Superstrings in Higher Order Extensions of Finsler Superspaces

13
ELSEVIER Nuclear Physics B 494 [ PM] (1997) 590-656
Superstrings in higher order extensions of

Finsler superspaces
Sergiu I. Vacaru 1
Institute of Applied Physics, Academy of Sciences, 5 Academy str., Chi~in~tu2028, Moldova
Received 6 November 1996; accepted 21 January 1997
Abstract
The work proposes a general background of the theory of field interactions and strings in spaces
with higher order anisotropy. Our approach proceeds by developing the concept of higher order
anisotropic superspace which unifies the logical and mathematical aspects of modem Kaluza-
Klein theories and generalized Lagrange and Finsler geometry and leads 1o modelling of physical
processes on higher order fiber bundles provided with non-linear and distinguished connections
and metric structures. The view adopted here is that a general field theory should incorporate
all possible anisotropic and stochastic manifestations of classical and quantum interactions and,
in consequence, a corresponding modification of basic principles and mathematical methods in
formulation of physical theories.
The presentation is divided into two parts. The first five sections cover the higher order
anisotropic superspaces. We focus on the geometry distinguished by non-linear connection vector
superbundles, consider different supersymmetric extensions of Finsler and Lagrange spaces and
analyze the structure of basic geometric objects on such superspaces. The remaining five sections
are devoted to the theory of higher order anisotropic superstrings. In the framework of supersym-
metric non-linear sigma models in Finsler extended backgrounds we prove that the low-energy
dynamics of such strings contains equations of motion for locally anisotropic field interactions,
Our work is to be compared with important previous variants of extensions of Finsler geometry
and gravity. There are substantial differences, because we rely on modeling of higher order
anisotropic interactions on superbundle spaces and do not propose some "exotic" Finsler models
but a general approach which for trivial or corresponding parametrization of non-hnear connection
structures reduces to Kaluza-Klein and other variants of compactified higher-dimension space-
times. The geometry of non-linear connections (not being confused with connections for non-linear
realizations of gauge supergroups) is first considered for superspaces and possible consequences
on non-linear connection fields for compatible propagations of strings in anisotropic backgrounds
are analyzed. Finally, we note that the developed computation methods are general (in some
aspects very similar to those for Einstein-Cartan-Weyl spaces which is a priority comparing with
other cumbersome calculations in Finsler geometry) and admit extensions to various Clifford and
spinor bundles. © 1997 Elsevier Science B.V.
0550-3213/97/$17.00 t~) 1997 Elsevier Science B.V. All rights reserved.

PH S0550-3213(97)00089-8
S.I. Vacaru/Nuclear Physics B 494 [PM] (1997) 590-656 591
PACS: 02.40.-k; 04.65.+e; I 1.90.+t; 12.10.-g; 12.60.J; 12.90.+b

Keywords: ExtendedFinsler superspaces;Locallyanisotropicsigma models; Superstrings
1. Introduction
The differential supergeometry has been formulated with the aim of getting a geo-
metric framework for the supersymmetric field theories (see for the theory of graded
manifolds Refs. [ 19,66,67,65], for the theory of supermanifolds Refs. [126,88,13,58]
and for detailed considerations of geometric and topological aspects of supermanifolds
and formulation of superanalysis, Refs. [33,26,72,53,120,122]). In this work we apply
the supergeometric formalism for a study of a new class of (higher order anisotropic)
superspaces.
The concept of local anisotropy is largely used in some divisions of theoretical and
mathematical physics [ 121,56,57,78 ] (see also possible applications in physics and biol-
ogy in [6,5] ). The first models of locally anisotropic (la) spaces (la-spaces) have been
proposed by Finsler [38] and Cartan [29] (early approaches and modern treatments of
Finsler geometry and its extensions can be found, for instance, in [90,7,8,74] ). In our
works [ 102-104,106,109,113,119,116] we try to formulate the geometry of la-spaces in
a manner as to include both variants of Finsler and Lagrange, in general supersymmetric
extensions and higher dimensional Kaluza-Klein (super)spaces as well as to propose
general principles and methods of constructing models of classical and quantum field
interactions and stochastic processes on spaces with generic anisotropy.
We cite here the works [ 15,16] by Bejancu where a new viewpoint on differen-
tial geometry of supermanifolds is considered. The author introduced the non-linear
connection (N-connection) structure and developed a corresponding distinguished by
N-connection supertensor covariant differential calculus in the framework of the De
Witt [ 126] approach to supermanifolds in the framework of the geometry of super-
bundles with typical fibres parametrized by non-commutative coordinates. This was the
first example of superspace with local anisotropy. In our turn we have given a general
definition of locally anisotropic superspaces (la-superspaces) [ 106] and higher order
anisotropic superspaces (ha-superspaces) [108]. It should be noted here that in our
supersymmetric generalizations we were inspired by the Miron, Anastasiei and Atanasiu
works on the geometry of non-linear connections in vector bundles and higher order La-
grange spaces [ 75-77]. In this work we shall formulate the theory of higher order vector
and tangent superbundles provided with non-linear and distinguished connections and
metric structures (a generalized model of la-superspaces). Such superbundles contain as
particular cases the supersymmetric extensions and various higher order prolongations
of Riemann, Finsler and Lagrange spaces.
I E-mail: lises@cc.acad.md
592 S.L Vacaru/Nuclear Physics B 494 [PM] (1997) 590-656
Superstring theory holds the greatest promise as the unification theory of all funda-
mental interactions. The superstring models contains a lot a characteristic features of
Kaluza-Klein approaches, supersymmetry and supergravity, local field theory and dual
models. We note that in the string theories the non-local one-dimensional quantum ob-
jects (strings) mutually interacting by linking and separating together are considered as
fundamental values. Perturbations of the quantized string are identified with quantum
particles. Symmetry and conservation laws in the string and superstring theory can be
considered as sweeping generalizations of gauge principles which consists the basis of
quantum field models. The new physical concepts are formulated in the framework of
a "new" for physicists mathematical formalism of the algebraic geometry and topology
[48].
The relationship between two-dimensional o--models and strings has been considered
[71,39,27,96,4] in order to discuss the effective low energy field equations for the mass-
less models of strings. Non-linear tr-models makes up a class of quantum field systems
for which the fields are also treated as coordinates of some manifolds. Interactions are
introduced in a geometric manner and admit a lot of applications and generalizations
in classical and quantum field and string theories. The geometric structure of non-linear
sigma models manifests the existence of topological non-trivial configuration, admits
a geometric interpretation of counterterms and points to a substantial interrelation be-
tween extended supersymmetry and differential supergeometry. In connection to this
a new approach based on non-local, in general, higher order anisotropic constructions
seem to be emerging [ 101,107,108]. We consider the reader to be familiar with ba-
sic results from supergeometry (see, for instance, Refs. [33,67,126,88] ), supergravity
theories [41,92,125,123,124] and superstrings [54,95,63,64].
In this work we shall present an introduction into the theory of higher order anisotropic
superstrings being a natural generalization to locally anisotropic (la) backgrounds (we
shall write in brief la-backgrounds, la-spaces and la-geometry) of the Polyakov's co-
variant functional-integral approach to string theory [85]. Our aim is to show that
a corresponding low-energy string dynamics contains the motion equations for field
equations on higher order anisotropic superspaces and to analyze the geometry of the
perturbation theory of the locally anisotropic supersymmetric sigma models. We note
that this work is devoted to supersymmetric models of locally anisotropic superstrings.
The work is organized as follows: Section 2 contains a brief review on supermanifolds
and superbundles and an introduction into the geometry of higher order distinguished
vector superbundles. Section 3 deals with the geometry of non-linear and linear dis-
tinguished connections in vector superbundles and distinguished vector superbundles.
The geometry of the total space of distinguished vector superbundles is studied in Sec-
tion 4; distinguished connection and metric structures, their torsions, curvatures and
structure equations are considered. Generalized Lagrange and Finsler superspaces and
their higher order prolongations are defined in Section 5. Section 6 contains an intro-
duction into the geometry of two-dimensional higher order anisotropic sigma models
and a locally anisotropic approach to heterotic strings. In Section 7 the background field
method for o--models is generalized for a distinguished calculus locally adapted to the
S.l. Vacaru/Nuclear Physics B 494 [PM] (1997) 590-656 593
N-connection structure in higher order anisotropic superspaces. Section 8 is devoted to

a study of Green-Schwarz action in distinguished vector superbundles. Fermi strings in
higher order anisotropic spaces are considered in Section 9. An example of one-loop and
two-loop calculus for anomalies of locally anisotropic strings is presented in Section 10.
A discussion and conclusions are drawn in Section I 1.
2. Distinguished superbundles
In this section we establish the necessary terminology on supermanifolds (s-manifolds)

[ 126,88,89,58,120,53,72,13,26,33 ] and present an introduction into the geometry of di s-
tinguished vector superbundles (dvs-bundles) [ 110]. Here we note that a number of
different approaches to supermanifolds are broadly equivalent for local considerations.
For simplicity, we shall restrict our study only to geometric constructions on locally
trivial superspaces.
2.1. Supermanifolds and superbundles
To build up s-manifolds [88,58,120] one uses as basic structures Grassmann algebra

and Banach space. A Grassmann algebra is introduced as a real associative algebra
A (with unity) possessing a finite (canonical) set of anticommutative generators /3A,
[flA,flh]+ = flAfl~ + fldfl;t = 0, where A,/~ . . . . . 1,2 . . . . . L. In this case it is also
defined a Z2-graded commutative algebra A0 -4- A1, whose even part A0 (odd part A1 )
is a 2L-l-dimensional real vector space of even (odd) products of generators flA.After
setting A0 = 7E -4- A0 t, where 7~ is the real number field and A0 t is the subspace of A
consisting o f nilpotent elements, the projections ~ : A ---+ 7~ and s : A ~ Ao' are called,
respectively, the body and soul maps.
A Grassmann algebra can be provided with both structures of a Banach algebra and
Euclidean topological space by the norm [88]
L
X-",.,A,...L~ .
II~II = ~ Ia;~'"'A~I.Zg = Z._.,~ ,-%. .3~ .
Ai r--O
A superspace is introduced as a product
A n'k = A0× • • • × A o ×A, × . . . ×AI

Y
which is the A-envelope of a Z2-graded vector space V"'k = Vo®Vt = ~ n G 7~k is

obtained by multiplication of even (odd) vectors of V on even (odd) elements of A.
The superspace (as the A-envelope) possess ( n + k ) basis vectors {/3i, i = 0, 1 . . . . . n - 1 } ,
and {fli,~ = 1,2 . . . . k}. Coordinates of even (odd) elements of V n'k are even (odd)
elements o f A. We can consider equivalently a superspace V "'k as a (2 L-1 ) (n + k)-
dimensional real vector spaces with a basis {/~i(A),/~/(A)}"
594 S.I. Vacaru/Nuclear Physics B 494 [PM] (1997) 590-656
Functions of superspaces, differentiation with respect to Grassmann coordinates, su-

persmooth (superanalytic) functions and mappings are introduced by analogy with the
ordinary case, but with a glance to certain specificity caused by changing of real (or
complex) number field into Grassmann algebra A. Here we remark that functions on a
superspace A~'t which takes values in Grassmann algebra can be considered as mappings
of the space T¢{2~L-'~){n+k) into the space 7"~2L. Functions differentiable on Grassmann
coordinates can be rewritten via derivatives on real coordinates, which obey a generalized
form of Cauchy-Riemann conditions.
A (n, k)-dimensional s-manifold M can be defined as a Banach manifold (see, for
example, Ref. [68] ) modelled on A n'k endowed with an atlas ~b = {U{i),¢{i) : U{i)
A n'k, (i) E J} whose transition functions ~ ( i ) a r e supersmooth [88,58]. Instead of
supersmooth functions we can use Goo-functions [ 88 ] and introduce G °°-supermanifolds
(Goo denotes the class of superdifferentiable functions). The local structure of a Goo-
supermanifold is built very much as on a C~-manifold. Just as a vector field on a
n-dimensional Coo-manifold written locally as
n-1
Eii(x ) e--
tgXi '
i=0
where fi are C°°-functions, a vector field on an (n, k)-dimensional Goo-supermanifold

AT/can be expressed locally on an open region UChT/as
n-l+k n-I k
s,(x') # = a o3) a ~ '
1=0 i=O i=1
where x = ( ~ , 0 ) = {x I = (.~i,0~)} are local (even, odd) coordinates. We shall use

indices I = (i,~), J = ( j , ) ) , K = (k, ~:) . . . . for geometric objects on /17/. A vector field
on U is an element XCEnd[Goo(U)] (we can also consider supersmooth functions
instead of Goo-functions) such that
X ( f g ) = ( X f ) g + (-)lfllXl fXg,
for all f, g in Goo (U), and
X ( a f ) = (-)IXllalaxf,
where IX[ and lal denote correspondingly the parity (= 0, 1) of values X and a and for
simplicity in this work we shall write (-)Ifllxl instead of ( - 1 )IfllXl.
A super Lie group (sl-group) [89] is both an abstract group and a s-manifold,
provided that the group composition law fulfills a suitable smoothness condition (i.e. to
be superanalytic, for short, sa [58] ).
In our further considerations we shall use the group of automorphisms of A (n'k),
denoted as GL(n, k, A), which can be parametrized as the super Lie group of invertible
matrices
S.L Vacaru/Nuclear Physics B 494 [PM] (1997) 590-656 595
D '
where A and D are respectively ( n × n ) and (k×k) matrices consisting of even Grass-
mann elements and B and C are rectangular matrices consisting of odd Grassmann
elements. A matrix Q is invertible as soon as maps o-A and o-D are invertible ma-
trices. An sl-group represents an ordinary Lie group included in the group of linear
transforms G L ( 2 L-1 (n + k ) , R ) . For matrices of type Q one defines [ 19,66,67] the
superdeterminant, sdet Q, supertrace, strQ, and superrank, srank Q.
A Lie superalgebra (sl-algebra) is a Z2-graded algebra A = A0 @ Aj endowed with
product [, } satisfying the following properties:
[l, I'} = - ( _ ) Itlll'l [1,, I},

[1, [i', i"}} = [ [1, t'}, i"} + (-)I~ll~'l [i'[1, i")),
IEA[/I, lIcAit,i , where Ill, II'1 = 0, 1 enumerates, respectively, the possible parity of
elements I, F. The even part A0 of an sl-algebra is a usual Lie algebra and the odd
part AI is a representation of this Lie algebra. This enables us to classify sl-algebras
following the Lie algebra classification [59]. We also point out that irreducible linear
representations of Lie superalgebra A are realized in Z2-graded vector spaces by matrices
( A D0) f°r even elements and ( 0f ° rB° d) d e l e m e n t s a n d t h a t ' r ° u g h l y s p e a k i n g ' C 0
A is a superalgebra of generators of an sl-group.
An sl-module W (graded Lie module) [88] is introduced as a Zz-graded left A-
module endowed with a product [, } which satisfies the graded Jacobi identity and
makes W into a graded-anticommutative Banach algebra over A. One calls the Lie
module G the set of the left-invariant derivatives of an sl-group G.
The tangent superbundle (ts-bundle) T/f/ over an s-manifold .~7/, 7r : T/f/ ~ A7/ is
constructed in a usual manner (see, for instance, [ 68 ] ) by taking as the typical fibre the
superspace An'~ and as the structure group the group of automorphisms, i.e. the sl-group
GL(n,k,A).
Let us denote by .T" a vector superspace (vs-space) of dimension (m, l) (with respect
to a chosen base we parametrize an element y E C as y = (9, ( ) {ya = (fa, (~)}, =
where a = 1,2 . . . . . m and h = 1,2 . . . . . l). We shall use indices A = (a,&),B =

(b, b) . . . . for objects on vs-spaces. A vector superbundle (vs-bundle) ~ over base /Q
with total superspace/~, standard fibre f" and surjective projection 7rE :/~--*M is defined
(see details and variants in [26,122]) as in the case of ordinary manifolds (see, for
instance, Refs. [68,75,76] ). A section of ,~ is a supersmooth map s : U---,/~ such that
7re. s = idu.
A subbundle of ~ is a triple (/3, f, f / ) , where/3 is a vs-bundle on/f/, maps f :/3---~
and f : AT/---~Mare supersmooth, and (i) 7reof = f%TrB; (ii) f : 7r~l ( x ) _ o T r - ~ l o f ( x )
is a vs-space homomorphism. We denote by
u . ( x , y ). . O,~9,()
(Yc, . {u. a (xl,y a ) (~ci, o~, yâ , ( h ) = ( $ i , x'i ,yâ ,y~)}
the local coordinates in 8 and write their transformations as
x t' = x ~ ' ( x l ) , srank \-~-xl ] = ( n , k ) , (1)
yA' = y2' ( x ) y a , where Y2' (x) CG(m, l, A).

For local coordinates and geometric objects on the ts-bundle Tif/ we shall not dis-
tinguish between indices of coordinates on the base and in the fibre and write, for
instance,
u = ( x , y ) = (Yc, O , ~ , ( ) = {u a = ( x l , y 1) = (Xî ,0 i ,yî , ( i ) = (2i, x i , ~ i , y i ) } .
We shall use Greek indices for marking local coordinates on both s-vector and usual
vector bundles.
2.2. Distinguished vector superbundles
Some recent considerations in mathematical physics are based on the so-called k-jet
spaces (see, for instance, Refs. [94,93,9] ). In order to formulate a systematic theory
of connections and of geometric structures on k-jet bundles, in a manner following the
approaches of Refs. [ 128,75,76], Miron and Atanasiu [77] introduced the concept of a
k-osculator bundle for which a fiber of k-jets is changed into a k-osculator fiber repre-
senting an element of a k-order curve. Such considerations are connected with geometric
constructions on tangent bundles of higher order. On the other hand, for developments
in modern supersymmetric Kaluza-Klein theories (see, for instance, Ref. [92] ) a sub-
stantial interest would present a variant of "osculator" space for which the higher order
tangent s-space distributions are of different dimensions. The second part of this section
is devoted to the definition of such type of distinguished vector superbundle spaces.
A vector superspace br(z) of dimension ( m , l ) is a distinguished vector super-
space (dvs-space) if it is decomposed into an invariant oriented direct sum br(z) =
•~"(1) @ ~ ' ( 2 ) @ ' ' ' @ ~(Z) of vs-spaces .Y'(p),dim.F'(/, ) = ( m ( p ) , l ( p ) ) , where ( p ) =
(1), (2) . . . . . ( z ), ~-]~Pp~m(p) = m, ~-~Pp=_~l(p) = l.
Coordinates on .T"(p) will be parametrized as
y(P) = (y(l), y(2) . . . . . y(p) ) = (~(1), ((1), 33(2), ((2) . . . . . ~9(p), ((p))
= {y(A) = (~(a),~-(a)) = (~(a),y(a))},
where bracketed indices are correspondingly split on Ufp)-components:
(A) = (A(I),A(2) . . . . . A ( p ) ) , (a) = (a(1),a(2) . . . . . a(p)),

(~) = (~(1)~(2) . . . . . ~(p)), (2)
For simplicity, we shall also write (2) as (A) = (A1, A2 . . . . . A/, ) , (a) = (a., a2 . . . . . ap)
and (fi) = (ala2 . . . . . ~ , ) if this will give not rise to ambiguities.
A distinguished vector superbundle (dvs-bundle) glz) = (/~(z), ¢r(d),.y-(d), ~ ) , with

surjective projection ¢r (z) :/~(z) ~ ~ , where M and/~(z) are respectively base and total
s-spaces and the dvs-space f-(z) is the standard fibre.
A dvs-bundle g(z) is constructed as an oriented set of vs-bundles or(p) : ~(i,) __~ ~'(p-J)
(with typical fiber .U(P),p = 1,2 . . . . . z); ~(0) = /f/. We shall use index z ( p ) as to
denote the total (intermediate) numbers of consequent vs-bundle coverings of M.
Local coordinates on g(P) are denoted as
U(p) = (X, y(p) ) = (X, y(1), Y(2) . . . . . Y(p))
= (X,O,y(p),((p)) = (-~,0, y(I),~'(I),Y(2>,~'(2) ..... y(p),((p))

: {U(a) = (X I, y(a) ) = (~i, 0~, ~:(a), ((a)) = (2i, x~i, ~(a), y(a) )
= ( X I = y Ao , y Al . . . . . y Ap . . . . . y a : ) }
(in our further considerations we shall consider different variants of splitting of indices
of geometric objects)•
Instead of (1) the coordinate transforms for dvs-bundles {u(") = (xl,y (A))}
{U(a') = (X 1', y ( a ' ) ) } are given by recurrent maps:
x ' ' = x t' ( x )1,

:0x,
srank~,0x I/ = (n,k), (3)
a't) _KA'Itx~.A1
Y(l - A1 " : Y ( 1 ) ' K AAI' l ( x ) C G ( m ( l ) , l ( l ) , A )
A/',) = Kaa! (u(p-1 ) ) y(p

Ap), K A,
A;' ( U(p_l ) ) ¢ G ( m(p ), l(p ), A ) ,
y(p
a'. '
y(;) = K2: (U(z_l))y:z), KaA: ( U ( z _ , ) ) E G ( m ( z ) , l(: ), A ) .
In brief we write transforms (3) as
X f :_ x f ( x l ) , y(A') = .~{A
i((A')a
> Y (A} •
('~') = (KI1' '~(A)/C(AZ)~:'where K tr - O x l ' / O x / .

