R R2R1

Oliver Knill, Harvard Summer School, 2010 ~ =

F~ dr 4y dydx = 2y 2|10 x|20 = 4.
C 0 0

Example (in class): Find the area of the region en-

closed by
Chapter 6. Integral Theorems sin(πt)2 2
~r(t) = h , t − 1i
t
for −1 ≤ t ≤ 1. To do so, use Greens theorem with the
vector field F~ = h0, xi.
Section 6.1: Greens theorem
Why is Greens theorem true? Look first at a smallR square G = [x, xR + ǫ] × [y, y + ǫ]. The
If F~ (x, y) = hP (x, y), Q(x, y)i is a vector field in two dimensions and C : ~r(t) = hx(t), y(t)i, t ∈ line ~ ǫ ǫ
R ǫ integral of F = hP,R Qiǫ
along the boundary is 0 P (x + t, y)dt + 0 Q(x + ǫ, y + t) dt −
[a, b] is a curve, the line integral 0
P (x + t, y + ǫ) dt − 0 Q(x, y + t) dt. This line integral measures the ”circulation” at the
Z Z b place (x, y). Because Q(x + ǫ, y) − Q(x, y) ∼ Qx (x, y)ǫR ǫ Rand P (x, y + ǫ) − P (x, y) ∼ Py (x, y)ǫ,
ǫ
~ =
F~ dr F~ (x(t), y(t)) · ~r ′ (t) dt the line integral is (Qx − Py )ǫ2 is about the same as 0 0 curl(F ) dxdy. All identities hold in
C a the limit ǫ → 0. To prove the statement for a general region G, chop it into small squares of
size ǫ. Summing up all the line integrals around the boundaries gives the line integral around
measures the work done by the field F~ along the path.
the boundary because in the interior, the line integrals cancel. Summing up the vorticities on
the rectangles is a Riemann sum approximation of the double integral.
The curl of a two dimensional vector field F~ (x, y) = hP (x, y), Q(x, y)i is defined as the scalar
field
curl(F )(x, y) = Qx (x, y) − Py (x, y) .
The curl(F ) measures the vorticity of the vector field. One can write ∇× F~ = curl(F~ ) because
the two dimensional cross product of (∂x , ∂y ) with F~ = hP, Qi is Qx − Py .

Examples.
1) For F~ (x, y) = h−y, xi we have curl(F )(x, y) = 2.
2) If F~ (x, y) = ∇f , then the curl is zero because if P (x, y) = fx (x, y), Q(x, y) = fy (x, y) and
curl(F ) = Qx − Py = fyx − fxy = 0 by Clairot’s theorem.

Green’s theorem (1827) If F~ (x, y) = hP (x, y), Q(x, y)i is a vector field and G is a region for
which the boundary C is a curve parametrized so that G is ”to the left”. Then
Example:
R RR Green’s theorem allows to express the coordinates of the centroid = center of mass
~ =
F~ · dr curl(F ) dxdy Z Z Z Z
C G
( x dA/A, y dA/A)
G G
George Green (1793-1841) was a physicist of the nineteenth century. He was a self-taught
mathematician and miller. His work greatly contributed to modern physics. using line integrals. With the vector field F~ = h0, x2 i we have
Z Z Z
Lets look at a special case, when F~ is a gradient field. In this case, both sides of Green’s x dA = ~ .
F~ dr
theorem
R are zero: RR G C
F ~ is zero by the fundamental theorem for line integrals. and
~ · dr curl(F ) · dA is zero
C G Example: An important application of Green is to compute area. With the vector fields
because
F~ (x, y) = hP, Qi = h−y, 0i or F~ (x, y) = h0, xi have vorticity curl(F~ )(x, y) = 1. The right hand
curl(F ) = curl(grad(f )) = 0.
side in Green’s theorem is the area of G:
Z
The identity curl(grad(f )) = 0 can be checked directly by writing it as ∇ × ∇f and using that
Area(G) = x(t)ẏ(t) dt .
the cross product of two identical vectors is 0. Just treat ∇ as a vector. C

Example: Find the line integral of F~ (x, y) = hx2 − y 2 , 2xyi = hP, Qi along the bound- Example: Let G be the region under the graph of a function f (x) on [a, b]. The line integral
ary of the rectangle [0, 2] × [0, 1]. Solution: curl(F~ ) = Qx − Py = 2y − 2y = −4y so that around the boundary of G is 0 from (a, 0) to (b, 0) because F~ (x, y) = h0, 0i there. The line

