturbations/initial
conditions in the long-time limit
Some do not vanish because the ground states are not
supersymmetric
Interpretation: remembers perturbations/initial conditions
even in the infinitely long time limit – the butterfly effect !
External perturbation
39Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Butterfly Effect
1
0For perturbation of the (for (m ) )b bt t t
1 0 the infinitesemal shift of the tr( ( )) ajectoryf x t
Perturbed trajectory
Unperturbed trajectory
40Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Expectation Values and Statistics
ˆ ˆ1 ( ) 1
0
Expectation value of an operator, , at moment
For wide class of operators including
where the "ergodic" probability density
ˆ ,
ˆ ˆ ˆ( ) lim
ˆ=
ˆ ( ) ( )
(
T t H tHT gn gT
gX
g
O t
O t Z n e Oe n N g O g
O O
O O x P x
P 1 1
Statistics works no matter if the supersymme
) ( ) ( )
try is broken or
.
n
.
ot
. Dg g gg
x N x x dx dx
U(x),P(x) U(x),P(x)-?
- Forgets initial conditions
- Thermalizes to a steady-state total probability distribution (the ground state)
- Never forgets initial conditions because the variable is not stable
- The steady-state total probability distribution is meaningless . The ground state is
Langevin 1D dynamics
41Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Deterministic Ground States
( )g P x dx1g 1g ( )g P x dx
( ) 1xg g P dx ( ) 1g P x dxg
The ground state is not a distribution in unstable variables
42Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Deterministic Ground States
homoclinic tangle
Ket – Poincare dual of global unstable manifoldsBra- Poincare dual of global stable manifolds
g
g
Saddle
chain ofboundary of dual Poincarechain of dual Poincareˆ d
Integrable (non-chaotic) flows: the supersymmetric ground states are Poincare duals of global unstable manifolds
Non-integrable (chaotic) flows: the non-supersymmetric ground states are Poincare duals of global unstable manifolds modified by a functional dependence on position
Langevin ODE on 2D torus
Langevin function is the height in the “3rd“ direction
1 , circle, 1-form
1 , circle, 1-form
0 , point, 2-form
2 ,circle, 1-form
3 entire space,
constant function
0 , entire space,
constant function
2 , cirlce, 1 form
3 , point, 2 form
0 1
1 2
( 1) 3 3 ( 1) 2 2
( 1) 1 1 ( 1) 0 (Euler characteristic of t0 )0 orus
W
43Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Deterministic Ground States
Lorenz model
Unstable manifold
(picture by Hinke Osinga)
Type I
Params.
TE
C
Physical Laplacians do not break supersymmetry.
With the increase of noise temperature, the susy will eventually get restored – strong enough noise destroys chaotic long-range order
44Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Stochastic Phase Diagram
, Temp.
High temp. ˆ ˆ ˆ ˆ ˆ: ˆ a a a aF e e e eH L L L L L
TE - Thermodynamic equilibrium
C - "Ordinary" chaos
Type II
Params.
TE
CN N
• Weak noise introduces exponentially weak overlap or noise-induced tunneling processes (instantons) between “local” ground states on different attractors.
• When tunneling processes break susy, Goldstinos representing parameters (modulii) are gapless. Result: power-law or scale free statistics.
• Mostly “regular” behavior along attractors interrupted by sudden unpredictable processes with scale-free statistics. Exists on the border of “ordinary” chaos. Typical description of Self-Organized Criticality!
• At higher temperatures, N-C is a crossover because external observer can not tell between different tunneling events
45Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Stochastic Phase Diagram
, Temp.
Noise-induced Chaos
TE - Thermodynamic equilibrium
C - "Ordinary" chaos
N - Noise-Induced Chaos
Experimental evidence: neuroavalanches exhibit power-law statistics. This suggests brain has its susy broken by these processes – the N-phase
46Excellence Cluster Universe, Garching2015/7/17
Example: Neurodynamics
A. Levina, J.M. Herrmann, and T. Geisel (2006).
Dynamical Synapses Give Rise to a Power-Law
Distribution of Neuronal Avalanches
Numerical Evidence:Neuroavalanches exhibit power-law for a range of parameters.