More generally, we shall consider matrices K (,,)
As a consequence, the local coordinate bases of the module of ds-vector fields
=(g<z>),
O~(a) = (~I,Og(A)) = ( 0 1 , 0 ( A I ) , 0 ( A 2 ) . . . . . O(A:)
0 o9 a a 0
(4)
- Ou ('~) - ( ~x t' Oy~,') Oy(A~) Oy¢;)
(the dual coordinate bases are denoted as
d (~) = ( d l , d ( a ) ) = ( d l , d ( a , ) , d (&) . . . . . d (A'-))

= du (a) = (dx I, dy (al), dy(a2) . . . . . dy (A=)) ) (5)
are transformed as
O(a ) -~ ( Cgl, C~(A} ) -~ ( C~I, Cg(A,), O( A2) . . . . . O( az) ) ""+ O(a)

= (O3/,t~(A}) = (O31,O~(A,),O~(A2) . . . . . O~(a:))
O3 1' Oq
eX] = K] Ox]"---771- Yi4'lo)l~l{)Y(a2'°Oy(A-~22)
' O
)/
+
' Oq
"~'''
.A'.
"JI- Y(z:O)I ay(--A'
O~
)
0 _ KA'I O........~+ y A~ ~ +..._.~ ~,Ah ) A, o

aye,') A, ay~,l) (2,I)AI ayeS) ay(z
)a''
a = KA~ ¢9 y, A3, _~ a ~.
m2 A 2 7 "~- (3,2)A2_ a~ + ' ' " "~- Y(z~2)a2~--At, '
aY(2) aY(2) 0Y(3) °Y(z)
O ' 01 A I" C~
__ K A z -1
A"~'-I a.-I 7
A'~'-I ~-Y(z:z-1)A,_l~ A t•
3y(z_l) " Oy(z_l ) ay(~)
a =/qA: ,9
A'. " (6)
Oyez ) " By(z )
Y-matrices from (6) are partial derivations of corresponding combinations of K-

coefficients from coordinate transforms (3),
t
yA,', 0 ( K AAVpyAp )
Ay -- o~yAf ' f < P"
In brief we denote respectively ds-coordinate transforms of coordinate bases (4) and

(5) as
, <o/+
where matrix K (d)(,~), its s-inverse KI:! ) , as well Y((~} and Y ~ are parametrized according
to (6). In order to illustrate geometric properties of some of our transforms it is useful
to introduce matrix operators and to consider in explicit form the parametrizations of
matrices under consideration. For instance, in operator form the transforms (6)
a =~O',
are characterized by matrices of the type

0
Oxt C'~XIt
0 0
0 = c9(,~) =
2I
OA2
OA
=
OY~lI)
0
A2
OY}2)
01 = 0(,~,) =
CgA'1
OA~
OA'.
=
Oy'/l'
O
OyeZ)
0 0
A ~.
\ OY'~z) \ Oy(~) /
and
/' y A't A~ A~.
Kt (1,0)1 Y(2,0)/ ' ' ' Y(z~o)l
0 K A'~ a'2 A'.
AI Y(2,1 )al Y(z~l)al
= 0 0 K a~ At"
A2 V(z~2)A2
t
0 0 0 ... KAA~
W e note that we obtain a s u p e r s y m m e t r i c generalization o f the M i r o n - A t a n a s i u [ 77]

osculator bundle (OscZif4, Tr, if4) if the fiber space is taken to be a direct sum o f z
vector s-spaces o f the same d i m e n s i o n dim 5t" = d i m / ~ , i.e. 5r<d) = f " • 5t- ® . . . ® 5t'. In
this case the K and Y matrices from (3) and (6) satisfy the identities
KArl _ KA; A:
At - A2- = "'" = K A y '
A' At A'
Y(1,0) a = Y(2,I)A . . . . . Yiz,z-l)a'
A' A' At
Y(l,,O)A=Yip+l,l)a ..... Y(z,z-1)A, (p =2 ..... Z -- 1 ) .
For s = 1 the Oscl~l is the ts-bundle T/14.

I n t r o d u c i n g projection ~-~ - ~.(z) : {(z) ~ ~ we can also consider projections
rl"
P2 " ~(m) ___,~(m) (P2 < Pl ) defined as
77.szSl( x, y( l ) ' " .. , Y(P' ) ) = ( x , y ( 1 ) , . . . ,y(P2)).
The s-differentials d~P~ : T ( g (m)) ~ T ( g (p:)) of maps zrP~2 in turn define verti-
cal d v s - s u b b u n d l e s Vh+l = Kerdcrt~h ' ( h = 0, 1 . . . . . pl - 1) of the tangent d v s - b u n d l e
T ( g (z)) (the d v s - s p a c e Vl = V is the vertical d v s - s u b b u n d l e on {(z). The local fibres of
d v s - s u b b u n d l e s Vh d e t e r m i n e s this regular s-distribution Vh+t : u E ~(z) ~ Vh+l (u) C
T ( g (z)) for which one h o l d s inclusions Vz C Vz-1 C . . . C Vt, The enumerated proper-
ties o f vertical d v s - s u b b u n d l e s are explicitly illustrated by transformation laws (6) for
d i s t i n g u i s h e d local bases•
3. Non-linear connections in dvs-bundles
The purpose of this section is to present an introduction into geometry of the non-
linear connection structures in dvs-bundles. The concept of non-linear connection (N-
connection) was introduced in the framework of Finsler geometry [30,29,60] (the
global definition of N-connection is given in [12] ). It should be noted here that the
N-connection (splitting) field could play an important rule in modeling various variants
of dynamical reduction from higher dimensional to lower dimensional s-spaces with
(or not) different types of local anisotropy. In monographs [75,76] there are contained
detailed investigations of geometrical properties of N-connection structures in v-bundles
and different generalizations of Finsler geometry and some proposals (see Ch. XII
in Ref. [75], written by S. Ikeda) on physical interpretation of N-connection in the
framework of "unified" field theory with interactions non-localized by y-dependencies
are discussed. We emphasize that an N-connection is a different geometrical object from
that introduced by using non-linear realizations of gauge groups and supergroups (see,
for instance, the collection of works on supergravity [92] and approaches to gauge
gravity [98,86] ). To make the presentation to aid rapid assimilation we shall have
realized our geometric constructions first for vs-bundles and then we shall extend them
for higher order extensions, i.e. for general dvs-bundles.
3.1. N-connections in vs-bundles
Let us consider the definitions of N-connection structure [ 106] in a vs-bundle ~ =

(/~, rrE, aqr) whose type fibre is be. and ~rr : T~--+T191 is the superdifferential of the
map rrE (It 7 is a fibre-preserving morphism of the ts-bundle (T~, re,/17/) to/~ and of
ts-bundle (TM, r, 37/) to M). The kernel of this vs-bundle morphism being a subbundle
of (TE, rE,/~) is called the vertical subbundle over ~ and denoted by V~ = ( V/~, rv,/~).
Its total space is V~ = [.Ju~g vu, where V~ = ker~-7, uC~. A vector
y = ya c9 t ~ a 0 = yiO..~_ y'iO ya 0 ya 0
+V A ox , + eo , + +
tangent to ~ in the point u E ~ is locally represented as
( u , Y ) = (u'~,Y ~') = ( x l , y A , y * , Y A) = (sci, o ~ , p " , ( a , ? i , Y ~ , ? " , Y a ) .
A non-linear connection, N-connection, in vs-bundle g is a splitting on the left of the

exact sequence
O, ,V{, i Tg,, ,Tg/Vg, ,0, (7)
i.e. a morphism of vs-bundles N : T~ E V~ such that Noi is the identity on V~.

The kernel of the morphism N is called the horizontal subbundle and denoted by
(HE, rE, E).

S.I. Vacaru/NuclearPhysicsB 494 [PM](1997)590-656 601
From the exact sequence (7) it follows that the N-connection structure can be equiva-
lently defined as a distribution {/~u ~ HuE, Tu/~ = HuE'@Vu~:} on ~; defining a global
decomposition, as a Whitney sum,
T g = n g + Vg. (8)
To a given N-connection we can associate a covariant s-derivation on M:
Xt { OYA
VxY = Oxt + NA(x, Y)}sa, (9)
where SA are local independent sections of g, Y = YASA and X = Xtst.

S-differentiable functions N a from (3) written as functions on x t and yA, N a ( x , y ) ,
are called the coefficients o f the N-connection and satisfy the following transformation
laws under the coordinate transforms (1) in £:
A' tgXt' = A/IA'MA OM a' (X) yA.

Nt' ~X l ' " A ' ' l OXt
If the coefficients o f a given N-connection are s-differentiable with respect to the
coordinates ya we can introduce (additionally to covariant non-linear s-derivation ( 9 ) )
a linear covariant s-derivation/9 (which is a generalization for vs-bundles of the Berwald
connection [21] ) given as follows:
where
I~IA/(x, y) -- o N A t ( x ' Y) (10)

OyB
and
~ c ( X , y) = o.
For a vector field on g, Z = Z I 0-Trx

a + YA 0y--X
a and B = BA(y)O@x being a section in
the vertical s-bundle (VE, rv,F.) the linear connection (10) defines the s-derivation
(compare with ( 9 ) )
L)zn ~ [Zl ( aRt -~- ~[AIBB) ~- yBanA1 •

\ ~x ~ c)yB J a y A"
Another important characteristic of an N-connection is its curvature (N-connection

curvature):
1-2 = ~1 j,2Ajdx! A dx ] Q Oy---

a~
with local coefficients

a:j ~gg: g A
= (_~ltJI oN:] + NtB I~Bj
A BÂ
- (_)ltJI Nj N m, (11)
OnXJ ", ., OX1
where for simplicity we have written (_)IKIIJI = (_)IKJI

We note that linear connections are particular cases of N-connections locally para-
metrized as N'~(x,y) = N A t ( x ) x t y B, where the functions NAt(x), defined on M, are
called the Christoffel coefficients.
3.2. N-connections in dvs-bundles
In order to define an N-connection into dvs-bundle ~(z) we consider an s-subbundle

N (~(z)) of the ts-bundle T ( g ( z ) ) f o r which one holds (see Refs. [94,77] respectively
\ /
for jet and osculator bundles) the Whitney sum (compare with (8))
N ( f f (z)) can be also interpreted as a regular s-distribution (horizontal distribution being

supplementary to the vertical s-distribution V ( g (z))) determined by the maps N : u E
g(z> _~ N(u) c T~(g(z>).
The condition of existence of an N-connection in a dvs-bundle ~(z) can be proved as
in Refs. [75-77]: It is required that g(z) is a paracompact s-differentiable (in our case)
manifold.
Locally an N-connection in g(z) is given by its coefficients
r.tAl (U), (N:~2)t(u) , A2 At,

lv(01)l, N(12)A, ( U) ) , . .. ( N(ov )t( u ), N(AI,[p)A 1( U ) ....
Ap
N(/,_lp)Ap_ ' (u)) . . . . .
A= A=
( N(oz)/(u), N(lz)AI(U) . . . . . N:pz)At, (U) . . . . . N:z_Iz)A:_, (U) ),
ap
where, for instance, ( N(op)l(U), at, (U) . . . . . N(p_lp)A~,_
N(lp)AI Ap l ( U) ) are components of
the N-connection in the vs-bundle 7r(p) : E(P) ~ ~(p-l). Here we note that if an N-con-
nection structure is defined we must correlate to it the local partial derivatives on ~(z)
by considering instead of the local coordinate bases (4) and (5) the so-called locally
adapted bases (la-bases)
~(o~) = ( 8 1 , 8 ( A ) ) ~- ( 8 1 , 8 ( A I ) ,8(A2) . . . . . 8(A~))

8 8 ~ 8 8
' ' A2 ' ' ' ' '
(12)
- ~ (~ Oy:l' ) OY(2) OY:z) )
(the dual la-bases are denoted as
~(ot) = ( ~ I , 8 ( A ) ) : (SI,~(AI),8(A2) ..... ~(A:))
= 8u(,~) = (Sx t, ~y(AO By(A2). . . . , ~Sy(A~) ) ) (13)
with components parametrized as

SI = 31 -- N l
A[
t~a I --
A2
N t OA2 -- • • • -- N I
A:_,
Oa: 1 --
NA:OA: , (14)
• A2~A - A3~. -- Az-I A:
~AI = OAI -- IVA I 2 -- N A t ~3 • • • -- NAt aA:_t -- N & aA:,
A3 A4 Az - t A.
~A2 = C~A2 -- N A 20A~ -- ga20A4 -- • • • -- NA2 C~A:- I -- NA20A:,
A.
~A:_I = OA:_I -- NA~_ I OA:'
or, in matrix form, as
a . = N ( u ) x O,,
where
:A'l)
(0__,
Ox / #X 1
6
("1
8At OY'~It Oy~
[ t~A2 = 8 O. = 0(4) = 0A2 = 0
• Oy(~2 A2
OY(2
\aA= c)
\ OYez /
1 - N At - N A2 . . . - N A:
0 1 --NA~ "'"- N AA,:
NA:
0 0 1 "''-- A2
0 0 0 ... 1
In generalized index form we write the matrix (6) as "~(~)

( 8 ) ' where, for instance, ]V~ =
~]j, AAI AI N¢I _NAI ~az NA: ~a: __NA:
N B j = ~Bt . . . . , = ' • " "' At = -- At ' A2 = A2 ' " • "
So in every point u E ~(z) we have the following invariant decomposition:
Tu / \~(d))= N o ( u ) (~ N , ( u ) • . . . • N z - , ( u ) G Vz ( u ) , (15)
where 8t E NO,SA~ C N l . . . . . 8a:_t E g z _ l , a A = E Vz.

We note that for the osculator s-bundle ( O s c Z l f l , rr, 1(4) there is an additional (we
consider the N-adapted variant) s-tangent structure
J: x(OscZ ) x(osczr4)
defined as
3y/l ) - J ~x l . . . . . 3y[z_l-----~)- J 3y/~-_2) 3y[z ) J (16)
(in this case I- and A-indices take the same values and we can not distinguish between
them), by considering the vertical J-distributions
No=N, N I = J ( N o ) . . . . . N z - l = J ( N z - 2 ).
In consequence, for the la-adapted bases on (OscZifl, Tr, if'l) we have the following
N-connection matrix:
I _NJ(1)I -N(2)I
J •.. -NJ(z)1
0 1 -N(1)IJ -- N (sz - l ) 1 [
N = ~(J) (17)
"" (1) = 0. .0 . 1. -NJ~i-2)1 ] .
0 0 0
There is a unique distinguished local decomposition of every s-vector X E X ( ~ (z))

on the la-base (12):
X = X (H) + X ( ~ ) + . . . + X ( ~ ) , (18)
by using the horizontal, h, and verticals, vl, v2 . . . . . vz projections:
X (H) = h X = X 1 t 3 l , X (V~) _ ~ u 1 X = X ( A I ) 6 A I . . . . . X (V:) _ ~ U z X = X ( A z ) ~ A z .
With respect to coordinate transforms (4) the la-bases (12) and ds-vector components
(18) are correspondingly transformed as
6 3x 1' 6 6 =KA'p B
cgxl -
cgxl Oxll ' Ap
Oy(p) av Oy(p)
a,', ' (19)
and
X 1' = Ox 1'
Xt x(Av, ) = Ka~X(,,),
A ~ AI
Vp = 1,2,. z.
cgX1 ' , •.,
Under changing of coordinates (3) the local coefficients of a non-linear connection

transform as follows:
v<d)g?<t~'> = ~</~)t~,'<#') +. ,-<#'),

x<~> j
(we can obtain these relations by putting (19) and (6) into (14) where ~(#') satisfy
6(~,)= (4) (t~)).
For dual la-bases (13) we have the following N-connection "prolongations of differ-
entials":
(~X I = d x I,
~yal }l.,~A~ ~'~
= dy~l, ) + ""(1)t A..I ,
A2 it4A2 tl,~Al IIAA2 t i l l

t~Yaz = dY(2) + "'" (2)AI e . r ( I ) "q- " (2)1 " a " '
6yA~ =dy~j~)- ..As . A! _ ,,As . A~ A. I

~Vl(s)AtaY(l) ~ ~Vt(s)A2aY(~) + . . . + M ( z ) t d x , (20)
where M~°)° are the dual coefficients of the N-connection which can be expressed
explicitly by recurrent formulas through the components of the N-connection ~,(t)
''(A)"
To
do this we shall rewrite formulas (20) in matrix form:
6* = d* x M ( u ) ,
where
a ° = ( ~ X 1 t~y a' t~y A2 . . . t$y a` ) , d ° = ( d x I dy~') t~y(2

A2) . . . (~Y~s))
and
1M(1)I
Ai A2
M(2)l ... M(~)t
A.
A2 A.
0 1 M(2)A I ... M(~)A t
A.
M = 0 0 1 ... M(~A 2
0 0 0 ... 1
and then, taking into consideration that bases Oo ( 8 . ) and d*(8 °) are mutually dual, to
compute the components of matrix M being s-inverse to matrix N (see (17) ). We omit
these simple but tedious calculus for general dvs-bundles and, for simplicity, we present
the basic formulas for osculator s-bundle (Osc z ,(4, 7r, ,(4) when J-distribution properties
(16) and (17) alleviates the problem. For common type of indices on /(/ and higher
order extensions on Osc z ~1 the dual la-base is expressed as
t~X I = d x I ,
J)
6y[2 ~ = dY[2 ) q- M(1)jdy(l
1 q- M~2)jdx J,
6y[~) = dy[z ) + M o1) j d y ( sJ_ l ) 1 J

+ M(2)jdy(z_2) + . . . + M ~ z ) j d x J,
with the M-coefficients computed by the recurrent formulas:

I (21)
M~I )~ = N(1)s,
I 1 1 K
M(2)j = N(2)j -~- N ( 1 ) K M ( 1 ) j ,
606 S.l. Vacaru/Nuclear Physics B 494 [PM] (1997) 590-656
1 1 1 K 1 K
M(s) J = N(s) J + N ( s _ l ) r M ( i ) j + . . . + N(2)KM(z_2)j + N~ l )KM(z
r -1)J"
These transformation laws for dual coefficients (21) hold with respect to coordinate
transforms (3):
K " ff K' I'
M(I)JY(o,o)K -- M(1)K, Yio,o)J + Y( 1,0)J,
K 1' _ 1/ K' Ij Kt It
M(2)JYio,o)K - M(2)K'Yio,o)J + M(1)K'Y(I,O)J + r(2,0)J,
+ +... + +
(The proof is a straightforward regrouping of terms after we have put (3) into (21)).
Finally, we note that curvatures of an N-connection in a dvs-bundle ~(z) can be
introduced in a manner similar to that for usual vs-bundles (see (11 )) by a consequent
step by step inclusion of higher dimension anisotropies:
1 l,~At, 8
f 2 ( p ) = ~ (p)Otp_lflp_l ~U Ott,-I /~ (~Uflp-I @ - - At, ' p=l,2 . . . . . Z,
Oy(p)
with local coefficients
At' t~N~P-I )[fll,-lyt,-i] ~N~A;-|

J'~(p)flt,-t'gp-i - ,Q y p - i (- ~uflp_ I
vU(p_l) tp-1)
^ 7
where ga',yt,_~ At' De (we consider yAo ~ X I ).
= 6N:jt,_,/Oy<p)
4. Geometric objects in dvs-bundles
The geometry of the dvs-bundles is very rich and could have various applications in
theoretical and mathematical physics• In this section we shall present the main results
from the geometry of total spaces of dvs-bundles.
4.1. D-tensors and d-connections in dvs-bundles
By using adapted bases (12) and (13) one introduces algebra D T ( ~ (z)) of dis-
tinguished tensor s-fields (ds-fields, ds-tensors, ds-objects) on ~(z), 7" = 7"pp~m....p: qqlq2...qz '
which is equivalent to the tensor algebra of vs-bundle 7rhvjv2...v: : H ~ (z) @V1~(z) ~ V2~(z ) @
• " " O ~ ( z ) - ~ £ (z), hereafter briefly denoted as ~dz. An element Q c 7"pplp2...p:
qqlq2...qz ' a ds-field
° f type ( P Pl
ql P2
q2 .".". Pz
qz ) ' can be written in l°cal f°rm
S.I. V a c a r u / N u c l e a r Physics B 494 [PM] (1997) 5 9 0 - 6 5 6 607
®OAI ® • • • ® ~mpl ® 3 B I 3 B I @ • • " @ 3801 ® OE1 @ • • • @ 3Et, z @ ~C1 @ . . .
® 3 ca2 ® • • • ® OD~ ® • •. ® 3Dp: ® 6 F~ ® • • • ® 3 e': . (22)
In addition to ds-tensors we can introduce ds-objects with various s-group and coor-
dinate transforms adapted to the global splitting (15).
A linear distinguished connection, d-connection, in a dvs-bundle ~(z) is a linear con-
nection D on ~(z) which preserves by parallelism the horizontal and vertical distributions
in g(z).
By a linear connection of an s-manifold we understand a linear connection in its
tangent bundle.
Let us denote by ~ ( M ) and ~ ( g ( P ) ) , respectively, the modules of vector fields on s-
manifold/Q and vs-bundle g(P) and by .T'(M) and .T'(g (t')), respectively, the s-modules
of functions on M and on g(P).
It is clear that for a given global splitting into horizontal and verticals s-subbundles
(15) we can associate operators of horizontal and vertical covariant derivations (h- and
v-derivations, denoted respectively as D (h) and D ( " : 2 " : ) ) with the properties
D x Y = ( X D ) Y = D h x Y + D . ~ x Y + D,,zxY + • • • + D.:xY,
where
D(xh)Y = DhxY,, D(xh) f = ( h X ) f
and
D~")Y=DvpxY, D~[P)f=(vpX)f, ( p = l . . . . . z)
for every f E 9r(217/) with decomposition of vectors X, Y C E ( g ( z ) ) into horizontal and

vertical parts, X = h X + v i X + . . . + v z X and Y = h Y + v~Y + . . . + vzY.
The local coefficients of a d-connection D in g(z) with respect to the local adapted
frame (5) separate into corresponding distinguished groups. We introduce horizontal
:It K ( bl ) . . . . . LA:K(U)) of D ¢h)
r(A) ~ = ( L I j K ( u ) , L A , x ( u ) , L B 2A2
local coefficients ~.~JK'~(B)K:
such that
' -~- (p = 1 . . . . . z),