1 2
integral is also zero from (b, 0) to (b, f (b)) and (a, f (a)) to (a, 0) because N = 0. The line An other important derivative besides curl is the divergence of F~ = hP, Q, Ri is the scalar
Rb Rb
integral on (t, f (t)) is − a h−y(t), 0i · h1, f ′(t)i dt = a f (t) dt. Green’s theorem assures that field div(hP, Q, Ri) = ∇ · F~ = Px + Qy + Rz . The divergence of a vector field F~ = hP, Qi
this is the area of the region below the graph. is div(P, Q) = ∇ · F~ = Px + Qy . The curl measure the ”vorticity” of a field. The divergence
measures the ”expansion” of a field.
Finally, lets mention a theoretical application: if curl(F~ ) = 0 in a simply connected region
G, then F~ is conservative
R there: given a closed curve C in G, it encloses a region R. Greens If a field has zero curl everywhere, the field is called irrotational.
theorem assures that C F~ dr~ = 0. So F~ has the closed loop property in G and is therefore a If a field has zero divergence everywhere, the field is called incompressible.
gradient field there.
With the ”vector” ∇ = h∂x , ∂y , ∂z i, we can write curl(F~ ) = ∇ × F~ and div(F~ ) = ∇ · F~ . Writing
An engineering application of Greens theorem is the planimeter, a mechanical device for things with the ”Nabla vector” is called Nabla calculus. This works both in 2 and 3 dimen-
measuring areas. It had been used in medicine to measure the size of the cross-sections of sions.
tumors, in biology to measure the area of leaves or wing sizes of insects, in agriculture to
measure the area of forests, in engineering to measure the size of profiles. There is a vector The following combination of divergence and gradient often appears in physics:
field F~ associated to a planimeter which is obtained by placing a unit vector perpendicular to
∆f = divgrad(f ) = fxx + fyy + fzz .
the arm). One can prove that F~ has vorticity 1. The planimeter calculates the line integral of
F~ along a given curve. Green’s theorem assures it is the area. It is called the Laplacian of f . We can write ∆f = ∇2 f because ∇ · (∇f ) = div(grad(f ).
We can extend the Laplacian also to vectorfields with ∆F~ = (∆P, ∆Q, ∆R) and write ∇2 F~ .
While direct computations can verify the following identities, they become evident with Nabla
Section 6.2: Curl and Div calculus from formulas for vectors like ~v ×~v = ~0, ~v ·~v × w
~ = 0 or u × (v × w) = v(u · w) − (u · v)w.
We have seen that the curl of a 2D vector field is a scalar. In three dimensions, the curl of a
3D vector field F~ = hP, Q, Ri is defined as the vector field div(curl(F~ )) = 0. ∇ · ∇ × F~ = 0.
curl(P, Q, R) = hRy − Qz , Pz − Rx , Qx − Py i . curlgrad(F~ ) = ~0 ∇ × ∇F~ = ~0.
curl(curl(F~ )) = grad(div(F~ ) − ∆(F~ )). ∇ × ∇ × F~ = ∇(∇ · F~ ) − (∇ · ∇)F~ .
It is a vector. We can write curl(F~ ) = ∇ × F~ .
These identities can be used to answer questions like:
Note that the third component is just the curl of a 2D vector field F~ = hP, Qi is Qx − Py .
While the curl in 2 dimensions is a scalar field it is a vector in 3 dimensions. In n dimensions, ~ such that F~ = hx + y, z, y 2i = curl(G)?
Question: Is there a vector field G ~
it would have dimension n(n − 1)/2, the number of coordinate planes in n dimensions. Answer: No, because div(F~ ) = 1 is incompatible with div(curl(G)) ~ = 0.

The curl is often visualized using a ”paddle wheel”. If you place such a wheel into the field into The derivatives curl, div and grad appear in physics. For example in electrodynamics or fluid
the direction v, its rotation speed of the wheel measures the quantity F~ · ~v. Consequently, the dynamics, where partial differential equations describe phenomena in life like light, sound and
direction in which the wheel turns fastest, is the direction of curl(F~ ). Its angular velocity is weather. Here are examples:
the length of the curl. The wheel could actually be used to measure the curl of the vector field
div(B) = 0 No monopoles there are no magnetic monopoles.
at any point. In situations with large vorticity like in a tornado, one can ”see” the direction of
the curl near the vortex center. curl(E) = − 1c Bt Faraday’s law change of magnetic flux induces voltage
curl(B) = 1c Et + 4π
c
j Ampère’s law current or change of E produces magnetic field
div(E) = 4πρ Gauss law electric charges are sources for electric field
The constant c is the speed of light. In fluid dynamics, we are interested in the velocity field
F~ = ~v , the density ρ and the potential function f if it exists.
Incompressibility div(~v ) = 0 incompressible fluids, in 2D: v = grad(u)
Irrotational curl(~v ) = 0 irrotational fluids
Incompressible irrotational ∆(f ) = 0, ~v = ∇f incompressible, irrotational fluids
Continuity equation ρ̇ + div(ρ~v ) = 0 no fluid gets lost
Example: An example of a field, which is both incompressible and irrotational is obtained by
taking a function f which satisfies the Laplace equation ∆f = 0, like f (x, y) = x3 − 3xy 2 , then
look at its gradient field F~ = ∇f . In that case, this gives
F~ (x, y) = h3x2 − 3y 2 , −6xyi .