Recovery of consciousness goes
through a set of phase transitions
coma
consciousanesthetic
47Excellence Cluster Universe, Garching2015/7/17
unconscious
Example: Neurodynamics
Atkinson and Shiffrin model: short-term memory and long-term memory
Recovery of consciousness goes
through a set of phase transitions
coma
consciousanesthetic
48Excellence Cluster Universe, Garching2015/7/17
unconscious
Scale:~<1minPhysics: fast electrochemical dynamics of light ionsSDEs: well know, e.g., Hudgkin-Huxley model
Scale:~>1min Physics: bio-chemical and mechanical rewiring of synapses
Example: Neurodynamics
Atkinson and Shiffrin model: short-term memory and long-term memory
Recovery of consciousness goes
through a set of phase transitions
coma
consciousanesthetic
49Excellence Cluster Universe, Garching2015/7/17
unconscious
Scale:~<1minPhysics: fast electrochemical dynamics of light ionsSDEs: well know, e.g., Hudgkin-Huxley model
Scale:~>1min Physics: bio-chemical and mechanical rewiring of synapses
Determined by the
same neurodynamical
phenomena
Example: Neurodynamics
coma
consciousanesthetic
50Excellence Cluster Universe, Garching2015/7/17
unconscious
Example: Neurodynamics
The ChT-SDE picture
-) Conscious brain is in the N-phase. One can tell between neuroavalanches-this is weak-noise regime.
-) C-phase must correspond to neurodynamical phenomenon of “seizure” (like in epilepsy) – neurons fire non-
stop, non-integrable flow. Again, because the N-C transition is sharp, the brain is in the weak-noise regime
-) Coma is in the TE-phase where there is no chaotic dynamical memory. Therefore, chaotic dynamical memory
is the short-term neurodynamical memory.
-) Existence of short-term memory is a necessary condition for being conscious. Consciousness is within the N-
phase
Params.
Temp.
TE-phase N-phase C-phase
Seizure
51Excellence Cluster Universe, Garching2015/7/17
Anti-anestheticconsciousunconscious
coma
Example: Neurodynamics
anesthetic
The ChT-SDE picture
52Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Low-Energy Effective Theory
Model Symmetry broken
Order Parameter Goldstone-Nambu Particle
Low-Energy Effective Theory
Ferromagnet O(3) Local Magnetization Spinwave LLG Equation
Crystal Spatial translation
Local Stress Transverse sound
Theory of Elasticity
Supersonductor (Global) U(1) “Wavefunction” of Bose-Condensate of
Cooper pairs
Zero sound GL Theory
… … …
Chaotic DS Top. Susy Unstable variables Goldstinos ?
Example: 2D vortex “dominated” turbulence- Order parameter: (half-of) spatial positions of vortices- Goldtinos are supersymmetric partners of
position of vortices- LEET: could it be Schwartz-type TFT? If yes,
does the concepts of braiding, topological quantum computing etc. apply somehow?
Time
The newly found approximation-free theory of stochastic dynamics
• reveals the mathematical origin of the ubiquitous dynamical/chaotic long-range – the spontaneous breakdown of topological supersymmetry that all natural/stochastic dynamical systems possess. “Chaos” (absence of order) is a misnomer in a certain sense because dynamical chaos is a low-symmetry or “ordered” phase.
• clarifies the concepts of thermodynamic equilibrium and ergodicity
• demystifies the controversial concept of self-organized criticality
• shows that dynamical properties of chaotic DSs can not be described by statistics. Low-energy effective theories of chaotic models, such as turbulent water, are those of gapless goldstinos/fermions – supersymmetric partners of unthermalized variables
• because of its multidisciplinary character and widest applicability, has a potential to bring together specialists in DSs theory, statistics, and topological field theories and/or algebraic topology that will result in cross fertilization of these mathematical disciplines
• has a potential to bring the studies of non-quantum systems to the new level of mathematical rigor and beauty, i.e., the level of supertsymmetric quantum (field) theories
53Excellence Cluster Universe, Garching2015/7/17
Conclusion
54Excellence Cluster Universe, Garching2015/7/17
Supersymmetry in Classical World…
Ergodic Dynamics
Infinitely long dynamics with the global ground state. Periodic B.C.
Transients (weak noise)
Finite-time dynamics starting and ending at different points; composite instanton
Examples: crumbling paper, Barkhausen jumps in ferromagnets, glasses, cascades and chain reactions of various types …
55Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Transient Processes
Examples: turbulent water, brain …
outin
out
Time of
observation
d-symmetry is intrinsically broken on instantons –crackles must (and they do) exhibit power-law statistics
Transients can be though of being in the N-phase only in a sense of equivalent “ergodic” theory.