~(~.~, ) 3 x g = L t j K ( U ) 3X 1 , ~(8.~ ) 3y~t~) .,,,)
D(h) 3q
and p-vertical local coefficients

¢", I t'-,( A ) t"~ a I (U) A2 a. (p = 1 . . . . . z)
•~jic>,,~(s>~c>) = (CJcp(U),,..B,c, ' ,CB2cp(U) ..... CB:c,,(u))
such that
D(Vr) ~_ 8 D(Vp) ~ Af
~_~7 ) 8x J = C~c, ( u ) ~x ~ , ( ~ ) ~.,B: = CB:c,, A: '
oy I' ~ y f f ) 6y(f)
D(V.) ~ 8q
( ~ ) q - OyC,, '
where q c .T(g(z)), f = 1 . . . . . z.
The covariant ds-derivation along vector X = X 1 ~6 + yAl 6 + . , , + yA~ ~ of a
ds-tensor field Q, for instance, of type ( p Pr ~ , l ~<r ~< z, see (22), can be written
\ q qr /
as
DxQ=D x(h) 0~ + O~x",)a + . . . + D~'=)Q,
where the h-covariant derivative is defined as
D ~xh) Q = XK Q tjar K~l @~ar Qd l @~By,
with components
~g")lAr
rl z'~HAr Z A, ['llCr -- r H I-liAr ICy DIAr
QJB~IK
=lAy cgXK~JBr_~_ ~ HK~a~JBx -~- CIK~JBI L, JK~HBr --~BrK~JC~
and the vt,-covariant derivatives are defined as

O~'P) a = X Cp Qjs~±cp61®Oa~®d
tAr l ®8 n r ,
with components
~l') lAy
= ,, C1 ,,'.hHA~ CAr t')lFr (',H g'liAr t,",F~ [')IA~
OJBr±CplAr ~JBRoycp "~- HCt' ~JBR + FrCp~JBR -- ~JC v ~HFR -- ~BrC e ~'~JFR"
The formulas presented above show that D F = ( L . . . . . L(p ) . . . . . C . . . . . C(p ) . . . . ) are

the local coefficients of the d-connection D with respect to the local frame (~-~r~, ~). a
If a change (3) of local coordinates on gIz> is performed, by using the law of transfor-
mation of local frames (19), we obtain the following transformation laws of the local
coefficients of a d-connection:
1' Oxlt OxJ OXM 1 Oxl/ 02xM

L jtM' -- ~X I OXj-'7cgxM, L JM '~ c~xM c~XJ, OxM,, (23)
p~
, ., a_M Af aK~ j,
Af = l(.,qfl(Bf uA [ KA 'Y
L(f)B~f M' -LA . . . . ="W..,--
S t~f OXM (f)BIM + Cs ~ '-s
1' Ox 1' Ox s c~ t ' '

C(p)J'C/, -- Ox I ~xj'Kc~C(P)-tc,, ' ....
- - C AS
B tfC~t
= KAIKBI,
Af B/ KC,;cAI
C~ BICv , "'"
As in the usual case of tensor calculus on locally isotropic spaces, the transformation
laws (23) for d-connections differ from those for ds-tensors, which are written (for
instance, we consider transformation laws for the ds-tensor (22)) as
S.l. V a c a r u / N u c l e a r P h y s i c s B 4 9 4 [ P M ] (1997) 5 9 0 - 6 5 6 609
1t i t ,/ , t ~t =t ~ t ~ t
1... p.~t i...t~plL~l...~p2...L~,l....IJps
Qy,I ... j t qB t I ""Dql
o t ~ t f't ~t wt
t~I '''"~q2 ""rl ""rqs
--
Ox t; OxJ~ k,A t, ld.BI
O'
Kottl; IdFp, D I I ...lpAi ...A,,, El ...Ep2 ...DI ...D ....
C)XII C)XJ( . . . . . AI "=B~ . . . . . Fp¢ ~$Jl...JqBI...Bql CI...Cq2...FI...Fq~ "
To obtain local formulas on the usual higher order anisotropic spaces we have to
restrict ourselves to even components of geometric objects by changing, formally, capital
indices (I, J, K , . . . ) into (i, j, k, a . . . . ) and s-derivation and s-commutation rules into
those for real number fields on usual manifolds.
4.2. Torsions and curvatures of d-connections
Let g(z) be a dvs-bundle endowed with N-connection and d-connection structures.

The torsion of a d-connection is introduced as
T ( X , Y ) = [X, DY} - [ X , Y } , X, YC~(h7"/).
The following invariant decomposition (by using h- and v-projections associated to N)

hold true:
T ( X , Y ) = T ( h X , hY) + T ( h X , vlY) + T ( v t X , hX) + T(vlX, VlY) +,..

+ T ( V p - l X , Vp-lY) + T(Vp-IX, vpY) + T(vpX, V p - l X )
+T(upX, vpY) + . . . + T(Vz_IX, v z -1Y) + T ( v z - I X, vzY)
+T(vz X, Vz-lX ) "~-T(vzX, vzY).
Taking into account the skewsupersymmetry of T and the equations
h[vpX, vpY} = 0 ..... vf[vpX, vpY} = 0 , f ~ p,
we can verify that the torsion of a d-connection is completely determined by the fol-
lowing ds-tensor fields:
h T ( h X , hY) = [ X ( D ( h ) h ) Y } - h[hX, hY} . . . . .

vt, T ( hX, hY) = - v p [ h X , hY} . . . . .
h T ( h X , VpY) = -Dy~(VP)hX - h[hX, vpY} . . . .
vpT(hX, vpY) = D(xh)vpY - vp [hX, vt, Y } . . . . .
vfT( vfX, v f Y ) = [ X ( D ( ' J ) v f ) Y } - v f [ v f X , v f Y } . . . . .
vpT( vfX, vfY) = --Vl, [ v f X , vfY} .....
vfT( ufX, Up Y) = -D(y ''P)vfx - vf [ vfX, vl, Y} . . . . .

vpT(vfX, vpY) = D(x''I) vp Y - vp [ vfX, Vp Y} . . . . . f < p,
vz-lT(vz-lX, Vx-lY) = [X(D(V:-')Vz -t )V} - v: --1 [Vz --I X, UZ- I Y } .....
vzT(Vz-lX, Vz-lY) = - - V z [V z --1 X , V z - I Y } ,

V z - l T ( V z - l X , vz Y) = - D CyV:)Vz-I X - Vz-l [ V z - l X , vz Y} . . . . .
v z T ( v z - l X , vz Y) = D (xV=-OVz Y - Vz [ Vz-l X, v z Y } .
where X , Y C ~ ( ~ ( z ) ) . In order to get the local form of the ds-tensor fields which
determine the torsion of the d-connection D F (the torsions of D F ) we use
8x.1 , axg I = R~ A) dy(a) ' ~X J ' ay(B) } - ay(B ) ay(A) '
where
RS> = (->'"i a N :
ax x ax J ,
and introduce the notations
h T ( ~ K, t~ ) = T t j K ~
t~XJ I'
8 8 ) -a, 8
ax.
8 8 ) pl 6
h T ( ay(a ) , aX J = J(a) ~ X I .....
t~ t~ p (a } t~
vpT( ayB,, ' 8X J ) = JBp ay( A} . . . . .
v,T( 8 8 ~,(A} & (24)

ayC~, ' OyB: ) = oBiCe cgy(a) "
NOW we can compute the local components of the torsions, introduced in (24), with
respect to a la-frame of a d-connection D F = ( L . . . . . L(p) . . . . . C . . . . . C(p ) . . . . ):
TIjK = L l j K -- (-- )IJKI LtK.t,

~,(jA)= R(A) jK 'pl(B > = CIj(B),
8 N ) A)
P(A)j(B)- cgy(B) L(A)(B)J'
s(A) (B)(C) = C(A) (B)(C) -- (--) I(B)(C)Ic(A) (C)(B). (25)
The even and odd components of the torsions (25) can be specified in an explicit form
by using decompositions of indices into even and odd parts (1 = (i, 7), J = (y, j ) . . . . ),
for instance,
Tijk = Lijk -- L'kj, Tijk = L ijk^ + L

^ A ^
Tijk = Lijk -- Likj . . . . .
and so on (see Refs. [75,76]).

Another important characteristic of a d-connection D F is its curvature:
R(X,Y)Z = D[xDr) - Dtx,r}Z,

S.I. V a c a r u / N u c l e a r Physics B 494 [ P M ] (1997) 5 9 0 - 6 5 6 611
where X, y Z E ~ ( / ~ ( z ) ) . F r o m the properties o f h- and v-projections it follows that
vpR(X,Y)hZ=O ..... hR(X,Y)vpZ=O, vfR(X,Y)vrZ=O, f 4= p, (26)
and
R ( X , Y ) Z = h R ( X , Y ) h Z + v I R ( X , Y ) v l Z + . . . + vzR(X,Y)vzZ,
where X, y Z E ~ ( / ~ ( z ) ) . Taking into account properties (26) and the equations
R( X, Y) = - ( - ) lXrIR( y X),
we prove that the curvature o f a d - c o n n e c t i o n D in the total space o f a vs-bundle/~(:)

is completely determined by the following ds-tensor fields:
{ l~(h)l)(h) _ FI(h)
R(hX, hY) hZ = ~ [ x ~r} ~[hX, hY}
_D(VO D(V:-t)
[hX, h } ' } - • " " [hX, h } ' } - Dlh~hV.})hZ,
R( hX, hY)vpZ = (1)(h) l)(h) --
" ~ [ X ~Y}
D (h)
[hX, hY}
- D (vl)
[hX, hY}
)vpZ - . . . - D [hX,
(vp-~) ~" Z
hY}'Vl '
(vp)
-- D[hX, hY})UP
Z'
R(vpX, h Y ) h Z [r~(Vp) r~(h) - - l ) ( h )

= I,z.,I[X~'y} ~ [ v v X , hY }
_ r~~[ v(vl) ahT_

t , X, h Y } : . . .
. . .. _ /3 (v"-l) ( " ) h )hZ '
[vpX,hY} - D[..x,_Y}
R(vfX, hY)vpZ :n(v/)r~(h) -- D (h)
= t~'[X ~Y) [vfX, hY}
- - U [ v / X , hY} ) V 1L . . . [ vfX, hY} I Up L ,
R(vfm, vpY)hZ --- [~F[}x( U f ) --r}

T~(OP) -- F)(Ul)
~[.:x,.~r} )hZ -
•••
_D (v:-|) _ D(v:)
[,,:x.,,,,r} ~[,,:x,,,.r} )hZ,
-- (r~(V/)l-~(vq) r-~(vl) x , '7
R( v f X, vqY)vpZ - ~~[ x ~y} - I J [ v / X , vqy) ) U l ~ - """
rdVe-l) ~. 7 -(vl,) )v~Z, (27)

--I"I[vfX, v q Y } ) u p - I L " -- D [ v f X t _ q Y } I
where
DIX ~(Y ) h=Fl(h)

~ ')
Fl(h)• ( ( ~ lh
X ~rY -- ~,--: ~ Y) ~ X
XY[Fl(h) l)(h)
'
DUOn(ve) ~.l)(h)Fl(Vp) _)Lxv,,rlo(rVe)o(xh),

IX ~ r } ~x ~r -(
D(.,,)n(h) = n(v,,)n(h) ", I.p x}'l n ( h ) n ( v e )
Ix ~r} ux ~r --(--: "--'~" ~ X .
D(Vf) 1"~(vv) = D ( v f ) FI(vP ) ~lvyX.erl n(vp) r~(V/)
[x ~r} ~x ~r - (-: "-'r " - ' x •
The local components o f the ds-tensor fields (27) are introduced as follows:
R( 6K, 6J)6H = RHIjKSI, R((3K, 6J)~(B) = Ri(BA¢.jK6(A) , (28)

1 .(a)
R ( 6 ( c ) , t~X)t~J -~ PJK(C)81, R (tS(c) , tSg) 6(B) = P(B).K(C)tS(A) '
R(6(c) , 8(B))Sj = Sj.(B)(C)¢~I, .I

R(tS(o ) , t~(c ) )t~(B ) -~
t"(A)
O(B).(C)(D) O(A)"
Putting the components of a d-connection D F = (L . . . . . L(p) . . . . . C . . . . . C(v ) . . . . ) in

(28), by a direct computation, we obtain the following locally adapted components of
the curvature (curvatures):
R H I j K = tSKLIHj -- (--) IKJIt~jLIHK

+LMnjLtMK _ ( _ ) Iral LMnrLtMj + CIH(A)R (A)JK,
R'(A)
(B).JK--_ t ~ .^L (a) (B)J-- ¢"~,--,IKJI6jL(A)(B)K
+L (C) (B)J L(A) (C)K -- ( -- ) IKJI L (C) (B)K "~ C (a) (B) (C) R (c) Jr,
PJ(K(A) = ~(a) L I j K -- CIj(A)IK 4- C I j ( B ) P (B) K(A),
P(B) (A)K(C) = t~(C) L(A) (B)g -- c(A) (B)(C)IK 4- c ( a ) (B)(D)p(D) K(C),

.I
Sj.(B)(C) = ~(c)CIj(n) -- (--)I(B)(C)ItS(B)CIj(c)
+ c ( H ) J(B) C I (H)(C) -- ( - - ) I ( B ) ( C ) I c ( H ) j ( c ) C I ( H ) ( B ) ,
S(B) (A) (C)(D) ----tS(D) C(A) (O)(C) -- (--)I(C)(O)I t~(c)C(A) (B)(O)
"~-c(E) (B)(c)C(A) (E)(D) -- (--)I(C)(D)Jc(E) (B)(D)C(A)(E)(C). (29)

The even and odd components of the curvatures (29) can be written out by splitting the
indices into even and odd parts, for instance,
Rh jk -- t~kLihj t~jLihk + LmhjLimk -- LmhkLimj + Cih(a)R(a) jk ,
Rh ij k -_t ~ k L t h"j - J - t ~ j L h ki ' 4 - L m hj L i mk -'}-~

f m hk^L i mj. + C ' h"( a ) R ( a ) j k . . . .
(we omit formulas for the even-odd components of the curvatures because we shall not
use them in this work).
4.3. Bianchi and Ricci identities
The torsions and curvatures of every linear connection D on a vs-bundle ~(z) satisfy
the following generalized Bianchi identities:
[ ( D x T ) ( Y , Z ) - R ( X , Y ) Z + T ( T ( X , Y ) , Z ) ] =0,
sc
[ ( D x R ) ( U , Y , Z ) + R ( T ( X , Y ) Z ) U ] =0, (30)
sc
where )--]sc means the respective supersymmetric cyclic sum over X, Y,Z and U. If D
is a d-connection, then by using (26) and
vp(DxR)(U,Y, hZ) = 0 , h(DxR(U,Y, vpZ) = 0 , vf(DxR)(U,Y, vpZ) = 0 ,

the identities (30) become
Z [ h ( D x T ) (Y,Z) - hR(X, Y) Z + hT(hT(X, Y), Z)

sc
+hT(v1T(X,Y),Z) +... + hT(vzT(X,Y),Z) ] = 0,
Z [vf(DxT) (yZ) - vfR(X,Y) Z

SC
+vfT( hT( X, Y), Z)+ Z vfT(vpT( X, Y), Z ) ] = 0,

p~f
Z [h(DxR) (U, Y,Z) + hR(hT(X, Y), Z)U
sc
+hR(vIT(X,Y),Z)U +.. +hR(VzT(X,Y),Z)U] =O,
Z [vf(DxR) (U, Y,Z) + vfR(hT(X, Y), Z)U
SC
÷ Z vfR(vpT(X,Y),Z)U] = 0. (31)
p>~f
In order to get the component form of these identities we insert correspondingly in
(31 ) these values of triples (X, y Z), (= (t3j, 6K, 6L), or (t~(o), 6(c), 6(B) ) ), and put
successively U = tSH and U = 6(a). Taking into account (24), (25) and (27), (28) we
obtain
Z [TIjKIH ÷ TMjKTIHM ÷ R(A)jKCIH(A) -- RjIKH] = O,

SC [L,K,J}
Z [ R(A) JKIH ÷ TMjK R(A) HM ÷ R(B)JK P(A) H(B) ] = O,
SCI L,K,J}
Cj(B)IK -- (--)IJKICIK(B)Ij -- TIjKI(B) ÷ cMj(B)TIKM
I
--(--) IJKICMK(B)T1jM ÷ TMjKCIM(B) ÷ P(D) j(B)CIK(D)

-(--)IKJIp(D)K(8)Ctj(D) + PjIK(B) -- (-- )IKJI PKIj(B) = 0,
p(A) j(B)IK -- (--)IKJI p(A) K(B)IJ -- R(A) jKI(B) ÷ cMj(B)R(A) KM
-- ( -- ) IgJI CMK(B) R(A) JM ÷ TMjK P (m) M(B) ÷ P (O) J(B) P (a) X(O)
~.(a) _
-- ( -- )IKJIp(D) K(B) P(A) J(D) -- R(°>JK S(A) B(D) ÷ "'(B).JK -- O,
CIj(B) I (C) -- (-- )I(B)(C)IcIj(c)_I_(O) -I- cMj(c)CIM(B)

.I
--(--)I(B)(C)IcMj(B)CIM(c) ÷ s(D) (B)(c)CIj(D) -- Sj.(B)(C) =0,
p (A) J(B)A_(C) -- ( -- ) I(B)(c)I p (a) J(C)J_(B)

÷ S (a) (B) (C)IJ ÷ cMj(c) P (A) M(B)
_ ( _ ) I(B) (C)IcMj(B) p (A) M(C) ÷ P (D) J(8) S(a) (C) (D)
--(--)I(C)(B)Ip(D) j(c)S(A) (B)(D) ÷ s(D) (B)(c)p(A) j(D)

+P(n) (A) J(C) - - ( - - ) [(C) (B)I P(c) (A) J(B) = O,
Z [ s(A) (B)(C).L(D)
SC[ (B),(C),(D) }
"q-s(F)(B)(C) S(A)(D)(F) -- S(B)(A)(c)(D) ] ~'0'
E [RKIHJIL -- TMHJRKILM -- R(A)HjP~'L(A) ] = O,

SC [H,J,L}
Z I-p'(A) "r'M D'(A) -- R(c) (A) ] = O,
t "'(D).HJIL -- I Hj,X(D).LM HJP(D) L(C)
SC l H,J,L}
•1 ILJI .;
PK.J(D)[L -- (--) PK.L(D)IJ + RKILJ-L(D)"}- CML(D) RKIJM
ILJI M I M .I
-(-) C j(o)Rr LM -- T JLP~.M(D)
+P(A) L(D)PKI.j(A) -- (--) IZJIn(a)
r n.l
j(D)I-K.L(A) -- R(A) JLS4.(A)(D) = O,
__ l?'(A)
P(c) (A)j(D)IL (--)ILJIp(c) (A)L(D)Ij + "'(C).LJI(D)
+CML(D)n(c) (m) JM -- ( - ) ItJI cgj<o> R<C) (a) LM

_TMjLP(c) (a) M(O) + p(e) L(O) P(c) (a) J<r)
--(--)ILJIp(F) j(D)P(c) (A)L(F) -- R(F) jLS(c) <A)r(O) = O,
•l _ (_)I(C)(D)ID.I I
PK.J<D)_I_(C) r K.J(C)-L(D) "~- Sx (D)(C)IJ
M .I
+C j(D)P~.M(C) -- (--)I(C)(D)IcMj(c)PKI.M(D)
+ p {A)j(c)SK.(D)(A
).1 __ (_)I(C)(D)I p(A) j(D)S.IK.(C)(A)
.1
-I-S (a) (C)(D)P~.j(A) = 0,
P(B) (A)J(D)-L(C) -- ( _ ) I(c)(O)l P(B)(A)J(C)_l_(D) q- S(B) (a) (C)(O)lJ
-[-cMj(D)P(B) (A) M(C) -- ( -- ) [(C)(D)I cMj(c) p<z)(A)M<D)
+p(F) J(C)S(B)(A) (D)(F) -- ( - ) I(C)(D)Ip(V) j(D)S(B) (A) (C)(F)
+s<F) (C) (D) P<a) (A) J(F) = O,

E .1 <A) .I
[ SK.(B)(C)±(D) -- S.<B)(C) S~(.(D)(A) ] = O,
SC[(B),(C),(D) }
Z i-.¢.(A) __ K,(E) v.(A)
tO(F).(B)(C)3_(D) O.(B)(C) O(F).(E)(A) ] = O,
SC{ (B),(C),(D) }
where ~-]SCI<8),(C).(D)} is the supersymmetric cyclic sum over indices (B), (C), (D).
As a consequence of a corresponding arrangement of (27) we obtain the Ricci
identities (for simplicity we establish them only for ds-vector fields, although they may
be written for every ds-tensor field):
S.L Vacaru/NuclearPhysics B 494 [PM] (1997) 590-656 615
D(h) r~(h)t'7
IX'-"Y} "'~ = R(hX, hY)hZ + Dl~)hy
, } hZ + ~ ~lhX.hr}
D O'D hZ,
f=l
z
~Y} " ~ = R(vpX, hY)hZ + D (hI