3 4
 Section 6.3: Flux and Stokes Theorem
If S is a surface in space, and F~ is a vector field. We are going to define the flux of F~ through
the surface S.

In dimension d, there are d fundamental derivatives:

1 7→ 1 f ′ derivative
Assume S is parametrized as ~r(u, v) = hx(u, v), y(u, v), z(u, v)i over a domain G in the uv-
1 7→ 2 ∇f gradient
plane.
2 7→ 1 ∇×F curl

1 7→ 3 ∇f gradient
3 7→ 3 ∇ × F curl The flux integral of F~ through S is defined as the double integral
Z Z
3 7→ 1 ∇ · F divergence
F~ (~r(u, v)) · (~ru × ~rv ) dudv .
~ G
Example: If we rotate the vector R field F = hP, Qi by 90 Rdegrees = π/2, we get a new vector
~ = h−Q, P i. The integral
field G F · ds becomes a flux γ G · dn of G through the boundary With the short hand notation dS ~ = (~ru × ~rv ) dudv which represents an infinitesimal normal
C RR
vector to the surface. this can be written as ~
F~ · dS.
of R, where dn is a normal vector with length |r ′ |dt. With div(F~ ) = (Px + Qy ), we see that S

curl(F~ ) = div(G)
~ .
dS
Green’s theorem now becomes
Z Z Z
~ dxdy =
div(G) G ~ ,
~ · dn F
R C

where dn(x, y) is a normal vector at (x, y) orthogonal to the velocity vector ~r ′ (x, y) at (x, y).
This new theorem has a generalization to three dimensions, where it is called Gauss theorem or RR
divergence theorem. Don’t treat this however as a different theorem in two dimensions. Note The interpretation is that if F~ = fluid velocity field, then ~ is the amount of fluid
F~ · dS
S
that this is just Green’s theorem disguised. There are only 2 basic integral theorems in the passing through S in unit time.
plane: Green’s theorem and the fundamental theorem of line integrals. The just mentioned
variant of Greens theorem can explains however divergence means in two dimensions: it is an Example: Let F~ (x, y, z) = h0, 1, z 2 i and let S be the upper half sphere parametrized by
expansion rate similar than the original Green theorem explains that curl is a rotation rate:
~r(u, v) = hcos(u) sin(v), sin(u) sin(v), cos(v)i
Take a small square in the plane. Greens theorem tells us that the curl of F integrated over that
square is the line integral along the boundary. The divergence version of Greens theorem tells us , ~ru × ~rv = sin(v)~r so that F~ (~r(u, v)) = (0, 1, cos2 (v)) and
that the divergence of F~ integrated over the square is the flux of the field through the boundary. Z 2π Z π
h0, 1, cos2 (v)i · hcos(u) sin2 (v), sin(u) sin2 (v), cos(v) sin(v)i dudv .
The curl at a point (x, y) is the average work of the field done along a small circle of radius r 0 0
around the point in the limit when the radius of the circle goes to zero. R 2π R π
The integral is 0 π/2 sin2 (v) sin(u) + cos3 (v) sin(v) dudv. The first integral is zero. The sec-
Rπ π/2
The divergence at a point (x, y) is the average flux of the field through a small circle of radius ond is π/2 cos3 v sin(v) dv = − cos4 (v)/4|0 = 1/4.
r around the point in the limit when the radius of the circle goes to zero.

5 6
Because ~n = ~ru × ~rv /|~ru × ~rv | is a unit vector normal to the surface and on the surface, F~ · ~n
is the normal
R R componentR of R the vector field with respect to the surface. One could write there-
fore also ~ =
F~ · dS F~ · ~n dS where dS is the surface element we know from when we
S
computed surface area. TheR Rfunction F~ · ~n is the scalar projection of F~ in the normal direc-
tion. Whereas the formula 1 dS gave the area of the surface with dS = |~ru × ~rv |dudv, the
flux integral weights each area element dS with the normal component of the vector field with
F~ (~r(u, v) · ~n(~r(u, v)). For actual computations, we of course refrain from this, since there is
more work to compute ~n. We just compute the vectors F~ (~r(u, v)) and ~ru × ~rv and integrate its
dot product over the domain.