It is in this sense, that earthquakes are in N-phase
Tectonic
plates
Tectonic
“rings”
Additional Slides
56Excellence Cluster Universe, Garching2015/7/17
probabilisticdeterministic
General Form O(partial)DEs
57Excellence Cluster Universe, Garching2015/7/17
Where to Search?
))(()( txFtx
-) Hamilton/Classical, Conservative dynamics
-) Electric circuitry
-) Magnetodynamics (LLG)
-) Schrodinger equation
-) (Magneto-)Hydro-, Bio-, Econo-, Neuro- Dynamics
Everything in Nature !
Flow vector field: Phase Portrait
System’s variables, points from a topological manifold called the phase space.
58Excellence Cluster Universe, Garching2015/7/17
2ˆ( / 2 )i p m V
Where to Search?
))(()( txFtx
/
/
x H p
p H x
-) Hamilton/Classical, Conservative dynamics
-) Electric circuitry
-) Magnetodynamics (LLG)
-) Schrodinger equation
-) (Magneto-)Hydro-, Bio-, Econo-, Neuro- Dynamics
Everything in Nature !
Flow vector field: Phase Portrait
System’s variables, points from a topological manifold called the phase space.
59Excellence Cluster Universe, Garching2015/7/17
Where to Search?
))(()( txFtx
))(2/ˆ( 2 xVmpi xdtdxVmpiS d
t ))(2/ˆ(( 2
dtxpHxpS ))(( /
/
x H p
p H x
Hamilton dynamics
Least Action Principle,
probabilisticdeterministic
Quantum Theory
General Form O(partial)DEs
60Excellence Cluster Universe, Garching2015/7/17
Euler Equations
- Hard Condensed Matter- Atomic and Molecular Physics- Elementary Particle Physics
Low-Energy Eff. Theories
General Form S(partial)DEs
Quantization
Where to Search?
- Sort Condensed Matter- Hydro-,Aero- … dynamics- Everything “non-quantum”,
i.e., above the scale of quantum degeneracy/coherence
))(()( txFtx
NoisetxFtx ))(()(
0)(/ txS
minS
/ˆ iSeU
Deterministic Chaos is the major discovery within DS theory-) “the three-body problem” by Poincare (1887) -) Numerical rediscovery by Lorenz (1963) and others.
In hydrodynamics, chaos is known as turbulence
Deterministic chaos in ODEs– non-integrability in the sense of DS theory
This definition does not lead to the explanation of the Butterfly Effect
ODE is the domain of the DSs Theory
Vs.
61Excellence Cluster Universe, Garching2015/7/17
Where to Search?
( ) ( , ) ( ) 2 ( )a
t ax t F x t e x t
temperature
SDE = ODE + Noise
trajectoryFlow
Diffusion
Deterministic Probabilistic
SDEODE
Initial point
Initial point
Flow
Average trajectory
noise
A set of vector fields
62Excellence Cluster Universe, Garching2015/7/17
Where to Search?
Chaos
Thermodynamic
equilibrium
All natural DS’s
-) Ground state has unstable or “unthermalized” variables (positive Lyapunov exponents), in which it is not a probability distribution
-) The “butterfly effect”
-) This is an ordered (or low-symmetry) phase –opposed to the semantics of word “chaos”
-) Statistics not applicable (example, replica trick)
-) The stationary total probability distribution is (among) the ground state(s). Note: not the issue of its existence!
-) Thermodynamics, statistics are applicable (Markov chains)
-) Forgets initial conditions/perturbations
63Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Butterfly Effect
Comparison with Supersymmetry in Quantum Theory:
-) Hadron Collider – the primary goal is to find supersymmetry in quantum
theory of elementary particles
ChT-SDE: supersymmetry exists (at least) everywhere else from quantum
theory
-) Supersymmetry (if exists) in quantum world must be spontaneously broken.
Prioblems, however, with theory of susy breaking – susy’s hard to break.
ChT-SDE: Chaos is the spontaneously broken susy.
In particular, all life forms are DS’s with spontaneously broken susy
64Excellence Cluster Universe, Garching2015/7/17
65
Stephen Hawking: “it is in complexity that I think the most important developments of the next millennium will be."