('>)/)(h)h7
DIx [vvX,hr} h Z -Jr-Z D IveX,
O's) hY)- hZ,
f=l
z
DI~')r~O'')~v} =R(vpX, vpY)hZ + Z D I,,,,x,,',,-,
(~'j) ,,~ hZ (32)
f=l
and
z
nL~IX
¢ h ) nt.Zy} 7 --
( h ) , , ~p~-- R(hX, hY)vpZ + D(h)
[hX,hy}VpZ + Z
(t,f) hY}OPZ'
D[hX,
f=l
z z
D(Vt) ,-.(h) .7
IX 1)y} Up/-, = R ( v f X , h Y ) v p Z + v " ,-.(,,,,) ~ X r~o,,,) hy}Vpl~,
2...aLIlvfX, hy}VpL + Z_ ~lvrX, . .7 TM
q=l q=l
D IX
0'~)~Y}
n(~'D v p Z = R(oqX, v f Y ) v p Z + ~ r~(t')
~{vrx, vry}vpZ. (33)
s=l
Putting X = Xt(u) gTx'

~ + X (A) (u) ~ and taking into account the local form of the h-
and v-covariant s-derivatives and (24), (25), (27), (28) we can, respectively, express
identities (32) and (33) in the following form:
x(A) IKIL -- (--) IKLIx(A) ILIK= R(B) (A) KLX(B} __ THKLX(A) IH

--R(B) KLX(A) ±(B) ,
XIIK±(D) -- (-- )]K(D)IxI.L(D)[ x = Pg.K(D)X H -- CHK(D)XIIH
--P(A) K(D)XI±(A),
XI±(B)±(C) -- (--)I(B)(C)IxIi(c)&(B) = S'IH.(B)(C)XH -- s(A) (B)(C) XIi(A)
and
x(A) IKIL -- (--) IKLIX(A) ILIK= R{B) (A)KLX(B) -- THKLX(A) IH

-R(n) KLX(A) ± ( B ) ,
x(A) I g ± ( n ) -- ( - - ) I ( n ) g l x ( a ) ±(B)IK = P(B)(a)KcXC _ cHK(B)X(A)IH
_p{D) K(B) X(A) I(D),
x(A) ±(B)A_(C) -- (--)I(C)(B)I x(A) ±{C)±(B) =S(D) (A)(B)(c) X(D)
--s(D) (B)(C) X(A) ±(D).
We note that the above-presented formulas [ 110] generalize for higher order anisotropy
the similar ones for locally anisotropic superspaces [ 106].
4.4. Cartan structure equations in dvs-bundles
Let us consider a ds-tensor field on ~(z):
t = t~a)~l®~ (a).
The d-connection 1-forms w / and co(B

-(a)) are introduced as
D t = ( Dtla) )¢~I®t~ (A)
with
I J -- t°(a)'(B)
Dt~a) = dt~A) + wjt(A) ~.(B),I . = tla)ljdXJ + t~A)J_(e)~y(O)
For the d-connection 1-forms of a d-connection D on {(z) defined by w~ and U,(n

)~'cA) the
following structure equations hold:
d ( d 1) - d n A tolH = --$'2, d(t~ (a)) - t~(B) i ~(B)~(A)----_j~(A),

1 =-OJ,I r/%(A) ~(C) f ~~,(A)
d w ~ - toy A w H ~ ( B ) - ~(B) ( c ) = __t~(A)
"~(~)'
in which the torsion 2-forms s2t and ~(A) are given respectively by the formulas
f2 t -~ ~,t
1 w!
JK"a J i dx 1 ,-.I j(C)U, J i 8(c) ,
+ ~t..
~(A) ~ 2 R ( A ) j K d J /\dK--[- ~ e ( A ) j ( c ) d J / \ ~ ( c ) --[- Is(A)(B)(C)~(B) /\~(C) '
and
$21 = 1 o 1 .4X d H 1 o.1 .~x A 6 (c) 1 c.t ~(a) 8(C)

-~xj KHU A + -~Ij.K(C)t~ "~- ~,j.K(C)t, A ,
(o) = ½Ri !K. A d" l p (B)ia)K(c)dK

+ -2 A (cl + 1o
2°(n) (C)(O) "
We have defined the exterior product on s-space to satisfy the property
8 (~) A 8 (t~) = - ( - ) I(-) (~)1 t3(~) A 6 ('~) .
4.5. Metrics in dvs-bundles
The base /9/ of dvs-bundle g(z) is considered to be a connected and paracompact

s-manifold.
A metric structure on the total space/~(z) of a dvs-bundle g(z) is a supersymmetric,
second-order, covariant s-tensor field
G = G(,~)(#)c~ ('~) ® 0 (8)
which in every point u E g(z) is given by the non-degenerate supersymmetric matrix

G(,~)(/3) = G(O(~), 0(~)) (with non-vanishing superdeterminant, sdetG 4= 0).
The metric and N-connection structures on ~(z) are compatible if there the following
conditions are met:
G(~I,a(A)) =0, G(~af,Cgmj,)= 0 , Z >~p > f >~ 1,
or, in consequence,
~,(B) i.
GI(A) -- iv I rt(a)(B) = O, GA/A,, -- NB~hA,,B,, = 0, (34)
where
GI(A) = G(OI, O(A ) ), GA/At, = G(c~af, ~ga,, ).
From (34) it follows that
N~ B) = h(B)(A)GI(A), NA~ = hA"BpGA/B, . . . . . .
where matrices h (a)(n) , h aeBe . . . . are respectively s-inverse to matrices
h(a)(B) = G( a(a), a(B) ) , hat, Bp = G( c3ap, aBp ).
So, in this case, the coefficients of N-connection are uniquely determined by the com-
ponents of the metric on ~(z).
A metric compatible with an N-connection on ~(z) is written in an irreducible form
as
G(X,Y) =G(hX, hY)+G(vlX, v~Y)+...+G(v:X,v:Y), X, Y E ~ ( ~ ( z ) ) ,
and looks locally as follows:
G = g , ~ ( u ) 6 a • 6 ~ = g t j d t ® d J + h(m)(B)~ (a) Q ~(B)

= g t j d t Q d J + haiBt~a~ • ~Bl + ha2B2~a2 (~ ~B2 q_ . . . _[_ hazB~a: ~ ~B:. (35)
A d-connection D on ~(z) is a metric, or compatible with a metric G, if the conditions
D(,~)G(#)(~,) = 0
are satisfied.
A d-connection D on {(z) provided with a metric G is a metric d-connection if and
only if
D(xh)(hG) =0, D(xh)(vpG) =0, o- x(vv) ,hG)

( =0, D ~ , D (vpG) = 0 (36)
for every f , p = 1,2 . . . . . z, and X C ~ ( { ( z ) ) . Conditions (36) are written in locally

adapted form as
glJiK =O, glj_l_(A) = 0 , h(A)(B)IK =O, h(A)(B)I(C ) = 0 .
In every dvs-bundle provided with compatible N-connection and metric structures

there exists a metric d-connection (called the canonical d-connection associated to G)
depending only on components of G-metric and N-connection. Its local coefficients

t"r
" ~ = t~'~:r,
~I ~(A)
"-'<8>K't~l
"~J<C>' t.%(A)
"-'<8)<C>) are as follows:
LI K = -~g
1 IH ( 6 r g n g + t3ggnK -- 8Hggx),
L(A) ~ xr(A) l h(a)(c ) [ t~(r) h(B)(C) -- ( t~(B) N(D) ) h(D)(C)

<B)K = O(B) IVK "~
<D)
--(~(c)NK )h(D)(B)],
"I 1 _IK6 _
CJ(C) = "~g (C)gJK,
~(A>
(B>(C> = 2 h(a>(D> (6(c)h(o)(8> + ~<B)h(o>(c> - 6(o) h(B>(c>. (37)
We emphasize that, in general, the torsion of the CF-connection (37) does not vanish.
It should be noted here that on dvs-bundles provided with an N-connection and a
d-connection and a metric, actually a multiconnection ds-structure is defined, i.e. we
can use in an equivalent geometric manner different types of d-connections with various
properties. For example, for modeling some physical processes we can use an extension
of the Berwald d-connection
trl ~ ~r(A) t'. r ( A )
BF = ~ Jr, o(B>lvr , v , ~(B)(C> )' (38)
t~(A)
where L t j K = LIK and c<A) (B)(C) = ~(B)(C)' which is an hv-metric, i.e. it satisfies the
conditions
D(Xh) h G = 0 . . . . . D ~ '') opG = 0 . . . . . D ~ '~) vzG = 0
for every X E ~ ( { ( z > ), or in locally adapted coordinates,
glglK = O, h(A)(B)I(C ) = O.
We can also introduce the Levi-Civita connection
t<o>l
(fl> <T> ,
= -2 G< ><#>(0(#> G (¢) (,> + O(r) G (,) <#) - 0<,) G <#>(,)),
constructed as in the Riemann geometry from the components of the metric G(~)<#>
o ,9 0
by using the partial derivations a<~) = a,<-> - (a--V, a~7a>.), which is a metric but not a
d-connection.
In our further considerations we shall heavily use the Christoffel d-symbols defined
similarly as the components of the Levi-Civita connection but by using la-partial deriva-
tions,
r,<'~>
(#>(~,> = 1G(,~)(~) ( t3<#)G(.:)(~,> + 6(~,) G(r)<#> - 6(T) G(#)(~,> ) , (39)
having the components
C [" = ( LIjK, O, O, C (A) (B)(C) ) ,

where the coefficients L t j x and C <a) <B)(c) must be computed as in formulas ( 3 8 ) .

We can express an arbitrary d-connection as a deformation of the background d-
connection ( 3 8 ) :
_ ~(,,)
F(") (~)(~') - --.(~)(r) + P ( ' ) (/~)<:') ' (40)
where P<")(~)(:,) is called the deformation ds-tensor. Putting splitting (40) into (25)
and (29) we can express the torsion T<")(~)<r) and the c u r v a t u r e R(fl)(a)(y)(8) of a
d-connection F(")(fl)<y) as the respective deformations of the torsion /~l<r) and the
torsion ~ ( " ) for connection ? ( " ) •
_/~(,,> ..
T('*) (~) (r) - . <B)(:,) + T.((~)(r)
and
R(fl) (") (~)(8) =/~'(") ~,"(,')

<fl)'(:')@ + "'(~).(~)<a) '
where
/~<")
<fl)(~,) = -~(")
( ~ ) (~,) _ (_~l(~)<y)l/-<")
" <~,)<~) + w (,,) (Y)(~)'
ip<,,)
<#)(:,) = ~<a)
-(B) (r) - ( - " ~ [<B)<T>Ip<a)
(~,)<~)'
and
).<a)
(B)'(~')(,~) -- 6 (8)~(B)(y)
~(a) - ( - ) I(y)@l~°<~')~(B)(8)
r,(a)
+~(~) p<") _)l<~)<@p<~) p<") p<") ,~ <w)

1(B)(:,)~(~o)(8) - ( (B)<a) (~)(~,) + ~(B)<~)" (Y)<~)'
(#)'<y)(~) = / ) @ p ( a )
R.(,,) (#)(~') _ ( _ ) I<r)<8)l D<:,) p<,~ ) (B)(,~)
+ P (~) (B) (r) P (a) <~) (8) - ( - ) I(:') @1 p <~0)(B) (8) P <'~)<~) (:,)
+ P ( " ) (B)(~) w(~') (:')@'
the n o n - h o l o n o m y coefficients w(")<,)<~) are defined as
[~(,,), 8<fl)} = ~<,,)6(fl) - (-)1<")(/~>16</~)8<,,) = w (T>(,~)<#)6(T).
We emphasize that if from a geometric point of view all considered d-connections are
"equal in rights", the construction of physical models on la-spaces requires an explicit
fixing of the type of d-connection and metric structures.
5. H i g h e r order tangent S-bundles
The aim of this section is to present a study of supersymmetric extensions from A4

to T M and Osc(Z)Al and to consider the corresponding prolongations of Riemann and
generalized Finsler structures (on classical and new approaches to Finsler geometry,
its generalizations and applications in physics see, for example, Refs. [38,30,90,74-
76,7,8,78,5,6,22,14] ).
The basic results on dvs-bundles ~(z> provided with N-connection, d-connection and
metric structures, presented in the previous section, can be correspondingly adapted to
the osculator s-bundle (OscZ~l, 1r, 191). In this case the dimension of the base space and
typical higher orders fibre coincides and we shall not distinguish between the indices of
geometrical objects.
The coefficients of a d-connection D F ( N) = ( L~ta, C (1l ) m . . . . . C[z)m) in OscZi91,
with respect to a la-base are introduced as to satisfy
3 ~ 5 ,, S
O h 8y[f) - L ~ -8-y ~ ) ' O_~,y,,,, 8y[f) - Cg)H~y_ ~)' (41)
( f = 0 , 1 . . . . . Z ; p = l . . . . . Z, andY[0)= xl).
A metric structure on Osc z M is a ds-tensor s-symmetric field
g l j ( U( z ) ) = gtj( x, Y(1), Y(z) . . . . . Y(z ) )

of type (0, 2), sranklgijl = (n, m). The N-lift of Sasaki type of gtJ (see (35)) defines
a global Riemannian s-structure (if M is an s-differentiable, paracompact s-manifold):
G = glj(U(z))dx I ® dx J + gij(U(z))dy[I ) ® dy{1 ) + . . . + gtj(U(z))dy[z ) ® dy{z ).

(42)
The condition of compatibility of a d-connection (41) with metric (42) is expressed as
DxG =0, VX E ~(OscZif4),
or, by using d-covariant partial derivations ](p) defined by (L~M, C(/I)JM, . . . . C[z)sm),
glJlM---o, glJl~,,~M = 0 (p = 1 . . . . . Z ).
An example of a connection compatible with the metric is given by the Christoffel
d-symbols (see (39))"
1 MK ( SgKJ ~3gm 6glj )

L M = ~g \--O--~i-xl + ~Ox c~xK ~ '
C(P)H=2 g \Oy(p)+Oy~p) Oy~,)]' p = l , 2 . . . . . z.
5.1. Supersymmetric extensions of Finsler spaces
We start our considerations with the ts-bundle T/f/. An s-vector X E ~ ( T ~ / ) is

decomposed with respect to la-bundles as
X = X ( l l ) l ~ l --~ Y(u)10I,
S.l. Vacaru/Nuclear Physics B 494 [PM] (1997) 590-656 621
where u = u '~ = (xl,y J) are local coordinates. The s-tangent structures (16) are trans-
formed into a global map
J : _=(T&/) ~ _=(T~7/)
which does not depend on the N-connection structure:

0
J ( 6 - ~ ) - ayt
and
c~
J(aT) =0.
This endomorphism is called the natural (or canonical) almost tangent structure on TM;
it has the properties
( 1) j2 = 0, (2) Im J = Ker J = V/'/~/
and (3) the Nijenhuis s-tensor,
N j ( X , Y ) = [JX, JY} - J [ J X , Y} - J [ X , JY] (X,Y C ~(TN))

identically vanishes, i.e. the natural almost tangent structure J on T.~/is integrable.
A generalized Lagrange superspace, GLS-space, is a pair GL "'m = (l(4, glJ(X, y ) ) ,
where g l j ( x , y ) is a ds-tensor field on T ~ = T l g l - { 0 } , s-symmetric of superrank
(n,m).
We call glJ the fundamental ds-tensor, or metric ds-tensor, of GLS-space.
There exists a unique d-connection C F ( N ) which is compatible with gH(U) and
has vanishing torsions TIjK and SIjK (see formulas (25) rewritten for ts-bundles).
This connection, depending only o n glj(U) and Ntj(u), is called the canonical metric
d-connection of GLS-space. It has the coefficients
1 /H
LIjK = ~g (~JgHK + ~HgJK -- ~HgJK),
1 In
CIjx = ~g (OJgHK -~" OHgJK -- cgHgJX).
There is a unique normal d-connection DF(N) = (L!jK, CIjK) which is a metric and
has a priori given torsions TtjK and SIjK. The coefficients of DF(N) are
1 IH. ,'rR
L!jK = LIjK -- "~g tgJR" HK -1- gKRTRHj -- gHRTRKj),
1 m . I,gJR S RHK q- gKRSRHj -- gHRSRKj),

C.l K = CIjK -- ~g
where LIjK and CIjK are the same as tbr the CF(N)-connection (41).
The Lagrange spaces were introduced [61 ] in order to geometrize the concept of a
Lagrangian in mechanics (the Lagrange geometry is studied in detail in Refs. [75,76] ).
For s-spaces we present the following generalization:
A Lagrange s-space, LS-space, L n,m = (,~7/,gH), is defined as a particular case of

GLS-space when the ds-metric on M can be expressed as
1 O2L
g t j ( u ) = 2 Oylay J' (43)
where L : T h T / ~ A, is an s-differentiable function called an s-Lagrangian on AT/.

Now we consider the supersymmetric extension of Finsler space: A Finsler s-metric
on ,Q is a function Fs : T1(4 ~ A having the properties
(1) The restriction of Fs to T ~ = TM \ {0} is of the class G ° and F is only
supersmooth on the image of the null cross section in the ts-bundle to ,Q.
(2) The restriction of F to T,(/is positively homogeneous of degree 1 with respect to
( y t ) , i.e. F ( x , Ay) = A F ( x , y ) , where A is a real positive number.
(3) The restriction of F to the even subspace of T M is a positive function.
(4) The quadratic form on A "'m with the coefficients
1 02F 2
glj(tt) - 2 OylOy J
defined on T79/is non-degenerate.

A pair F n,m = (291, F ) which consists of a supersmooth s-manifold/9/and a Finsler
s-metric is called a Finsler superspace, FS-space.
It is obvious that FS-spaces form a particular class of LS-spaces with s-Lagrangian
L = F 2 and a particular class of GLS-spaces with metrics of the type (58).
For an FS-space we can introduce the supersymmetric variant of a non-linear Cartan
connection [ 30,90] :
0 .t
U;(x,y) =
where
G. 1 = "~g
1 _.is (\ ~a2e
y K -- OX
aeJ ) , 8(tt) = g l j ( u ) y l y J,
and g , lJ is inverse to g~j(u) = la2~/OylOyJ. In this case the coefficients of the canon-
ical metric d-connection (25) gives the supersymmetric variants of the coefficients of
the Cartan connection of Finsler spaces. A similar remark applies to the Lagrange
superspaces.
5.2. Higher order prolongations o f Riemann, Finsler and Lagrange s-spaces
The geometric constructions on T,~ from the previous subsection have corresponding
generalizations to the Osc~Z)~l s-bundle. The basic idea is similar to that used for
prolongations of geometric structures (see Ref. [79] for prolongations on the tangent
bundle). Having defined a metric structure g H ( x ) on an s-manifold/~ we can extend it
to the O s c Z M s-bundle by considering g l j ( U ( z ) ) = g i g ( x ) in (42). Miron and Atanasiu
[77] solved the problem of prolongations of Finsler and Lagrange structures on the
osculator bundle. In this subsection we shall analyze supersymmetric extensions of
Finsler and Lagrange structures as well as present a brief introduction into the geometry
of higher order Lagrange s-spaces.
Let F n'm = (If4, F ) be an FS-space with the fundamental function Fs : Tlf4 ~ A on
M. A prolongation of F on OscZlfl is given by a map
( F o 7rl)(uCz)) = F ( u ¢ l ) )
and the corresponding fundamental tensor

1 02F 2
gIJ(U(1)) -- 20y[l~OY~l ~ '
for which
(gIJ o 7r~ ) (U~z)) = glj(U~l)).
So, glg(u~l)) is a ds-tensor on O s k z M = OscZl(4/{O} = {(u~z)) E OscZlO, srankly[l)l =

1}.
The Christoffel d-symbols
TM(u(j) ) = 2gl MK (U(I)) \(agK/(U(I))0X

J q- OgJK(U(1))aX
I aglj(U(l))
K
define the Cartan non-linear connection [29]
i 1 0 . I K)y~)). (44)
G(N)J -- 20y-~l ) tYKMY~1
The dual coefficients for the N-connection (21) are recurrently computed by using (44)
and the operator
/ O 1 O 1 O
F=y(,)~x/+ 2 Y(2)8y[l----~)+... + z Y ( z ) 8 y [ L i ) ,
t 1 i i K
g ( 2 ) g = ~ [ FG(N)J + G(N)KM(1)g],
i 1 1 K
M~z)g = - [ F M ( z _ l ) g + G ( N ) K M ( z - I ) j ] •
Z
The prolongations of FS-spaces can be generalized for Lagrange s-spaces (on La-
grange spaces and their higher order extensions see Refs. [75-77] and on supersymmet-
ric extensions of Finsler geometry see Ref. [ 106] ). Let L "'m = (lf4, gig) be a Lagrange
s-space. The Lagrangian L : T M --* A can be extended on OscZ~/l by using maps of
the Lagrangian, (L o 7r~)(U(z)) = L (u(1)), and, as a consequence, of the fundamental

tensor (43), (gig o 7r~)(U(z)) = gH (u(l)).
5.3. Higher order Lagrange s-spaces
We introduce the notion of Lagrangian of z-order on a differentiable s-manifold M as

a map L z : OscZf4 ~ A. In order to have concordance with the definitions proposed by
[77] we require the even part of the fundamental ds-tensor to be of constant signature.
Here we also note that the questions to be considered in this subsection, being an
supersymmetric approach, are connected with the problem of elaboration of the so-
called higher order analytic mechanics (see, for instance, Refs. [34,70,69,94] ).
A Lagrangian s-differentiable of order z (z = 1,2,3 . . . . ) on an s-differentiable s-
manifold h4 is an application L (z) : OscZ~l --+ A, s-differentiable on O s k Z M and smooth
in the points of OscZffl where Y[I) = 0.
It is obvious that
1 02L (z)
g l y ( x , y ( l ) . . . . . Y(z)) - -~ 1 g
OY(z )OY(z )
is a ds-tensor field because with respect to the coordinate transforms (3) we have the
transforms
Ktl'KJ'g,,j, = g,j.
A Lagrangian L is regular if srank[gt~[ = (n, m).

