In differential geometry, the integration along fibers of a k-form yields a                     (        k        −        m        )                -form where m is the dimension of the fiber, via "integration".
Let                     π        :        E        →        B                 be a fiber bundle over a manifold with compact oriented fibers. If                     α                 is a k-form on E, then for tangent vectors wi's at b, let
                    (                  π                      ∗                          α                  )                      b                          (                  w                      1                          ,        …        ,                  w                      k            −            m                          )        =                  ∫                                    π                              −                1                                      (            b            )                          β                where                     β                 is the induced top-form on the fiber                               π                      −            1                          (        b        )                ; i.e., an                     m                -form given by: with                                                                         w                                  i                                            ~                                               the lifts of                               w                      i                                   to E,
                    β        (                  v                      1                          ,        …        ,                  v                      m                          )        =        α        (                  v                      1                          ,        …        ,                  v                      m                          ,                                                            w                                  1                                            ~                                      ,        …        ,                                                            w                                  k                  −                  m                                            ~                                      )        .                (To see                     b        ↦        (                  π                      ∗                          α                  )                      b                                   is smooth, work it out in coordinates; cf. an example below.)
Then                               π                      ∗                                   is a linear map                               Ω                      k                          (        E        )        →                  Ω                      k            −            m                          (        B        )                . By Stokes' formula, if the fibers have no boundaries, the map descends to de Rham cohomology:
                              π                      ∗                          :                  H                      k                                  (        E        ;                  R                )        →                  H                      k            −            m                                  (        B        ;                  R                )        .                This is also called the fiber integration.
Now, suppose                     π                 is a sphere bundle; i.e., the typical fiber is a sphere. Then there is an exact sequence                     0        →        K        →                  Ω                      ∗                          (        E        )                              →                          π                              ∗                                                              Ω                      ∗                          (        B        )        →        0                , K the kernel, which leads to a long exact sequence, dropping the coefficient                               R                         and using                               H                      k                                  (        B        )        ≃                  H                      k            +            m                                  (        K        )                :
                    ⋯        →                  H                      k                                  (        B        )                              →            δ                                    H                      k            +            m            +            1                                  (        B        )                              →                          π                              ∗                                                              H                      k            +            m            +            1                                  (        E        )                              →                          π                              ∗                                                              H                      k            +            1                                  (        B        )        →        ⋯                ,
called the Gysin sequence.
Let                     π        :        M        ×        [        0        ,        1        ]        →        M                 be an obvious projection. First assume                     M        =                              R                                n                                   with coordinates                               x                      j                                   and consider a k-form:
                    α        =        f                d                  x                                    i                              1                                                    ∧        ⋯        ∧        d                  x                                    i                              k                                                    +        g                d        t        ∧        d                  x                                    j                              1                                                    ∧        ⋯        ∧        d                  x                                    j                              k                −                1                                                    .                Then, at each point in M,
                              π                      ∗                          (        α        )        =                  π                      ∗                          (        g                d        t        ∧        d                  x                                    j                              1                                                    ∧        ⋯        ∧        d                  x                                    j                              k                −                1                                                    )        =                  (                      ∫                          0                                      1                                g          (          ⋅          ,          t          )                    d          t          )                                  d                      x                                          j                                  1                                                              ∧          ⋯          ∧          d                      x                                          j                                  k                  −                  1                                                                    .                From this local calculation, the next formula follows easily: if                     α                 is any k-form on                     M        ×        I        ,                
                              π                      ∗                          (        d        α        )        =                  α                      1                          −                  α                      0                          −        d                  π                      ∗                          (        α        )                where                               α                      i                                   is the restriction of                     α                 to                     M        ×        {        i        }                .
As an application of this formula, let                     f        :        M        ×        [        0        ,        1        ]        →        N                 be a smooth map (thought of as a homotopy). Then the composition                     h        =                  π                      ∗                          ∘                  f                      ∗                                   is a homotopy operator:
                    d        ∘        h        +        h        ∘        d        =                  f                      1                                ∗                          −                  f                      0                                ∗                          :                  Ω                      k                          (        N        )        →                  Ω                      k                          (        M        )        ,                which implies                               f                      1                          ,                  f                      0                                   induces the same map on cohomology, the fact known as the homotopy invariance of de Rham cohomology. As a corollary, for example, let U be an open ball in Rn with center at the origin and let                               f                      t                          :        U        →        U        ,        x        ↦        t        x                . Then                               H                      k                                  (        U        ;                  R                )        =                  H                      k                                  (        p        t        ;                  R                )                , the fact known as the Poincaré lemma.
Given a vector bundle π : E → B over a manifold, we say a differential form α on E has vertical-compact support if the restriction                     α                              |                                              π                              −                1                                      (            b            )                                   has compact support for each b in B. We written                               Ω                      v            c                                ∗                          (        E        )                 for the vector space of differential forms on E with vertical-compact support. If E is oriented as a vector bundle, exactly as before, we can define the integration along the fiber:
                              π                      ∗                          :                  Ω                      v            c                                ∗                          (        E        )        →                  Ω                      ∗                          (        B        )        .                The following is known as the projection formula. We make                               Ω                      v            c                                ∗                          (        E        )                 a right                               Ω                      ∗                          (        B        )                -module by setting                     α        ⋅        β        =        α        ∧                  π                      ∗                          β                .
Proof: 1. Since the assertion is local, we can assume π is trivial: i.e.,                     π        :        E        =        B        ×                              R                                n                          →        B                 is a projection. Let                               t                      j                                   be the coordinates on the fiber. If                     α        =        g                d                  t                      1                          ∧        ⋯        ∧        d                  t                      n                          ∧                  π                      ∗                          η                , then, since                               π                      ∗                                   is a ring homomorphism,
                              π                      ∗                          (        α        ∧                  π                      ∗                          β        )        =                  (                      ∫                                                            R                                                  n                                                              g          (          ⋅          ,                      t                          1                                ,          …          ,                      t                          n                                )          d                      t                          1                                …          d                      t                          n                                )                η        ∧        β        =                  π                      ∗                          (        α        )        ∧        β        .                Similarly, both sides are zero if α does not contain dt. The proof of 2. is similar.                     ◻