Werner Heisenberg: "When I meet God, I am going to ask him two questions: Why relativity and why turbulence? I really believe he will have an answer for the first."
Richard Feynman: “turbulence -the most important unsolved problem of classical physics"
Importance Emphasized by Classics
Excellence Cluster Universe, Garching2015/7/17
ChT-SDE: Previous Approach
66
' '
( ) ( ) ( ) ( )
( ) (
If , what would be according to the SDE ?
This is how it derived:
No
) ,
( ) ( ) ( )
( )
w
Furth
( ) (1 / 2) ( ) ... ( )
,
:
er
D
D
Ns
i i j Di ij
Ns
i
f t f x P xt d x f t t
x t t x t x
f t t f x x P xt d x
f x x f x x x f x P xt d x
x
i i j
''jF( )= F ( F )F2) ( ( ) ...
( 0, 1 / 2)
ˆ ˆ: ( ) 1 ( ) ... ( )
ˆ
Ito, Stratonovich,
Thus Fokker-Planck Eq
( )
uation is:
i i
D
t
t x x t x t x x
f x tH P xt d x
P H P
F1/ 2( ) ( ( )) (2 ) ( ( )) ( ) ( )at ax t F x t e x t t t
,
,
*,
*,
( ) ( ( )) ( )
( ( ) )
( ) ( )
( ) ( )
Dt t t
Ns
t t t
Dt t t
Ns
Dt t t Ns
f t t f M x P xt d x
M x x
f x M P xt d x
f x M P xt d x
( )t
1 1 n n nt t t
n
1n
1n
time
' '
1a b abn n nnNs t tt
Excellence Cluster Universe, Garching2015/7/17
67
Accidental pictureCLRO is a “critical” phenomenon - some excitation has zero gap because the DS is at a phase transition
Chaos
All stochastic DSs
Symmetry breaking picture: CLRO is a symmetry breaking phenomenon - a gapless excitation is the Goldstone particle.
Self-Organized Criticality (Bak&Tang&Wiesenfeld, 1987): proposition to believe that there is a mysterious force that self-tunes parameters of some SDEs to the phase transition into “chaos”:1) A mystery explain by another mystery. Technically, this is just a
green light for using well-developed RG approach.2) If this is true, where does the CLRO come from in chaotic DSs?3) There was no definition of chaos for stochastic dynamics
Excellence Cluster Universe, Garching2015/7/17
Long-Range Order = Gapless Excitation
Ubiquity of CLRO
ContradictionAll DSs must possess this (super)symmetry
Requirement
Chaotic Long-Range Order: Potential Origin
68
Accidental pictureCLRO is a “critical” phenomenon - some excitation has zero gap because the DS is at a phase transition
Chaos
All stochastic DSs
Symmetry breaking picture: CLRO is a symmetry breaking phenomenon - a gapless excitation is the Goldstone particle.
Self-Organized Criticality (Bak&Tang&Wiesenfeld, 1987): proposition to believe that there is a mysterious force that self-tunes parameters of some SDEs to the phase transition into “chaos”:1) A mystery explain by another mystery. Technically, this is just a
green light for using well-developed RG approach.2) If this is true, where does the CLRO come from in chaotic DSs?3) There was no definition of chaos for stochastic dynamics
Excellence Cluster Universe, Garching2015/7/17
Long-Range Order = Gapless Excitation
Ubiquity of CLRO
ContradictionAll DSs must possess a (super)symmetry
Requirement
Such a supersymmetry does exist !-) Langevin SDEs: Parisi&Sourlas (1979) -) Its topological meaning: Witten (1982) -> ChTs (1988).-) Pseudo-Hermitian Ev. Op. Mostafazadeh (2002)-) Stochastic Evolution Op.: Ruelle (1987)-) ChT-SDE for all SDEs (2013)
Chaos
All stochastic DSs
Chaotic Long-Range Order: Potential Origin
Cohomological or Witten-
type topological field theory,
Parisi-Sourlas Quantization
Pseudo-Hermitian
quantum theory
Dynamical
systems theory
ChT-SDE – approximation-free theory of SDEs
Probability,
Statistics,
Thermodynamics
69Excellence Cluster Universe, Garching2015/7/17
Geneology of ChT-SDE
ChT-SDE – dynamical generalization of statistics, or
stochastic generalization of DS theory
Mathematical Foundation: Cohomological or Supersymmetric Field Theories