A Lagrange s-space of z-order is a pair L (z,n,m) = (ffl, L(Z)), where L (z) is an s-
differentiable regular Lagrangian of z-order, and with ds-tensor gtg being of constant
signature on the even part of the basic s-manifold.
For details on non-supersymmetric osculator bundles we refer to Ref. [77].
6. Superstrings in higher order anisotropic S-spaces
This section considers the basic formalism for superstrings in dvs-bundles. We shall
begin our study with non-supersymmetric two-dimensional higher order anisotropic
sigma models. Then we shall analyze supersymmetric extensions and locally anisotropic
generalizations of the Green-Schwarz action.
6.1. Two-dimensional higher order anisotropic sigma s-models
Let ~(z) be a higher order anisotropic space (not superspace) with coordinates ~(~) =
"u(z) = (yi = -£(z), f(a) = "y(z) ) = (xi, y at . . . . . ya,,, . . . . ya= ), d-metric g(a)(~); we
use the notation (N2, yn~) for a two-dimensional world sheet with metric Y~e(z ~) of
signature ( + , - ) and local coordinates z = z u, where/~, ~,/i . . . . . 1,2.
The action of a bosonic string in a dv-bundle ~(z) is postulated as
1,~= l (45)
which defines the so called two-dimensional sigma model (o--model) with d-metric
g(,~)(B) in higher order anisotropic spaces (dv-bundles with N-connection) and A being
constant. A detailed study of different modifications of the model (45) is also given in
Refs. [ 107,108]. Here we shall consider a supersymmetric generalization of the string
action by applying the techniques of two-dimensional ( 1,1 )-supersymmetry by changing
the scalar fields u(z) into N = 1 real s-fields (without constraints; for locally isotropic
constructions see Refs. [11,23,36,42,51,83] ) ~(z, 0) which are polynoms with respect
to the Majorana anticommuting spinor coordinate 0:
7/(<,>( z, O) = ff(<~) ( z ) + -OA (`~) ( z ) + 2-OOF('~) ( z ) . (46)
We adopt the following conventions and notations with respect to the dimension-2
Dirac matrices 3,a and the matrix of charge conjugation C:
{Ya, 3,e} = 2r/a~l, tr(3,e3,a) = 2r/ea (Ys) 2 = 1,

~'=3,g0/~ , T --I = - y ~ ,
C3,aC C = -C r = C -I .
For Majorana spinors 0a, X~ . . . . the following relations hold:
O = 0+3, °, 0 a = Ca~O~,
Let us introduce in the two-dimensional (1,1)-superspace the covariant derivations
D~ = --a
a-On _ i( ffO)~,
satisfying the algebra
{D~, Da} = 2i(0"C)~ _= 2iO~a
and the integration measure on anticommuting variables with the properties
The (1,1)-supersymmetric generalization of (45) in terms of s-fields (46) is written as
1II
I<,s = 8iTrod d2z d20{g('~)(s3)(fi') - b(,~)(t~)(~')}Dfi'('~) ( 1 + ys)Dfi"(t~),
where A2 = 2~-d which is a higher order anisotropic generalization of the Curtright-

Zachos [36] non-linear sigma model. Integrating over 0, and excluding the auxiliary
fields F (<~) according to their algebraic equations we obtain from the last expression
I(rs = ~ d2z "g(a)(B)
..~. ^ ^
+ee'b(~)(a)Oeu(~)Oiu(/3) + ig(~)(#)~(~) .r e '-'~
~(-)A(B)
1 ~(_) ~(,~) }
~,a(;3)(,~)(~,)(8) ( 1 + Y5) ,@,)~(/3) ( 1 + Y5) ,~(8) , (47)
where
~g+)A(B) = tr 8(,e)O.. ~(B)

(~t) e -t-/-'(a)(y)Og~U , , -I- n(fl)
~y~ (a) (3')8.---0i)](~'}1A
eo J (y) ,
F(,~)(r) are the Christoffel d-symbol (39) on the dr-bundle. In order to have motions
of la-strings compatible with the N-connection structure we consider [ 107,108] the
following relations between ds-tensor b(,)(#), strength B(,~)(#) (r) = 81(,~)b(/~)(:,)} and
torsion T~)(y) (see (29)):
^ ~(8)
8(,~)v(~)6, ) = g(,~)(8) (/3)6,), (48)
with s-integrability conditions

af ^
12~pa~6~b(~)(r) =8[a,,T~I(~)(r) (f < p,s;p,s=O,l ..... z), (49)
af
where g2~a, are the coefficients of the N-connection curvature ( 11 ). In this case we can
express B(~)(#)(y) = T[(~)(~)(r)}. Conditions (48) and (49) define a model of higher
order anisotropic superstrings when the o--modes s-antisymmetric strength is introduced
from the higher order anisotropic background torsion. More general constructions are
possible by using normal coordinates locally adapted to both N-connection and torsion
structures on background s-spaces. For simplicity, we omit such considerations in this
work.
The ds-tensor R(,~)(~)(r)(8) from (47) denotes the curvature with torsion B:
~,(+)
(f~) (,~)(~,)(8) [~(-4-)] = h'(f~)(,~)6,)(8) =F D6,)B(o,)(~)(8 ) ~ D(8)B(,~)(~)(r)
+B(r)(,~)(y)B(8)(B)- (~)(,~)(8) {y)(B)'
where h'(#) (,~)(r)(8) is the curvature of the torsionless Christoffel d-symbols (39),
(#)(:,) - (~) (#)(y)'

A
~(±) B (+)
1 8
In order to define a locally supersymmetric generalization of the model (47) we con-

sider a supersymmetric calculus of the set of ( !, 1 )-multiplets of higher order anisotropic
matter (ff(~), ~(~) = ,~('~) ) with the multiplet of ( 1,1 )-supergravity (eT,~~be) in two di-
mensions (see Refs. [62-64] for locally isotropic constructions).
The global supersymmetric variant of action (47), for 2¢rd = 1, is written as
10[~o,A] = ~1 f a2z [~(,~)(8)oL~ ('~) O~u~8)

= + b(,~)(8)~.. ~O~(a
= ) Oa.~(8)
_
+i~(,~) (8) ~('~) Y~(D£A) (8) + iJ3(,~)(8) (r) ~('~) TsY~(8£fi'(8) ) a(r)
+ 1 R(8) (,,><r)(~) (X<") a<r> ) ( x<8>a<~>)
1 ~ ~, eq-(a)
---~D(e)n(,~)(8)(T)ta yS)t(8) ) (A(~) A(*))
1g g(~) , -r('~)- ~(8)
o(,~)(8)(r )n(r)(,~ ) t a "ySA ) (~(r)75A(~))], (50)
where the covariant derivation D(~) is defined by the torsionless Christoffel d-symbols.
The action (50) is invariant under global s-transforms with the Majorana spinor
parameter e:
e ~(") -(8)
(8)(~,) r5 ).
Defining Majorana-Weyl spinors A(,") (MW spinors) instead of ,~('~) we can rewrite
the action (50) in a more convenient form:
'/
.~ ,p-,+,,a+~~(8) +/~(,,>(8)~ <">( ~ - a _ ) < n
7<,,> ~tJ
+tg(a)(8)a+
1 ~+
+ ~ R (8) (,,) (r) (~) (~(') Y5A(8)) (~(r) Y5,~(~)) ], (50a)
with s-symmetric transformation law
A,¢(-) = A+~,{o'> + A_¢ (°') = ~+a(_'~) + ~_a.(+'~) ,

AA(, °') = -i(ffu ('~))e~: _ ~(o)(+)~(8)
-(8)(~,) "'+ A + ~p(~')
For simplicity, we shall omit the "hats" on geometrical objects if ambiguities connected
with indices for manifolds and supermanifolds do not arise.
In string theories one considers variations of actions of type (50) with respect to
s-symmetric transformation laws and decompositions with respect to powers of ,~. Co-
efficients proportional to ~5 vanish because they do not contain derivations of the
e( z )-parameters. In order to compensate the terms proportional to A and A? one adds
the so-called Noether term
/(N) = -~ 2g(,~)(8)(O~q~('~))(-~(8)y~/~/h 2)
i
3 B(~)(8)(:') (~(") YsY°-A ('°)) (~(~') ~ )
i
- ~B(,,)(3)(W (-X(')y£a(3))(-~6')Y5~: )],
where C~ is the higher order anisotropic generalization of the Majorana gravitino with
the s-symmetric transformation law
AC~ = -aLe + • • •
From the standard variation, but locally adapted to the N-connection, of the I (N)
with a next covariantization (with respect to (e~, Ce)) of the theory. In the result (it is
convenient to use MW spinors) we introduce this action
I = 1 2fjd 2 z e [raag(a)(3)Oau(a)Ouu (3) + e-lem~b(a)(3)Oau(a)auu(3)
.
tg(a) (3) -r<~>
A+ (/9+a+)<m Jr- ig(a)(3) A(-'~)( ~ - A_ ) (3)
+1~+4(3) (,~)(7) (a) (A('~)75 A(3> ) (X(Y)Ys A(a) ) ]
+ [g(-)(3)(2a"
u<~>+ ¢.a<+"> + ¢~ a<£>)
x (]T>{y + e-'dgC. + - e-':9C.)
2i ,.v(~)3'a-r<3)..-r(W-
+~-B(,~)<3)(Wtt't+ A+ ) t a + ~/'a). - (](~)ya](_#))(](_r)Ca)}
] _ (51)
where e = det [e.[.I, for which the higher order anisotropic laws of supersymmetric
transforms read
Ae7 = 2my~e, ACe = -Dee,

A~<-> = A+~o<">+ A_~ <">= ~+ a<_~> + ~_ a<+").
AA<~~> = -i(gu(")+ {](+">Ca+](-~)Ca}Ya)eT- ~(-)(+),(3)
--(3)(r) "'+ A+ ~(r)" (52)
Restricting our considerations in (51) and (52) only to a(+`~> spinors and e_ param-
eters, when e+, a(_'~) = 0, we obtain the action for the heterotic higher order anisotropic
string on background (g(,,H3), b<,,)(3) ) with ( 1,0)-local supersymmetry (Ca --+ Ca(-) ).
This action can be interpreted as the "minimal" interaction of the higher order anisotropic
( 1,0)-matter (q~("), A(+") ) with ( 1,0)-supergravity (eT, C~ ).
6.2. Locally anisotropic heterotic strings
As an illustration of the application of s-field methods in locally anisotropic s-spaces

we shall construct the action for a model of a higher order anisotropic s-string.
The ( 1,0)-superspaces can be parametrized by two Bose coordinates (z ~, z =) and one
Fermi coordinate 0 + ; / / = (z ~, z =, 0+). One represents vector indices as ( + + , - - ) =
(:~, =) taking into account that by ( + , - ) the spiralities + 1 / 2 are denoted.
S.I. Vacaru/NuclearPhysicsB 494 [PM](1997)590-656 629
The standard derivations

3
O~ = {O+,O~,3=}, D+ = ~ +i0+3+,
in the flat (1,0)-superspace [91] satisfy the algebra
{D+,D+} =2i3~, c9~3==D, [a,_,,D+] = [3a,3b] = 0 ,

and the s-space integration measure reads
/ dO+ - 0~ + , d3/i - = dZzdO+.
In the flat ( 1,0)-superspace one defines scalar and spinor s-fields
q~(z,O) =A(z)+O+,t+(z), ~p_(z,O)=rl_(z)+O+F(z)

and action
l= f d3iiL= f d2z(D+L)lo_-o
with a charged Lagrangian, L = L_, in order to have Lorentz invariance.

The ( 1,0)-multiplet of supergravity is described by a set of covariant derivations
VA = + = + an, (53)
where E L)/is an s-vielbein and w~/

.(M) is the Lorentz connection with L-generator M:
[M,a+] = +1~+.
z
The covariant constraints in s-space (for (1,0)-supergravity [24,40]) are given by
the relations
{V+, V+} = 2 i V , , [V+, V=] = -2iX+M, (54)

[V+,V~t] =0, [V~t,V=] = - X + V + - R(2)M,
where R (2~ = 2V+X + and X + defines the covariant strength of (I,0)-supergravity in

s-space.
= ~ ~ +
As a consequence of (54), only E+, E__~,E+, E+ and E2_--are independent s-fields; the
rest of the components of the vielbein and connection can be expressed through them.
The conditions of covariance of the derivations (54) lead to the following transformation
laws with respect to the d-coordinate and local Lorentz transforms with the corresponding
parameters K 0-- and At:
VA = eK V;~ e-K, K = K~Do___+ ArM;

for vielbeins we have
/~E~ = -V;~K 0--+ K~2DyE0 + EOKY[Dr., DO}0-] + At[ M, E%],

from which it follows that

" 1
/NE~ = KO-Do__E~++ 2iE +K + - E~DfIK ~ + -~ATE+.
It is convenient to use the s-symmetric gauge (when E~+ = 0),

i -1
K + = - - ~ (E +) V + K ~,
and to introduce the Lorentz invariant scalar s-field S and Lorentz compensator L
satisfying correspondingly the conditions
E+ (E~=-)1/2 = e - s ' E+(E:__-)-I/2 = e L.
In a Lorentz invariant theory we can always choose the gauge L = 0; in this gauge
the s-conformal transforms are accompanied by a corresponding compensating Lorentz
transform with parameter
A t = ( E + ) - 1 2 7 + K + - - ~ (1 E _-=)-I~=K==~(V~K
1
~-'~=K =)+...
The solution of the constraints (54) in the s-symmetric gauge E~+ = L = 0 and in the
linear approximation is [40]
V+=(I-S) D+ + H + O = - ( D + S + O = H + ) M ,
V~=(1 - S ) O ¢ + i FL1(a=H+) + (D+S)I D+ ..I
- i [ D+H+ ]O= - ( a~S - iD+a=H+)M,

i
V= = ( 1 - S)3= - ~ [ (D+H~)]D+ + H~3¢ + (?=S + 3$H¢=)M,
.,y+ = 2i [D+(O$H~= + 20=S) + O~I-l=+] + ... ,
where the s-fields (H+, H i ) and S are prepotentials of the system. These s-potentials
have to be used in the quantum field theory.
The linearized expression for the s-field density E -1 = Ber(E~) is computed as
E -1 = s d e t ( E O) = e3S/2[ 1 + iH~(D+I-I=+ + H+O=H+)] -l
The action for the heterotic string in higher order anisotropic (1,0)-superspace ac-
counting for the background of massless modes of locally anisotropic graviton, antisym-
metric d-tensor, dilaton and gauge bosons is introduced in a manner similar to locally
isotropic models [47,55,63,64] but with the corresponding extension to the distinguished
and locally adapted to N-connection geometric objects:
'/
1Hs = 4.n.otI d3ii-E -I { iV+u('~)V=u(~) [g(o,)(t~) (u) +b(,~)(~)(u)]
S.L Vacaru/Nuclear Physics B 494 [PM] (1997) 590-656 63 I
+~Iz_I ) [81,[IJIV + + AI-~IIJI (u)]~I~) + d~(u) Z+ }, (55)
+ u) = AIzlIJI(~)V+u(~) is the gauge bosch background, q~(u) is the dilaton

where Alzllgl(
field and qrlz_l~ are heterotic fermions, Ill, IJI . . . . . 1,2 . . . . . g.
7. Background D-field methods for o'-models
The background-quantum decomposition of the superfields of (1,0) higher order

anisotropic supergravity considered in the previous section can be performed in a stan-
dard manner [40] by taking into account the distinguished character of geometrical
objects on locally anisotropic s-spaces; the constraints (54) should be solved in terms
of background-covariant derivations and quantum s-fields (H+, H__ $, S), the quantum s-
fields L and E$+ are gauged in an algebraic manner (not introducing into considerations
ghosts) and note that by using quantum scale transforms we can impose the gauge
S = 0 (also without Faddeev-Popov ghosts). Finally, after a background-quantum de-
composition of superfields of (1,0) higher order anisotropic supergravity in the manner
just pointed out we can fix the quantum gauge invariance, putting zero values for the
quantum fields (in the absence of supergravitational and conformai anomalies and for
topological trivial background configurations). In this case all (1,0) supergravity fields
can be considered as background fields.
The fixing of the gauge symmetry as a vanishing of quantum s-fields induces the
ghost action
IFp = l d 3 z - E { b $ = X T + c = + b--+XT=c~}. (56)

1
The aim of this section is to compute the renormalized effective action; more exactly,
its anomaly part on the background of s-fields (1,0) higher order anisotropic supergravity
for the model defined by the action (55).
In order to integrate over the quantum fields in (55) in a d-covariant manner we
use background-quantum decompositions of the action with respect to normal locally
adapted coordinates along autoparallels defined by the Christoffel d-symbols (39). We
emphasize the multiconnection character of locally anisotropic spaces; every geometric
construction with a fixed d-connection structure (for some purposes considered as a
simple or more convenient one) can be transformed, at least locally, into a another
similar one for a corresponding d-connection by using deformations of the connections
(40) (or (3.2) and (3.3) if we are interested in na-map deformations). Let
(?s-------
f- + - - = O,
where X (~') (s = 0) = X ('~) , X ('~) (s = 1 ) = X ('~) +7"r('~) and 7r('~) are quantum fluctuations
with respect to background X ('~) . Covariant quantum fields sr('~) ("normal fields") are
defined as
(<")- 8X(~> .,--=o= (<"> (s) k o.
We shall use derivations in (1,0) higher order anisotropic s-space covariantized in a

o--model manner
- (D+,D=) r!.A</3) ~- { <r) <~) J
on properties of derivation ~7)i see (47) and (48), into distinguished autoparallel (in a
o--model manner) d-covariant derivation with properties
D (s) T(,~)... = ((~)D</~)T(,) .... D ( s ) ( ( ~ ) (s) = O.
The derivation D, is defined as
2iD~ = {D+,D+};
we note that D, 4= V~°v. The following relations hold:
/~+(<~> = V+((t~) + L ~(fl) }V+X(~) (<,)
io<~) , : + ~v<~)w
D~( </3>= V~°v( </~>- ~,, <~.><,,><,,>w v+ X<.)r<~)
s ,
D (s) D~sr<~) (s) = sr<"> (s) ((T> (s) R(~))<,~><~>V)~X<'~>,
where
V~°vsr<t~) = v,sr(t~) + { ~ } V,X<T) sr<~>.
For heterotic fermions the d-covariant and gauge-covariant formalism of computa-

tion of quantum-background decomposition of action can be performed by using the
prescription
~IU) (s) : qtl'l)
_ (0) -- qtltl
(_), qrltl)(s = 1) = qtlt_l, + A(-),
IU D 2(-) ,/A/l(s)
- = 0,
where A~ I are quantum fluctuations with respect to the background ,F; the functions
~I/_I)(s) interpolate in a gauge-covariant manner between g~It_
) ' -I + AItl because of the
definition of the operator D (s):
~s AIIIIJISX(a)]'I"I~)
_ : [ (31IIISl +
D(s)!FIII) <'~> Os 6'Fltl-)+..(,~)
cgs al111Jlr<a)~Isl
.
As d-covariant and gauge-covariant quantum s-fields we use spinors
satisfying the conditions
O(s)xl (s) = O.
We also define the following derivation in (1,0) higher order anisotropic s-superspace:
- A Ill X7 X (~>'~ IJI

+ Ill<o> + ;
for which the following identities hold:
O(s)D+gt(_)l/I =D+X(_)III + FIll r(,~)v7 X(a)~ I11

iJl(¢r)(/~)s v + (_) ,
D (s) FII~II(~)(~) = ((r) 79(y) FIIII
Jl(a) (/3)'
where D<y) is both d-covariant and gauge invariant derivation and the strength d-tensor
of gauge fields is defined by using la-derivation operators,
FlZl I11 - t~(~)A ljII + altl algl _ al/I algl
IJl(a)(/~) = t~('~)AlJl(/3) (a) "*[gl(a)'*lJl(/3) "lxl(3)'*lJl(,~)"
The action (55) consists of four groups of terms of different nature:
IHs = l(l) +1(2) +1(3) +1(4). (57)
The first term,

l(l) _ i /
4¢r-at d3z - Eg(a)(8)(X) V+X (a)V=X (3) ,
is associated to the higher order anisotropic gravitational sector and has the following
background-quantum decomposition:
I(')[X + 7r(() ] = I0~]) + I(I) + 1~ l) + . . . .
where
fl)~ 1 d b l (1)
- b! ds b ; b=0,1,2 ....
s=O
Thes terms are computed in the usual (but distinguished from the N-connection struc-
ture) manner:
1(1) - - 27ra'i f d3z-Eg(,,>(~) (D+( ('~))V=X(~), (58)
1~')- 47ra'~- f d3z-E{ g<"><~>z~+g<">zL(<~>

+ R<a) (,~>(~)(,,) V + X (°'>V=X (') (<a> (<~>),
13(1)= 4ga' d3z 5 (R<a><c~><r)<v>V+X<cr)79=((v>

634 S.I. Vacaru/NuclearPhysics B 494 [PM] (1997) 590-656
+ R(a) (,~>(r) (v>V= X('~) 79+(<~) ) ((a) sr (~>
+ 3 D<.) R<a><,~)(r>(~) V+ X<'~) V=X<~) ( #<a)(<~> },
1(4' ) - 4ia,
- f d3z-E { ~(D(~) R<a)(~)(r)(v)V+X(~>D=T(v)
+DO. ) R<a><,~>(~><~>V=X ('~)D+((~)) ( (<a> ((~>

>
+ 3 R(a> (,r) <r)(v>((a> sr<~>79+ sr ('~>79=sr
1 R <8>(~).(r)(~'>
+ ~R(a><,~)(r><v)
15(l)=-4zra'i f a'z-E
x
{1g (D<u) R<a)<~)(~)(,>D+#<~>D=#<~>#<~'>#Ca>#<r>
+ . . . .}
16(1) =--4rr- t
2 R <a><r)<~><~>R <~><~>.
0') D + sr<~>D =sr<v)r<a>r<~)r(a>r(r>
s s s s +... },
where by dots, in this section, are denoted those terms (containing multiples ( T X ) )
which are not important for calculation of anomalies, see Section 11.
The second term in (57),
1(2) _ / fdSz-Eb<~><#>(X)V+X<~>V=X(#>,
4q-r~tJ
having the following background-quantum decomposition
1 (2) [X + r r ( ( ) ] = 10(2)+ 1(2)+ 12(2)+ . . . .
where
I~2) - 1 dbI (2) s=O" b = 0 , 1 , 2 ,

b! ds b . . . .
describes a Wess-Zumino-Witten-like model of interactions (in a higher order aniso-
tropic variant). The diagram vertices depend only on the intensity H of the antisymmetric
d-tensor b:
H(r>Oz)(v>= ~36 [(r)b(u.)(v)] = ~(

1
8(r)b(~)(v) + 6(~>b(v>(r) tS(~)b(tz)(r) )-
- -
By straightforward calculations we find the coefficients

S.L Vacaru/NuclearPhysicsB 494 [PM](1997)590-656 635
1~2) = - i [d3z_E((r)V+X(~)V=X(/3)H(r)(~)(~) '
2¢ra ~ J
12(2)- 4 ~ ' e ld3z-E
{ ((r)D+((a)V=X<B)H(r)(a)(B)
+(<r) V +X <a>D=( <~>H@)(<~)(~)
+(<a) ((T>V+X<,~)V=X(~)D(a) H(~.>(<~)<t~)},
1(2)=-3 4"n'a'-i/d3z-E{ 2<(')D+<(')D=<O')H(')(~')O')

+ 2 (<.) "~.<p)"T(Y)R (~)
.(y)(p)l ('D H0")I(r)<~>U+ X(<s) V= X0')
+~(<~-)((a) ( D+((u)V=X(V) + D=f(~>~7+X<~)) D<a>H(~)<u)<~

)
+3 (<o")((a)((r)V+X(~)~7:X<~)D(~)D<a)H<~><~)<fl)},
42)- 4"rra'
/ S d 3 z - E { I((~)((a)D+((~)D=((~)H(')(u)<~)
+6 c(a)~(~)((p) ((z')V +X<8>V=X(")D<a)( R<'~)'z)<P)E(~) H0')I 0)<u) )
+2~'sr(~)r(p)
s r(Y)
s R(U)
"(')')(p)l(8) H0')l(r)(~) x ( W+X(~)D=X(~)
+D+X @>V=X (8)) + (<~>(<~>(<'~>(<~>V+X@>V=X
x {l"r' ""
~-L.,<a>nO-)<,<,>[<o ) R <,,><o<>.<n)
(u) ( 2 J
ID(,~)D(~)D(~)H(,)(v)(p)~
+4(<~>(<a)((~)(D+(<~>
' V=X
<">
+D=((~') W+ X <u)) D(y)D<a)H<~.)<u)<~'>},
1(2)--5 4:a'-i Sdaz_E{l((r)(<~)((y)D+(<u)D=((v)
x[ID(r)D(a)H(,)(~,)@)+2 n,(p> ,i I <~>}]+....

. (y> (,DI (u> n<v>
16(2) = _ 4"trod
i / dBz_E(<,>(<a>(<~>
(<n>V+(+>D=(<~
>
l
X ~D(a >H (r)<u)l<p)R<,)l<a).<fl)
 1 ~
+]-~r-'(a) R<#Z)
.(y)(p)l(6)H0,)l(r)(u)
+ ~ D<<~)D<,)D<a>H<T)<~>
] + ....
636 S.L Vacaru/NuclearPhysics B 494 [PM] (1997) 590-656
where by dots are denoted terms not important for the calculation of anomalies, and the
operations of symmetrization () and antisymmetrization are taken without coefficients.
The third term in (57)
1(3)= 10(3) + 1~3) + i(3) + . . . + i}3) + . . .
is of the Fradkin-Tseytlin dilaton type with coefficients
/0(3/= - 4 I r ~1
a fd3z_Ea,~+.,
I(3) = - 4,trott
1 /d3z_e,Z+;<°)8<o). '
(3) l t
47"rff fdaz-Ea'X+l,2(~)g('~)79(.)D(~)q~ ....
I}3)
1 =' - 47ra' / a3z-E""~+-~ ...;<<D<o,).• Vl~ >~.
The fourth term in (57),
•(4) -- 4¢rcdl f d 3 z _ E ~(_)l,iD+qt(_)l,i,
has the following background-quantum decomposition:

1 (4) [~ -4- A(X), X 4- 7"r(() ] = 10(4) 4- 1(4) 4- 12(4) 4- . . . .
with coefficients
i(4) = ----!ll--47rce'f,ze {X(_)[tlD+x(_) Itl 4- g'(_)ltlD+g'(_)ltl
4-°/(_ ) III F[tlIJI(~)(~,)((~)V +X(U) tP'(_) IJI},

12(4)_ 4¢rot'
~ f d 3 z - E {X(-)IIID+X(-)Itl
+ 2 X (_ ) Itl FIIIIJI(~,)(,,)((~)V +X(") q"(_) Ijl

_ l((~,)((u),p,(_)ltl~(_)ljiD(a)FlillJl(u)(u) V+X (p)
-l o/(_)ltlqz(_)lJIFliiiJl(t~)O,)((~)D+((u)},
1(4) =---14tree' / d3z- E { X (-) III FI'IIJI(~')(~)((u)V +X(') X(- ) I''

+((a) ((~) X(-) lZlD(a)FlzllJl(~)(~) V +X(~) qZ(_) [JI
4-X(-) 111FI/IIJI (u)0,)sr (~)/)+((")~( _)lJI
_ 3((,~) (0')~(_) I11D(a)/71111Sl(~)(~,)~(-) IJI D+((~)
-l ((a) ((r) ((#)q"(-) Ithtt(-) lJlV+XO') }
×(~D(r)~D(~)Fltlljl(g)(v) - R(a)l(v ) .(~)Fl,llJl(g)(,,)),

I4 4) = - - _ 1
4era' /d3z_ E {~X(_)
1 I IFI,IIJI(u)<~)X(_) IJl((v) D+((,u )
z
-~ ((a) ((~) X (-) lll ( D(a) FlillJf(~)(v) ) X(_) IJIv +X (~)
4-~l'f(a) l'f(la') )((_) lll ( ~l~(a) FizllJl(lt){v) )ltr (_)[Jl~) +l'f (v} -+- ~l'f(Y) ((a) l~(l"t)X(_ ) {'l
×(D(y)D(a)FIIIIJI(u)(v) + R (y)(v).(A)(P)F[I[[J[(u)(P))~7+X(v)~T't(-) [J[

14 (<A)f(~) ({Y)~(-) IIIT~(- ) IJI( 3D(~z)D(A)F~tlIjI(y)(~)
+~, <p) 1
"'(a)(v)-(y) FllllJl(r)(p))D~((v)- ((r)((a)((Iz) ((Y)~(-)III~(-) IJ]
× [D(y)D(~)D(a)FlzltJl(~z)(v) + (D(y)R(a)(~).(u)(o) )FIIIIJI(~)(p)

+3(D<z> FlzllJl<~>) R <a><~>. >] ~7+X<~>}
<~,
1 ( 4 ) = - - 1 547ra' /d3z-E {-~ (<~>(<a>x(-)~1~

X (D(a)FltllJlOz)(v))X(-)IJ[D+bF(/z) -~- . . " ) ,
16(4) = _ ----SI fd3z_E{_~__~f(r)(lu)f(A)X()IIIx()Ij
47ra 1 _ _
X (3V(r)V(a) Fl,llJl(g)(v) + Fi,I}Jl(y)(p) R(a)(,,).(glP))V+((v) + ... (59)

)
The kinetic terms for the quantum fields ((~) and X(_)I JI in the decompositions
(2.14) and (2.15) define the propagators (27ra' = 1)
(((g) (/0((.) (//,)) = g(~,)(.) D+
=g(U)<~> 1
(27r)2/ddZp(--~D+[eip(z-z"6(- )(0 - 0')l,
(X(_)lll(ii)x(_)lJl(i~t))=i~llllJ[O=O+ [] ~(-)(&~')
3
~llllJl f .,dz.p= n [oip(z--z')~3 [;.

- (277.)2 ,* t, 7 ~ , + t ~ ,-,(_),,,U')].
Finally, we remark that the background-quantum decompositions of the action (55)

for the heterotic string define the Feynman rules (vertices and propagators) for the
corresponding generalization of the two-dimensional sigma model which are basic for a
perturbation quantum formalism in higher order anisotropic spaces.
8. Green--Schwarz action in dvs-bundles
The Green-Schwarz covariant action (GS action) for superstrings can be considered
as a two-dimensional o--model with a Wess-Zumino-Witten term and a fiat, dimension
d = 10, s-space as the tangent space [43,49,50,73]. The GS action was generalized
for the curved background N = 1, d = 10 of the superspace under the condition that
the equations of motion hold [ 127] and under similar conditions for N = 2, d = 10
supergravity [45 ].
The GS action in dimensions d = 3,4, 6, 10 can be represented as
1= 1 / d2z v/-~'geOeu<'~) °aiu(fl) t.l<,~>l(t~>.~

C£) (-~I)TI(~-)(~)
1 f
+ ~ J d 2 z e°;'Oeu (a>oaiu(8) B <,~)(8) (60)
by using the flat vielbein l ('~-> and 2-form
1 ,~
B = ~Su ()A6u(~>B(,~)<~)
in the flat d = 10 s-space with coordinates u (`~>. An important role in the formulation of
the GS action plays the fact that the 3-form H = d B is closed, dH = 0. Because H is
s-invariant, the 2-form B changes on complete derivation under s-transforms
6*H = O, 8*dB = d~*B = O, 6*B = dA,

where 6* is dual to d, which ensures the s-invariance of the GS action.
We generalize the action (60) for higher order anisotropic s-spaces by changing the
fiat vielbein .(,~>t("-)into the locally anisotropic, ~(-~)
~<,~), with a possible dependence of the
Lagrangian on scalar fields (see Ref. [87] for locally isotropic spaces),
(,~) <8> ^
1= ~1 d2z [ x/~ygiOeu(~>oaiu (A~>(E<-~)E<-~))r/(~_)(fl)
<8>B(~_)<~>
+e~-,.O~u<,~)O,u<,>El.~l E<~>
iZl],
+i4z~lLly~2(Oe6lZll_J I + E~'~>AILIILI(~))!/t (61)
e i,-~K~, V = d e t ( I ~ ) , u ('~) are coordinates of the

where P is a scalar function, "go-,= Ve-V'
higher order anisotropic s-space and !/xI-tl are two-dimensional (heterotic) M W fermions
in the fundamental representation of the interior symmetry group Gr.
As background s-fields we shall consider
E<Z).- .'4fi"(~>
. . P(~-)
(/s) ( z ) , B = 1E<Z>E(B->B<z>(z)
_ ,
A ILlI_Sl= A ILlI_JI(,~_)E<~->,
where
The s-fields AlLIIJI belong to the adjoint representation of the interior symmetry group
Gr.
The action (61) is invariant under transforms (AE<~> _= Au ('~)EI~I):
A ~(-~)=0, AE e ~t <~) =5 v£%,
Age ILl = _ (AE<~_) x, . .alLil_Sl~lJ

<._> -- -,I
A V~e_'= 1a
-~(Y + e~-,.).Mai,~
-~, (62)
where ca is defined as a two-dimensional tensor, the parameter ,-k~o) is anti-self-dual as

a two-vector, juggling of indices of d-dimensional Dirac matrices is realized by using
the d-dimensional matrix of charge conjugation and, for simplicity, we can consider the
matrices (F(._))~a as symmetric; we shall define below the value M (@.
The variation of action (61) under the transforms (62) can be written as
AI = d2z
+AEa(eP(Vya)~a-)~!#-)rl(,~)(#)Dap
_ _ - 2eP~ (V.ya) %~'-(@~,~(~)'~
.- ~(~_>a<_~)
+ A (VV/) ~l-tl y~(79~gt ) I-sl (63)
where De is the Gr-covariant derivation and the torsion 2-form T <~->,the strength 3-form
H and the supersymmetric Yang-Miils strength 2-form Fl_tllS_l are respectively defined
by the relations
T(~ ) = dE(@ + E<3) o(~_) 1 E<B)E(r> T(._)

-~*<_~>=~ - - <n<~>'
H=dB=IE(r-)E(3->E(~-)H(@<3><r>,
Fl-111-sl= dAILIIZI+ AILII-KIAl-Kll-sl= ~E<_#)E<_ r)F<~><~).
ILIIZI
The variation of the action (63) under the transforms (63) vanishes if and only if
the following conditions are satisfied:
( 1) the 3-form H is closed under condition
( F <a-q->) ?','(F(a.q_))a/~ + ( F<a-->)'ia(F(a_))ai + (F(a-) )i~(F(a_) )~'~ = O,
which holds for dimensions d = 3,4, 6, 10;

(2) the imposed constraints
f;~) = -i(?<a-))e/, ~7<y)((~>'<B))a= ~<=)<3)Ba, F~-iLII-sl = 0,
)~;,
640 S.I. V a c a r u / N u c l e a r Physics B 494 [PM] (1997) 5 9 0 - 6 5 6
H~(,~ (l~) = 2eP ( F(m) ) ai! (F(~)) ~H~,

o .
(64)
(3) the coefficient M °2i from (62) is taken in the form
M 6i- 4iE °--'i- 4ff;'i~_)(?(~-)) ~H,_. - qs It-Iy*qtl-JI w7,

= e.,_°,lLri~j _p ,
where
D~P + 2H£ - 2B~ = 0;
(4) The last term in (63) vanishes because of the conditions of chirality,
grill = _y5grlZl.
In locally isotropic s-gravity it is known [41,82,10] that s-field equations of the

type (64) are compatible with the Bianchi identities and can be interpreted as standard
constraints defining supergravity in the superspace. Considering geometric objects locally
adapted to N-connections and Eqs. (64) we obtain a variant of higher order anisotropic
supergravity (see Refs. [ 106,110-112] for details on locally anisotropic supergravity)
which for dimensions d = n + m = 10 contain, distinguished by the N-connection
structure, equations of motion of N = 1 of higher order anisotropic supergravity and
super-Yang-Mills matter.
The constructions presented above can be generalized in order to obtain a variant
of higher order anisotropic N = 2, d = 10 supergravity from so-called IIB-superstrings
[45] (which, in our case, will be modified to be locally anisotropic). To formulate the
model we use a locally adapted s-vielbein 1-form E (~-) = 6u (~) EI~I, an S O ( I , 9 ) ® U(1)
connection 1-form /2 (~_),
(~) a 2-form of the complex potential A and one real 4-form B.
The strengths are defined in a standard manner:
T(Z) = DE(~-) = ~E(~) + E (~)

- o(~-)
"~(~/' (65)
R(¢~) _(t~) + O<Z)

(~-) = 6sx(~-I (~-)$2(B_)
(z~ '
F = 6A, G = ~3B + A F - A E
On the mass shell (on locally anisotropic spaces we shall consider distinguished
metrics) ds-tensors (65) are expressed in terms of one scalar s-field V E SU( 1, 1):
(q s) q'O-s-g= l.
V= -ff ,
Excluding a scalar by using the local U( 1 )-invariance we can use the first components
of the complex s-fields (q, s) as physical scalar fields of the theory.
The constraints defining IIB supergravity in d = n + rn = 10 higher order anisotropic
s-space contain equations (on every anisotropic "shell", in locally adapted frames, they
generalize constraints of IIB supergravity [52] ) :
~"b~, c~, --- ~ "b e cI, =o, ~,~

bp cp =
--lo- ''bp cl'
•
T~b"p Cp -~"
-- b I, Cp = O, (66)
<---+--- ---+--.~ .,______+ ~_______~ ~
F~,, 6, ~,, = F~v bt' cl, ---- Fa,, bt, Q, = Fat, be Cp = Fa,, b,, c, = 0,
~ ,--,...--.-.+ .,_____+____+ ~._+__..+ ~____...+
dp
F.,,. b,, ce = - - i q ( o-a. ) b,, c,,. F~vb ,, c~ = - - q ( o-a.b,, ) c ~ A d v .
d 1,
F,,,, 6, ~,, = - i s ( o'a, ) Op c,, , Favbvc e = S(O'a,,b,,)%~ A d,,,
Hap b,, c,, = -- i ( q -- -g) ( O'a,, ) b,, cp , Ha, bv c,, = - i ( q - 2) ( o'a,, ) be ce.
____+~ ____+____+
dp dp __
H,,,,b,, ~.~ = -- ( q -- -g) ( o-e,o,,) U,, Ad,, , na,,Oe c,, = - i ( - ~ - s) (o-apbp)~Tt, Zd,,,
____+ ____, ~
where
d,, dp
.___+
d t,
( o-apo-b,, ) c,-~, = ( o-apbp ) c t, "1- 7"]apbp(~7~, , (67)
.---+ ___+
the same formula holds for . .< .. . ... - -underlined spinors, spinor Adv will be used below
____+
for fixing of the U(1) gauge and the 3-form H _ F + F = 6B is real and closed
(this condition is crucial in the construction of the GS action on the background of liB
supergravity, with respect to the usual isotropic string model, see Ref. [45] ).
The action (60) can be generalized for N = 2 higher order anisotropic s-spaces in
the following manner:
Is= d2z x/-~y~P(q,s)t:~. - ~- r/<~_></9+ - t:~ o<~><~> (68)

, - 2 ~ % ,
where P ( q . s) is a function of the scalar fields q and s, and
The variation of the Lagrangian in (67) under respective k-transforms of type (62)
can be written as
A L =./'--~,',2i~PE<~'>
v-r, ;. AE</3>T<~><y>e.- ^/<_~><#_>+ I e)"E~>
(-~> ..<~>T ' E~B>AE<r>H<~><~>(r>
1 /~ 7~ ~<a) ^</~> ^
+~[ (v/-z-yy"nP)+ x/7-yy APIE;- E.- r/<~_)<~>. (69)
Taking into account the constraints (66) we can express the variation (69) as
AL=({-i(e~Uyea(q-2) +y e r)tz~ t~. to-o(~>)IEa ka~
- - i [ ( - - T ) - l / 2 e " u g # a ( q -- s) + v/S-~y*"r~Plg~>~,Se>(o-<e>~<,>)Z~ - 6, k~2
t~ t~u r/</~><~>)(o-(r>)*-~-[ - s)A +__

+ ( _ 9 / ) -U2~'~'a~a(q _ ~)A& ] ka fl } + h.c.)

+.._
1 ~ ;.# 1 ~'~^0')~(3)
where h.c. denotes Hermitian conjugation.

Using relation (67) and fixing the U( I ) gauge so as to have
-- ot --~--
P=q - s = q - s and A P = ( q - s ) ( A E + - - A ~ , - AE*--A~ ) +---
we can obtain zero values of the coefficients before the (o'apb,)terms. The rest of the
terms in A L vanish for a corresponding fixing of the variation A(v/-sTy'" ).
So, in this section we have constructed a model of higher order anisotropic liB
superstring on the background of IIB supergravity with broken chiral U(1) subgroup
of the supersymmetry S U ( 1 , 1) group of automorphisms of N = 2,d = n + m = 10
supergravity. We omit in this work the calculus for supersymmetric fl-functions.
9. F e r m i s t r i n g s i n h i g h e r o r d e r a n i s o t r o p i c spaces
There are some types of Fermi strings depending on the number N = 0, 1,2, 4 of
supersymmetry generators (see, for instance, Refs. [54,63,64] for reviews and basic
references to this classification for locally isotropic strings). The aim of this section is
to present the basic results on Fermi and heterotic strings on higher order anisotropic
backgrounds: the construction of actions and the calculation of superconformal anoma-
lies.
We note that there are two possible interpretations of models considered in this
work. On the one hand, they can be considered as locally anisotropic supersymmetric
two-dimensional supersymmetric non-linear sigma models connected with supergravity.
Under quantization of such a type of theories superconformal invariance is broken;
the two-point functions of the graviton and gravitino, computed from the quantum
effective action, can become non-trivial in the result (this conclusion was drawn [ 1]
for locally isotropic sigma models and, in general, holds true for locally anisotropic
generalizations). On the other hand, our models can be interpreted as Fermi strings on
higher order anisotropic background. We shall follow the second treatment.
The effective "off-shell" action F for a (infinite) set of locally anisotropic fields is
introduced (in a manner similar to [39]) as
F [ G, H . . . . ] = ~ e"~x / [ Dy~-~f) ] [Du ('~) ] exp( (70)

./
,/{,
X
I - 27ra' a 2z ~ ~y~~) oa~('~)oa,~(#) G(,~)(3) (u)
0,.~(3> H(a)(3) (u) + . . . ~, (71 )

)
where dots are used instead of the possible sources (with higher order derivations),
compatible with the reparametrization invariance, of other types of perturbations and
y!e) is the Euclidean two-dimensional metric. In (71) we consider in an explicit form
the components of the locally anisotropic graviton and antisymmetric d-tensor and, for
simplicity, omit the dilaton field and topological considerations.
The problem of calculation of F is split into two steps: the first is the calculation of
the effective action on a higher order anisotropic with fixed Euler characteristic X, and
then the averaging over all metrics and topologies. In order to solve the first task we
shall compute
exp{-W[G, H,g]} /[Orl] exp{-l[u + v(z),gl}. (72)
The general structure of W from (72) is constructed from dimensional and symmetry
considerations [ 85],
+y(u) f ( e ( z ) ~ ) z E]~z~ ( e ( z ) ~ ) z ' d 2 z d Z z t, (73)
where the dimensional regularization (e = 2 - d 2 on world-sheet), Y(e) is the determinant

of the two-dimensional metric, R = R ~ ~i./. We have the relation fl(u) = 4 y ( u ) for the
dimensionless functions (we can compute them as perturbations on a ' ) and note that
the second term in (73) is the Weyi anomaly which shall be computed by using normal
locally adapted to N-connection coordinates on higher order anisotropic space.
For a two-dimensional conformal flat two-dimensional, d2 = 2, metric we find l~.~)
),(e)
/~-~. = e2°~/~-~., R ( u ) = - 2 e - 2 ° [ ] o-, ai = e-2O-~i~,
Y(e) ~ = e2O.
From the decomposition -yg~.

(e~ = 6~;-+ h~'i with respect to first and second order terms on
h we have
v r z(e) = l + e o ' ( z ) + L ( d - - 4 ) ° ' Z ( z )
when h~-t= 2o-he;, h --- h~ = 2od(z~. We give similar formulas for the frame decomposi-
tion of two metric:
el/2e~= \ l + ~ e o ' j ~ . e ~ , e-l/Ze~e _
which are necessary for dealing with spinors in curved spaces.

Transforming the quantum field sr --~ ~ r , where sr is the tangent d-vector in the
point u E £(z) of the higher order anisotropic space and using the conformally flat part
of the two-dimensional metric, we obtain the following effective action necessary for
further calculations
d) c) v/~Tq_ f)
a'R ~'R
a (~ a a ~ a a ~r
.) I,) .,)
Fig. I. The diagrams defining the conformal anomaly of a closed boson string in a higher order anisotropic
space.
1 , 1/2 e~
li~n~t= f d2z [ ~ ( 2 ~ ' a ) e H(4)(~)(,)(u)O~(")&,~(/3)~ '')
27ra'
+ ---6-- eo-( z ) R(4) (#) (r) (a) (u) a~sr (~) 3esr (a) sr (#}sr (r)
1 a a 27ra'
+ T2~ra'D (4) H(4) (t~)(~) ( u ) eesO~~r(Z) &~-(r) ~-(4) ~-(8)] , (74)
where (u) are higher order anisotropic space coordinates not depending on two coordi-
nates z. So, the anomaly in (74) takes the form
4of a2z (75)
if ~- in (70) is the connection on topologies constant the first term in (71) can
be absorbed by the renormalization of this connection constant. The set of two-loop
diagrams defining (73) is illustrated in Fig. 1. We note that we must take into account
tadpoles, because of the compactness of the string world-sheet there are no infrared
divergences.
In the one-loop approximation (Fig. 1 ) we find
fl(1)= 1 (n+ml+...+mz) 7(1) = 1

247r ' 96~- (n + ml + ... + mz ) ;
there is a correspondence with classical results [ 85] if we consider a trivial distinguish-
ing of the space-time dimension nE = n + ml + ... + mz.
The two-loop terms (Fig. 1) from (74) are computed as
7"ra'e2R / d2ko-(k)o.(_k) J dd2pdd2 q

(b) = 3(277.)6
( p . k - p2)E(q, k - q2) + (p. k - p2)2(p . q)( q. k - q2)
×
p2(k _ p)Eq2(k _ q)2
qrolte 2R
(c)-3(--~-~)4G(O) f d2ko'(k)o'(-k) f d d 2 P ( P ' k - p 2 ) 2
(d)= faa'p ,
2eZ~o/H~)
(e) + (f) = (27r)6 (a)(r) e~;e~ jf d 2 k t r ( k ) t r ( _ k ) . I ddZpddz q
l
X p 2 ( k _ p)2q2( k _ P _ q)2
( k . p - p 2 ) ( k , q - q2)p~( k - q)~.( k - q)aq~
× (k_q)2
( k "p - p2)2q~( k - q)~'qa( k - p)~ ]
+
where
l/
o'(z ) - ( 2 ~ ) 2 dZp o'(p) e x p ( - i p x ) .
The contributions of non-vanishing tadpole diagrams mutually compensate. The sum

of the rest of contributions results in the anomaly
y(2) = .,(
64"n"
-R + H2
= ~
-R - H2
'
where H 2 = H(,~)(/3)0,)H (') (/~)0') and R i.s the scalar curvature with torsion.
Computing W in the leading order in d for the closed boson string, it is not difficult
to find the effective action /" for massless perturbations of the string (of the metric
G(,~)(t~) and field H(,,)(/~)) on the tree (X = 2) level. Taking into account the identity
f (R(z) yv/y--~e))zD~z~ ( R ( z ) ~ ) z , d Z z d 2 z ' = 167r
for the metric on the sphere we find
F(°)[G,H] ,.-, (2,n.at)nv./2

.,(
1 + --~ - R + H 2 , (76)
where nE = n 4. ml 4 . . . . 4- mz is the dimension of the higher order anisotropic space.

Formula (74) generalizes for such types of spaces (scalar curvature R and torsion H
are for, distinguished by N-connection, on a la-space) to that presented in [27,28].
The cosmological constant in (74) arises due to the tachyon modes in the spectrum of
bosonic strings and is absent for superstrings. From the vanishing of fl-functions for
and taking into account the contributions of reparametrization ghosts [85] into the
anomaly of the bosonic string we obtain in leading approximation
24zr 4. -R- H2 4. . . . .
"E--26 Ott(__~ 2H2)

y = 96-----~- + 6 - ~ +...
In consequence, the correction to the critical dimension is
Dc = 26 + o/H 2 + O ( [ a ' ] 2 ) . (77)
We emphasize that the torsion in (77) can be interpreted in a different manner as

in the case of locally isotropic theories, where H... is considered as an antisymmetric
strength of a specific gauge field (see the Wess-Zumino-Witten model [ 124,127] ). For
locally anisotropic spaces we suggested the idea that the H... terms are induced by the
distinguished components of the torsions of, in our case, higher order anisotropic spaces.
The constructions presented in this section can be generalized for the case of N =
1 and N = 2 higher order anisotropic Fermi strings. Let us decompose the action
I s [ e ~ , u + ( ( z ) , ~ h ( z ) ] to within fourth order on quantum fields (('~) and ~h~!'~)(z).
After redefining ( ~ ~ ( , ff --* e - 1 / 4 ~ we obtain the following additional
to (72) term:
/F(int) = f d2z [/~:O'~'(a)~0~tb(a-)

/4 . . . .
- 2~ ( 1 - 2 o ' ) H(,~)(#)(v)~(a)y~ (0~((r)) ~h(t~)
+ i -7ra'
T (1 + 2o') R(~)(#)(v)(8) ~(~)ye (0el(8)) ~h(/~)((v)
+ -7ra'
~- ~ ~(-)(t~)(v)(8) ~(")
q~ ( 1 + ys)~h(r)~(¢~) ( 1 + ys)~h(8)J
Fixing the gauges
1
N= 1: e7 = e ~ 4'e = ~ , e a ,
N =2 : e~ = e ~ ~h~ = ~1y ~ A , A ~ = ~e Oap,
where A is the Majorana (N = 1) or Dirac (N = 2) spinor. Because of supersymmetry

it is enough [39] to compute only the coefficient before Weyl anomaly in order to get
the superconformal anomalies. The one-loop results are
N= 1 : ~ ( l ) = ne y(l) ne
167r' - 64~r'
N=2" /3(l) ne )= ne
= 8---~' y(l 327r'
from which, taking into account the reparametrization and superconformal ghosts we
obtain the following values of the critical dimension (y = 0):
N= 1 : fl}l) _ ne - 2 y(i) ne - 10
16~r ' - 64q'r '
N=2 : [3(I) = n_.L y(1) nF. -- 2
87r' - 327r"
Fig. 2. The diagrams defining snpergravitational and superconformal anomalies in the theory of higher order
anisotropic superstrings.
From a formal point of view, in the two-loop approximation we must consider dia-
grams ( b ) - ( m ) from Fig. 2. By straightforward calculations and using methods similar
to those presented in [ 39] we conclude that all two-loop contributions ( b ) - (m) vanish.
As a result we conclude that for Fermi strings the following formulas hold:
ne - 2 dH 2 nE - 10 ee'H 2
N=I : ,8= - - --+ .... y-
16rr 24rr 64rr 96¢r '
nE o/H 2 nE -- 2 ct'H 2
N=2 : ,8 = 87r 24----~+ . . . . Y= 32--~ 967r
or
2 i 2
N=I : Dc=lO+-~trH + ....
1 ~, I 2
N=2: Dc = 2 + ~o~ r* . . .
in leading order in o/.

Finally, we note that because H 2 contains components of the N-connection and the
torsion of a d-connection on higher order anisotropic space we conclude that a possible
local anisotropy of space-time can change the critical dimension of Fermi strings.
10. A n o m a l i e s in locally anisotropic it-models
Anomalies in quantum field theories were considered starting with Refs. [2,17].
Conformal and gravitational anomalies have been analyzed in Refs. [ 32,37,3] (see also
the reviews [ 129,80,63,64,44] ). The aim of this section is to investigate anomalies in
higher order anisotropic ( 1,0)-superspaces.
10.1. O n e - l o o p c a l c u l u s
Supergravitational and conformal anomalies of heterotic locally anisotropic o--models

connected with (1,0) higher order anisotropic supergravity are defined by the finite
(anomaly) parts of the diagrams of self-energy type (Fig. 2). In order to compute
the anomalies it is necessary to consider all the vertices of the theory with no more
than a linear dependence on potentials /'if+, H~ (see Subsection 6.2 and Section 9 for
notations on higher order anisotropic heterotic superstrings). We consider the following
"effective" action (without ghosts) :
I = I0 + lint, lint = 14 + I1, (78)
where
lo = - f d3 z - [ iD +((~) 0=((/3)'~(~>(/3) + XIIID+xIII],
I~ = - ~ 1 j[ d 3 z - { iD +(('~10=((131(((:,) AI~')
+( (8>( (~>B(8)(~) (~>(~) + ((~) ((8) ((~> ~ (~)(8)(~>(~) (~)
u_,l/I .IJI D r(/~)
(:(,~)~llllJI + :<z,):(,,)~l/ll~l ((~)((~)(<~)~-I~ll~l }

x ,~ (~)(~) ~ ~ (~)(~)(~)+ (~)(~)(~)(~)) ,
Ii =-~ d3z - (g(a)(~>+ (<~>A(~)<~>(~>
+ ( (8>((~>B(8>(~)(~>(~> + ((~)((8)((~>/3(~)(8)<~)(~)(~>
+( (~)((~) ((8) ((~) ~ (s)(8)(~)(~)(~)

(~)) + [iH~0=((a) O=((~)
+I D+((~)(D+H~=)D+((/~)
+ i(D+((~))H~=(
O,((/~))
]
+4-~i Z6 L¢(a,)p!. .((ap)/~(,,).

. . . ~(~,,)4~(D+OcH~=+O~=H+)},
p
where
I^ IA
B(8)(~)(a)(13) = -~R(8)(a)(l~)(~) + ~D(8) (~)(a)(t~),
1
+17)~( ) R8< >(>(~)(v)

. + ~
2 R(,(8)(~)i
) (~) <t~)164(~),
1
/14(~)(8)(~>(r)(~)(~> = ~6D(~)D(8)R(~)(~)(~>(~>
2 t~(/~)
~ (r)<~)(0)n(o)
~(8)(~)(a) + D(~)D(8)D(~)H(r)<~)(/3)
1 ^ ^<o) ^ 1~ ~, ~(o)
+ ]-{D(~)R(~)(6)[(a)H(~)](r>(o) + ~ t:(~)n(r)(o)[(~> I(8)(e)l(~)l'
S.L Vacaru/NuclearPhysicsB 494 [PM](1997)590-656 649
0= GO=
P P
Fig. 3. (1,0) supergraphdefining the one-loopanomaly.
D+~ II~_ = ~ ) tI~_

H*= H~= = tt~_ o= a= It~
Fig. 4. ( 1. O) supergraphsdefining the one-loopdilaton contribution to the anomaly.
H7 0 H*__
Fig. 5. The diagrams not essential for the calculation of anomalies.
~llllJI _ l ~ l / l l J I ~llllJI _ 1 ~ ~l/llJI

(,q(/3) - 2"('~)(/3)' ~(z,)(,~)(/3)- - 3 (v) r(,~)(/3),
~-I/llJI 1 z o ~ ,~ /~l/llJI _a_ "fflliIJIl~(r) ).
(~)(v)(,~)(/3) - 24 ~''~ (~)~(r) (,q(#) "'(~)(~)"(,~)(r)(/3)
For simplicity, in this section we shall omit the tilde over geometric objects (such
as curvatures and torsions computed for Christoffel distinguished symbols (39)) but
maintain the hats in order to point out the even components on the s-space).
We write the supergravitational anomaly in the following general form (see Ref. [ 124]
for locally isotropic models):
l
32,rr/d3z -
{T,D+H+~H+ -iT2D+H~=-~H~=},
03
where background-dependent coefficients will be defined from perturbation calculus on

or' by using (78). For computation of supergraphs we use the standard technique [40]
of reducing to integrals in momentum space, which is standard practice in quantum field
theory. For instance, the one-loop diagram corresponding to Fig. 3 is computed as
1 [ k2(k=+ p = ) 2 i p~
(2zr) 2 ~, d2k k2(k+p)2 - 247rp2.
Diagrams of the type illustrated in Fig. 4 give rise only to local contributions in the
anomaly and are not considered because of dimensional considerations. In Fig. 5 dia-
grams are given that are not essential for the calculation of anomalies.
We note that comparing with similar locally isotropic results [64] the torsions HI
are generated by components of the distinguished torsion of the higher order anisotropic
background.
650 S.L Vacaru/Nuclear Physics B 494 [PM1 (1997) 590-656
0=
Fig. 6. B-dependenttwo-loopcorrectionsto the anomaly.
10.2. Two-loop calculus
The two-loop B-dependent corrections to the anomaly are defined by the diagrams
illustrated in Fig. 6. The one-loop results including ghosts (56) are similar to the locally
isotropic ones [40]. Thus we present a brief summary (we must take into account the
splitting of dimensions in higher order anisotropic spaces):
I'I¢I-I°°P=
"eft
l f
96rr d3 z -
{(
n + ml + • • • + mz - 26 +
NE)
2
D+H=+.4
H;[]
a3 }
- 3 / ( n + ml + . . . + mz - IO)D+Ht-L_aH*- • (79)
2 - -
If the action (79) is completed by local counterterms
H+ [] H¢=, Se2=H=+, SetD+H~=, S&D+S

and the conditions N e - (n + ml + . . . + mz) = 22 and 3/1 = 3/2 - 3/are satisfied, we
obtain from (77) a gauge invariant action:
,
Weft + Wloc = 1 - ~ ( n + rnl + . . . + m z - 10)
/ d3z-~ + ~+, (80)
The action (80) is the conformal anomaly of our model ( 2 + depends on S). The
critical parameters of the higher order anisotropic heterotic string
n+ml+...+mz =10and Ne=32
make up the conditions of cancellation of their anomalies (the original locally isotropic
result was obtained in [46] ). We can also compute and add the one-loop dilaton (see
Fig. 5) contribution to (78):
w ( l , ~ ) = 7)2qb
"eft
/
128¢r2 a3z-
[ "
~ ~ .]
iD+I-I+=-~t-l*= - D + H + - ~ H + ,
3/(1,,~) ---- 1 D 2 ¢ ~
_
4~-
and compute
d2kd2q q2(q=+p=)(k=+p=)
(a) = l l ( p ) = 16rra ( q 2 _ l z 2 ) [ ( q q _ p ) 2 1 ~ 2 ]
D÷ D+
D + ~ i 3 = l) 0=~0= m)
D = D+ D+ +
O= D+ i) j) ~)
D+ O= D+ O- D+ O=
D = d) O= e) O= D+ f)
O= D+ O- D+ O= D+
c)
Fig. 7. A2-dependent two-loop corrections to the anomaly.
k2 1 p4
X ( k 2 _/./,2) [ ( k -nt- p)2 _/.~2] - 64qT 2 p2 q- O (/~2) ,
d2kd2q q=(q= + p=)

(b) =12(p) = 167r4 ( q 2 _ t x 2 ) [ ( q + p ) 2 _ t z 2 ]
k2(k= +p=)2 1 p4
x(kZ_l.tZ)[(k+p)Z_/.t2] - 327r2p 2 +O(/.t2),
where the mass parameter ~2 is used as an infrared regulator.

Two-loop A2-dependent diagrams are illustrated in Fig. 7. The supergraphs (f), (i)
and (j) are given by the momentum integral If = li = lj = I(p):
dZkdZq k=q=(k= + q= + p=)2 p4

l(p) = (27r)4 k 2 q 2 ( k + q + p ) 2 -96rr2p2;
also the anomaly parts of diagrams ( a ) - ( e ) , ( k ) - ( m ) are proportional to this integral.

The rest of the possible A2-type diagrams do not contribute to the anomaly part of the
effective action. After a straightforward computation of two-loop diagrams we have
,oo0=
16¢r
there are no dilaton contributions in the two-loop approximation.

The anomaly coefficient y is connected with the central charge of the Virasoro su-
peralgebra of the heterotic string on the background of massless modes (under the
conditions of vanishing of fl-functions or, equivalently, if the equations of motion are
satisfied, see Refs. [99,100,97,84,35,81,25]):
23' + a' ('D(,8)~/))2 = / ~ - B* _ _lag .<-><~>
n + ml + . . . + mz - lO a'
= 2 + -~- Leff,
<r>_ = 5 ~ + + ....
g
where (T)_ is the averaged supertrace, fl(,~)(~) is the metric fl-function, flo is the dilaton
fl-function and by dots we denote the terms vanishing on the equations of motion. We
note that the fl-functions and the effective Lagrangian are defined in the string theory
only with the exactness of a redefinition of fields [25]. According to standard calculus
of perturbation theory in a' we have
2
and
(0)
ff =- f d
10X v / - ~ [ L e f f(0)
=½f d'°XX/~I{-R-4D2a~+4(Z~(~)~)2+gH
}, 1 2
where
2 <~)
H<,~)<~) -- H(~) <r)(8)H(o)<r)
which is a higher order anisotropic generalization of the models developed in [ 20,3 l, 18 ].
II. Discussion and conclusions
We have explicitly constructed a new class of superspaces with higher order anisotropy.
We will now discuss the status of the results in this work and the relevant open questions.
From the generally mathematical point of view a definition of a supersymmetric
differential geometric structure imbedding both types of supersymmetric extensions of
Finsler and Lagrange geometry as well various Kaluza-Klein superspaces is possible.
The first type of superspaces, considered as locally anisotropic, are characterized by non-
trivial non-linear connection structures and corresponding distinguishing of geometric
objects and basic structure equations. The second type, as a rule, is associated to
trivial non-linear connections and higher order dimensions. Substantial interest in further
considerations presents the investigations of the physical consequences of the models
of field interactions on higher and/or lower dimensional superspaces provided with an

N-connection structure.
It is worth noticing that higher order derivative theories are one of a currently central
division in modern theoretical and mathematical physics. A rigorous formulation is
needed of the geometric background for developing higher order analytic mechanics
and the corresponding extensions to classical quantum field theories. Our results not
only contain a supersymmetric extension of higher order fiber bundle geometry, but also
propose a general approach to the "physics" with locally anisotropic interactions. The
formalism elaborated on in this work of distinguished vector superbundles highlights a
scheme by which supergravitational and superstring theories with higher order anisotropy
can be constructed.
To develop in a straightforward manner self-consistent physical theories, define local
conservation laws, give a corresponding treatment of geometrical objects and so on,
on different extensions on Finsler spaces with non-linear structure of metric form and
of connections, torsions and curvatures is a highly conjectural task. Only the approach
on modeling geometric models of the mentioned type (super)spaces on vector (su-
per) bundles provided with compatible non-linear and distinguished connections and met-
ric structures make "visible" the possibility (see, for instance, Refs. [75,76,102,113,103-
105,109,112 ] ), manner of elaboration, as well as common features and differences of
models of fundamental physical fields with generic locally anisotropic interactions. From
the viewpoint of fundamental ideas of string theory only some primarily changes in the
established material have been introduced in this work. But we did not try to give a
simple straightforward repetition of standard material in the context of some sophisti-
cated geometries. Our main purposes were to illustrate that the higher order anisotropic
supergravity is also naturally contained in the framework of low energy superstring
dynamics and to develop a corresponding geometric and computational technique for
supersymmetric sigma models in locally anisotropic backgrounds.
The above methods of perturbative calculus of anomalies of heterotic sigma models in
higher order anisotropic superspaces, as a matter of principle, can be used to every finite
order in d (for instance, by using the decomposition (76) for effective action we can,
in a similar manner as for one- and two-loop calculations presented in Section 10,
find corrections of anomalies up to fifth order inclusive) and are compatible with
the well-known results for locally isotropic strings and sigma models. We omit such
considerations in this work.
References
[ 11 E. Abdala and R.S. Jasinschi, Nucl. Phys. B 232 (1984) 426.

[21 S. Adler, Phys. Rev. 177 (1969) 2426.
[31 L. Alvarez-Gaume and E. Witten, Nucl. Phys. Nucl. Phys. 234 (1983) 269.
14] S.P. de Alwis, Phys. Rev. D 34 (1986) 3760.
151 P.E. Antonelli and R. Miron (eds), Lagrange and Finsler Geometry, Applications to Physics and
Biology (Kluwer Academic Publishers, Dordrecht, 1996).
[6] RL. Antonelli and T.J. Zastavniak, eds., Lagrange Geometry, Finsler Spaces and Noise Applied in
Biology and Physics, in Mathematical and Computer Modelling, ed. Rolin, E.Y. 20. N415 (Plenum
Press, New York, 1994).
17] G.S. Asanov, Finsler Geometry, Relativity and Gauge Theories (Reidel, Boston, 1985).
I81 G.S. Asanov and S.F. Ponomarenko, Finsler Bundle on Space-Time. Associated Gauge Fields and
Connections ($tiin~a, Chi~in~lu, 1988) [in Russian].
19l G.S. Asanov, Fibered Generalization of the Gauge Field Theory. Finslerian and Jet Gauge Fields
(Moscow University Press, Moscow, 1989) ]in Russian].
[101 J. Atick, A. Dhar and B. Ratra, Phys. Rev. D 33 (1986) 2824.
111] J. Bagger, in Supersymmetry: Proc. NATO Adv. Study Inst., Bonn, Aug. 20-31, 1984 (New York,
London, 1985) 45.
[12] W. Barthel, J. Reine Angew. Math. 212 (1963) 120.
[131 C. Bartocci, U. Bruzzo and D. Herm,'indes-Ruip6rez, The Geometry of Supermanifolds (Kluwer
Academic Publishers, Dordrecht, Boston, London, 1991 ).
[ 141 A. Bejancu, Finsler Geometry and Applications (Ellis Horwood, Chichester, England, 1990).
[151 A. Bejancu, A New Viewpoint on Differential Geometry of Supermanifolds, I (Timi~oara University
Press, Romania, 1990).
[ 16] A. Bejancu, A New Viewpoint on Differential Geometry of Supermanifolds, II (Timi~oara University
Press, Romania, 1991 ).
[ 171 J.S. Bell and R.A. Jackiv, Nuovo Cim. 60 47.
[181 S. Bellucci, Progr. Theor. Phys. 79 (1988) 1288.
119] EA. Berezin and D.A. Leites, Doklady Academii Nauk SSSR 224 (1975) 505 [in Russian]; Sov.
Math. Dokl. 16 (1975) 1218.
[201 E. Bergshoeff, M. de Roo, B. de Wit and E Nieuwenhuizen, Nucl. Phys. B 195 (1982) 97.
[211 L. Berwald, Math. Z. 25 (1926) 40; Math. Z. 26 (1927) 176 (E).
[22] G. Yu. Bogoslovsky, Theory of Locally-Anisotropic Space-Time (Izdatel'stvo Moscovskogo
Universiteta, Moscow, 1992) [in Russian].
[23] H.W. Braden, Ann. Phys. 171 (1986) 433.
124] R. Brooks, E Muhammad and S.J. Gates, Nucl. Phys B 268 (1986) 599.
[25] R. Brustein, D. Nemershansky and S. Yankielowicz, Nucl. Phys. B 301 (1988) 224.
[261 U. Bruzzo and R. Cianci, Class. Quant. Grav. 1 (1984) 213.
[271 C.G. Callan, D. Friedan, E.J. Martinec and M.J. Perry, Nucl. Phys. B 262 (1985) 593.
1281 C.G. Callan, I.R. Klebanov, and M.J. Perry, Nucl. Phys. B 278 (1986) 78.
1291 E. Cartan, Les Espaces de Finsler (Hermann, Paris, 1935)
130] E. Cartan, Exposes de G~omdtrie in Series Actualit~s Scientifiques et Industrielles 79 ( 1936); reprinted
(Herman, Paris, 1971).
1311 G.E Chapline and N.S. Manton, Phys. Lett. 120 (1983) 105.
1321 S.M. Christensen and M.J. Duff, Phys. Lett. B 76 (1978) 571.
[331 R. Cianci, Introduction to Supermanifolds (Napoli, Bibliopolis, 1990).
[34] M. Crampin, W. Sarlet, and F. Cantrijin, Math. Proc. Camb. Phil. Soc. 99 (1986) 565.
]35] G. Curci and G. Paffuti, Nucl. Phys. B 286 (1988) 399.
1361 T.L. Curtright and C.K. Zachos, Phys. Rev. Lett. 53 (1984) 1799.
[371 M.J. Duff, Nucl. Phys. B 215 (1977) 334.
[ 381 P. Finsler, Ueber Kurven und Fl[ichen in Allgemeiner Ramen ( Dissertation, Gfttingen, 1918) ; reprinted
(Birkhauser, Basel, 1951).
[39] S. Fradkin and A.A. Tseytlin, Phys. Lett. B 158 (1985) 316; Nucl. Phys. B 261 (1985) 1.
[401 S.J. Gates, M.T. Grisaru, U Mezincescu and P. K. Townsed, Nucl. Phys., B 286 (1986) 1.
I41 I S.J. Gates, M.T. Grisaru, M. Rocek and W. Siegel, Superspace or One Thousand and One Lessons in
Supersymmetry (Cummings Publ. Comp. Inc., Benjamin, 1983).
[421 S.J. Gates, C.M. Hull and M. Rocek, Nucl. Phys., B 248 (1984) 157.
]431 M.B. Green and J.H. Schwarz, Nucl. Phys. B 243 (1984) 285.
[441 M.B. Green, J.H. Schwarz and E. Witten, Superstring Theory, vol. 1 and 2 (Cambridge University
Press, Cambridge, 1987).
1451 M.T. Grisaru, P. Howe, L. Mezinchescu e.a., Phys. Lett. B 162 (1985) 116.
[461 D.J. Gross, J.A. Harvey, E. Martinec and R. Rohm, Phys. Rev. Lett. 54 (1985) 502.
[47] J.C. Henty, C.M. Hull and EK. Townsed, Phys. Lett. B 185 (1987) 73.
[481 E. d'Hoker and D.H. Phong, Rev. Mod. Phys. 60 (1988) 917.
[491 P.S. Howe, Phys. Lett. B 70 (1977) 453.
[501 ES. Howe, J. Phys. A 12 (1979) 393.
1511 ES. Howe, G. Sierra, Phys. Lett. 148 (1984) 451.
[521 P. Howe and P. West, Nucl. Phys. B 238 (1984) 181.
1531 J. Hoyos, M. Quiros, J. Ramirez Mittelbrunn and F.J. De Uries, J. Math. Phys. 25 (1984) 833, 841,
847.
1541 C.M. Hull, Lectures on Non-linear Sigma Models and Strings (DAMPT, Cambridge, England, 1987).
1551 C.M. Hull and P.K. Townsed, Nucl. Phys. B 301 (1988) 197.
1561 R.S. Ingarden, Tensor N.S. 30 (1976) 201.
1571 H. lshikawa, J. Math. Phys. 22 ( 1981 ) 995.
[58] A. Jadczic and K. Pilch, Commun. Math. Phys. 78 (1981) 373.
[591 V. Kac, Commun. Math. Phys. 53 (1977) 31.
[601 A. Kawaguchi, Tensor N. S. 6 (1956) 596.
1611 J. Kern, Arch. Math. 25 (1974) 438.
1621 S.V. Ketov, Nucl. Phys. 294 (1987) 813.
[ 631 S.V. Ketov, Introduction into Quantum Theory of Strings and Superstrings (Nauka, Novosibirsk, 1990)
[in Russian I.
1641 S.V. Ketov, Non-linear Sigma-Models in Quantum Field Theory and String Theory (Nauka,
Novosibirsk, 1992) [in Russian].
1651 B. Konstant, in Differential Geometric Methods in Mathematical Physics, Lecture Notes in Mathematics
570 (1977) 177.
[661 D.A. Leites, Usp. Math. Nauk 35 (1980) 3 1in Russian1.
1671 D.A. Leites, The Theory of Supermanifolds (Petrozavodsk, URSS, 1980) [in Russian I.
1681 S. Leng, Differential Manifolds (Addison-Wesley, Reading, MA, 1972).
169] M. de Leon and P. Rodrigues, Generalized Classical Mechanics and Field Theory (North-Holland,
Amsterdam, 1985).
1701 P. Libermann and Ch. M. Marie, Sympletic Geometry and Analytical Mechanics (D. Reidel Publ.
Comp., 1987).
[711 C. Lovelace, Phys. Lett. B 35 (1984) 75.
1721 Yu. I. Manin, Gauge Fields and Complex Geometry (Nauka, Moscow, 1984) [ in Russian I.
1731 E. Martinec, Phys. Rev. D 28 (1983) 2604.
1741 M. Matsumoto, Foundations of Finsler Geometry and Special Finsler Spaces (Kaisisha, Shigaken,
1986).
1751 R. Miron and M. Anastasiei, Vector Bundles. Lagrange Spaces. Application in Relativity (Academiei,
Romania, 1987) [in Romanian].
1761 R. Miron and M. Anastasiei, The Geometry of Lagrange Spaces: Theory and Applications ( Kluwer
Academic Publishers, Dordrecht, 1994).
1771 R. Miron and Gh. Atanasiu, Compendium sur les Espaces Lagrange D'ordre Suprrieur, Seminarul de
Mecanic~i. Universitatea din Timi~oara. Facultatea de Matematic/l (1994).
178] R. Miron and T. Kavaguchi, Int. J. Theor. Phys. 30 (1991) 1521.
[ 791 A. Morimoto, Prolongations of Geometric Structures, (Nagoya Univ., 1969).
1801 A. Yu. Morozov, Uspehy Fizicheskih Nauk, 150 (1986) 337.
[811 M. Niedermaier, Phys. Lett. B 207 (1988) 145.
1821 B.E.W. Nilsson, Nucl. Phys. B 188 (1981) 176.
1831 D. Olivier, Phys. Rev. D 33 (1986) 2462.
[841 H. Osborn, Nucl. Phys. 308 (1988) 629.
I85 ] A.M. Polyakov, Phys. Lett. B 103 ( 1981 ) 207; 211; Gauge Fields and Strings, Contemporary Concepts
in Physics (Harwood Acad. Publ., Chur e. a., Switzerland, 1987).
1861 V.N. Ponomarev, A.O. Barvinsky and Yu.N. Obukhov, Geometrodynamical Methods and Gauge
Approach to Gravity Theory (Energoatomizdat, Moscow, 1985) l in Russian[.
1871 T.D. Robb and J.G. Taylor, Phys. Lett. B 176 (1986) 355.
1881 A. Rogers, J. Math. Phys. 21 (1980) 1352.
[89] A. Rogers, J. Math. Phys. 22 (1981) 939.
[901 H. Rund, The Differential Geometry of Finsler Spaces (Springer, Berlin, 1959).
1911 M. Sakamoto, Phys. Lett. B 151 (1985) 115.
[92] A. Salam and E. Sezgin (eds), Supergravities in Diverse Dimensions, vol. 1 and 2, (Word Scientific,
Singapore, 1989).
[93] G. Sardanashvily, Gauge Theory in Jet Manifolds (Hadronic Press, Palm Harbor, 1993); Five Lectures
on the Jet Manifold. Methods in Field Theory E-print: hep-th/9411089.
[94] D.J. Saunders, The Geometry of Jet Bundles (Cambridge Univ. Press, Cambridge, 1989).
[95] J.H. Schwarz (ed), Superstrings: the first 15 years of superstring theory: reprints & commentary,
(World Scientific, Singapore, 1985).
[96] A. Sen, Phys. Rev. Lett. 55 (1986) 1846.
[971 G.M. Shore, Nucl. Phys. B 286 (1987) 349.
[98] A.A. Tseytlin, Phys. Rev. D 26 (1982) 3327.
[99] A.A. Tseytlin, Phys. Lett. B 178 (1986) 34.
[100] A.A. Tseytlin, Nucl. Phys. B 294 (1987) 383.
[101 ] S. Vacaru, in Foundations of Gravitation and Cosmology, Contributions to the International School-
Seminar, Odessa, September 4-10, 1995 (Russian Gravitational Society, Moscow, 1995) p. 131.
11021 S. Vacaru, J. Math. Phys 37 (1996) 508.
1103] S. Vacaru, Buletinul Academiei de ~tiinIe a Republicii Moldova, Fizica ~i Tehnica 3 (1995) 53.
[104] S. Vacaru, Buletinul Academiei de ~;tiinle a Republicii Moldova, Fizica ~i Tehnica 1 (1995) 54.
[ 105] S. Vacaru, Buletinul Academiei de ~tiinie a Republicii Moldova, Fizica ~i Tehnica 1 (1996) 62.
1106] S. Vacaru, Generalized Lagrange and Finsler Supergravity; E-print: gr-qc/9604016.
1107] S. Vacaru, Locally anisotropic gravity and strings, (accepted for publication in Ann. Phys. (NY),
1997); E-print: gr-qc/9604013.
[ 108] S. Vacaru, Superstrings in higher order extensions of Finsler Superspaces; E-print: hep-th/9611034.
[ 109] S. Vacaru, Spinors in higher dimensional and locally anisotropic spaces; E-print: gr-qc/9604015.
[1101 S. Vacaru, Locally anisotropic interactions: 1. Non-linear connections in higher order anisotropic
superspaces; E-print: hep-th/9607194.
[ 111 ] S. Vacaru, Locally anisotropic interactions: II. Torsions and curvatures of higher order anisotropic
superspaces; E-print: hep-th/9607195.
[112] S. Vacaru, Locally anisotropic interactions: II1. Higher order anisotropic supergravity; E-print: hep-
th/9607196.
[ 113] S. Vacaru and Yu. Goncharenko, Int. J. Theor. Phys. 34 (1995) 1955.
1114] S. Vacaru and S. Ostaf, Buletinul Academiei de Stiin~e a Republicii Moldova, Fizica ~i Tehnica 3
(1993) 4.
I 1151 S. Vacaru and S. Ostaf, in Coloquium on Differential Geometry, 25-30 July 1994 (Lajos Kossuth
University, Debrecen, Hungary) p. 56.
[ 1161 S. Vacaru and S. Ostaf, in Lagrange and Finsler Geometry, ed. P.L. Antonelli and R. Miron, (Kluwer
Academic Publishers, Dordrecht, 1996) p. 241.
[1171 S. Vacaru and S. Ostaf, Rep. Math. Phys. 37 (1996) 309; E-print: gr-qc/9602010.
[ 118] S. Vacaru, S. Ostaf and Yu. Goncharenko, Romanian J. Physics 39 (1994) 199.
I119] S. Vacaru, S. Ostaf, Yu. Goncharenko and A. Doina, Buletinul Academiei de Stiinte a Republicii
Moldova, Fizica ~i Tehnica 3 (1994) 42.
[ 1201 V.S. Vladimirov and I.V. Volovich, Theor. Math. Phys. 59 (1984) 317 [in Russian].
[121] A.A. Vlasov, Statistical Distribution Functions (Nauka, Moscow, 1966) [in Russian].
11221 I.V. Volovich, Doklady Academii Nauk SSSR 269 (1975) 524 [in Russian].
[ 123] J. Wess and J. Bagger, Supersymmetry and Supergravity (Princeton Univ. Press, Princeton, 1983).
[1241 J. Wess and B. Zumino, Phys. Lett. 37 (1971) 95.
[ 125] P. West, Introduction to Supersymmetry and Supergravity (World Scientific, 1986).
[ 126] B. DeWitt, Supermanifolds (Cambridge University Press, 1984).
11271 E. Witten, Nucl. Phys. B 266 (1986) 245.
[ 1281 K. Yano and S.I. lshihara, Tangent and Cotangent Bundles. Differential Geometry (Marcel Dekker,
New York, 1973).
[129] B. Zumino, Y.S. Wu and A. Zee, Nucl. Phys. 239 (1984).

Superstrings in Higher Order Extensions of Finsler Superspaces

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Superstrings in Higher Order Extensions of Finsler Superspaces

Uploaded by

Copyright:

Available Formats

13

ELSEVIER Nuclear Physics B 494 [ PM] (1997) 590-656

Superstrings in higher order extensions of

Received 6 November 1996; accepted 21 January 1997

0550-3213/97/$17.00 t~) 1997 Elsevier Science B.V. All rights reserved.

PACS: 02.40.-k; 04.65.+e; I 1.90.+t; 12.10.-g; 12.60.J; 12.90.+b

N-connection structure in higher order anisotropic superspaces. Section 8 is devoted to

In this section we establish the necessary terminology on supermanifolds (s-manifolds)

2.1. Supermanifolds and superbundles

To build up s-manifolds [88,58,120] one uses as basic structures Grassmann algebra

A superspace is introduced as a product

A n'k = A0× • • • × A o ×A, × . . . ×AI

which is the A-envelope of a Z2-graded vector space V"'k = Vo®Vt = ~ n G 7~k is

Functions of superspaces, differentiation with respect to Grassmann coordinates, su-

where fi are C°°-functions, a vector field on an (n, k)-dimensional Goo-supermanifold

where x = ( ~ , 0 ) = {x I = (.~i,0~)} are local (even, odd) coordinates. We shall use

for all f, g in Goo (U), and

[l, I'} = - ( _ ) Itlll'l [1,, I},

where a = 1,2 . . . . . m and h = 1,2 . . . . . l). We shall use indices A = (a,&),B =

the local coordinates in 8 and write their transformations as

x t' = x ~ ' ( x l ) , srank \-~-xl ] = ( n , k ) , (1)

yA' = y2' ( x ) y a , where Y2' (x) CG(m, l, A).

u = ( x , y ) = (Yc, O , ~ , ( ) = {u a = ( x l , y 1) = (X^i ,0 i ,y^i , ( i ) = (2i, x i , ~ i , y i ) } .

2.2. Distinguished vector superbundles

where bracketed indices are correspondingly split on Ufp)-components:

(A) = (A(I),A(2) . . . . . A ( p ) ) , (a) = (a(1),a(2) . . . . . a(p)),

A distinguished vector superbundle (dvs-bundle) glz) = (/~(z), ¢r(d),.y-(d), ~ ) , with

U(p) = (X, y(p) ) = (X, y(1), Y(2) . . . . . Y(p))

= (X,O,y(p),((p)) = (-~,0, y(I),~'(I),Y(2>,~'(2) ..... y(p),((p))

x ' ' = x t' ( x )1,

A/',) = Kaa! (u(p-1 ) ) y(p

In brief we write transforms (3) as

('~') = (KI1' '~(A)/C(AZ)~:'where K tr - O x l ' / O x / .

(the dual coordinate bases are denoted as

d (~) = ( d l , d ( a ) ) = ( d l , d ( a , ) , d (&) . . . . . d (A'-))

O(a ) -~ ( Cgl, C~(A} ) -~ ( C~I, Cg(A,), O( A2) . . . . . O( az) ) ""+ O(a)

0 _ KA'I O........~+ y A~ ~ +..._.~ ~,Ah ) A, o

Y-matrices from (6) are partial derivations of corresponding combinations of K-

In brief we denote respectively ds-coordinate transforms of coordinate bases (4) and

are characterized by matrices of the type

W e note that we obtain a s u p e r s y m m e t r i c generalization o f the M i r o n - A t a n a s i u [ 77]

For s = 1 the Oscl~l is the ts-bundle T/14.

77.szSl( x, y( l ) ' " .. , Y(P' ) ) = ( x , y ( 1 ) , . . . ,y(P2)).

3. Non-linear connections in dvs-bundles

3.1. N-connections in vs-bundles

Let us consider the definitions of N-connection structure [ 106] in a vs-bundle ~ =

tangent to ~ in the point u E ~ is locally represented as

( u , Y ) = (u'~,Y ~') = ( x l , y A , y * , Y A) = (sci, o ~ , p " , ( a , ? i , Y ~ , ? " , Y a ) .

A non-linear connection, N-connection, in vs-bundle g is a splitting on the left of the

O, ,V{, i Tg,, ,Tg/Vg, ,0, (7)

i.e. a morphism of vs-bundles N : T~ E V~ such that Noi is the identity on V~.

(HE, rE, E).

To a given N-connection we can associate a covariant s-derivation on M:

where SA are local independent sections of g, Y = YASA and X = Xtst.

A' tgXt' = A/IA'MA OM a' (X) yA.

I~IA/(x, y) -- o N A t ( x ' Y) (10)

For a vector field on g, Z = Z I 0-Trx

L)zn ~ [Zl ( aRt -~- ~[AIBB) ~- yBanA1 •

Another important characteristic of an N-connection is its curvature (N-connection

1-2 = ~1 j,2Ajdx! A dx ] Q Oy---

with local coefficients

where for simplicity we have written (_)IKIIJI = (_)IKJI

3.2. N-connections in dvs-bundles

In order to define an N-connection into dvs-bundle ~(z) we consider an s-subbundle