Example. Calculate the flux of the vector field F~ (x, y, z) = h1, 2, 4zi through the paraboloid
z = x2 + y 2 lying above the region x2 + y 2 ≤ 1.
We can parametrize with polar coordinates ~r(r, θ) = hr cos(θ), r sin(θ), r 2 i where ~rr × ~rθ = We see that for a surface which is flat, Stokes theorem is a consequence of Green’s theorem.
R R R If we put the coordinate axis so that the surface is in the xy-plane, then the vector field F
~ = 2π 1 (−2r 2 cos(v)−
h−2r 2 cos(θ), −2r 2 sin(θ), ri and F~ (~r(u, v)) = h1, 2, 4r 2i. We get S F~ ·dS 0 0
4r 2 sin(v) + 4r 3 ) drdθ = 2π. induces a vector field on the surface such that its 2D curl is the normal component of curl(F ).
The reason is that the third component Qx − Py of curl(F~ )hRy − Qz , Pz − Rx , Qx − Py i is the
Concept problem: Compute the flux of F~ (x, y, z) = h2, 3, 1i through the torus parameterized two dimensional curl: F~ (~r(u, v)) · h0, 0, 1i = Qx − Py . If C is the boundary of the surface, then
RR R
as ~r(u, v) = h(2 + cos(v)) cos(u), (2 + cos(v)) sin(u), sin(v)i, where both u and v range from 0 S
F~ (~r(u, v)) · h0, 0, 1i dudv = C F~ (~r(t)) ~r ′ (t)dt.
to 2pi. A general surface can be approximated by a polyhedron consisting of a mesh of small triangles.
No computation is needed. Think about what the flux means. When summing up line integrals along all these parallelograms, the line integrals inside the
surface cancel and only the integral along the boundary remain. On the other hand, the sum
The following theorem is a generalization of the fundamental theorem of calculus in which many of the fluxes of the curl through boundary adds up to the flux through the surface.
of the concepts of multivariable calculus come together.
Stokes theorem: Let S be a surface with boundary curve C and let F be a vector field. Then Stokes theorem was found by Ampère in 1825. George Gabriel Stokes (1819-1903) was
Z Z Z probably inspired by work of Green and rediscovers the identity around 1840.
~ =
curl(F~ ) · dS ~ .
F~ · dr
S C

Note that the orientation of C is such that if you walk along the surface. If your head points
into the direction of the normal ~ru × ~rv , then the surface to your left. G

Example: Let F~ (x, y, z) = h−y, x, 0i and let S be the upper semi hemisphere, then curl(F~ )(x, y, z) =
h0, 0, 2i. The surface is parameterized by ~r(u, v) = hcos(u) sin(v), sin(u) sin(v), cos(v)i on
G = [0, 2π]×[0, π/2] and ~ru ×~rv = sin(v)~r(u, v) so that curl(F~ )(x, y, z)·~ru ×~rv = cos(v) sin(v)2.
R 2π R π/2 Example: Calculate the flux of the curl of F~ (x, y, z) = h−y, x, 0i through the surface param-
The integral 0 0 sin(2v) dvdu = 2π.
eterized by ~r(u, v) = hcos(u) cos(v), sin(u) cos(v), cos2 (v) + cos(v) sin2 (u + π/2)i. Because the
~ = ~r ′ (t)dt = surface has the same boundary as the upper half sphere, the integral is again 2π as in the above
The boundary C of S is parameterized by ~r(t) = hcos(t), sin(t), 0i so that dr
R example.
~ 2 2
h− sin(t), cos(t), 0i dt and F (~r(t)) ~r (t)dt = sin(t) + cos (t) = 1. The line integral C F · dr
′ ~ ~
along the boundary is 2π. For every surface bounded by C the flux of curl(F~ ) through through the surface is the same.
The flux of the curl of a vector field through a surface S depends only on the boundary of S.
If S is a surface in the xy-plane and F~ = hP, Q, 0i has zero z component, then curl(F~ ) = Compare this with the earlier statement that for every curve between two points A, B the line
~ = Qx − Py dxdy.
h0, 0, Qx − Py i and curl(F~ ) · dS integral of grad(f ) along C is the same. The line integral of the gradient of a function of a
curve C depends only on the end points of C.

Stokes theorem is important in physics. Here is an example: The electric field E and the
magnetic field B are linked by a Maxwell ~ 1
R R equation curl(E) = − c Ḃ. Take a closed wire C
which bounds a surface S and consider B · dS, the flux of the magnetic field through S.
S RR RR
Its change can be related with a voltage using Stokes theorem: d/dt S B · dS = S
Ḃ · dS =

7 8
RR R
−c curl(E) ~ = −c E
~ · dS ~ = U, where U is the voltage measured at the cut-up wire.
~ dr G, the partial differential equation ρ̇ + div(F~ ) holds. It is called the continuity equation. It
S C
It means that if we change the flux of the magnetic field through the wire, then this induces gives a meaning to the expression div(F~ ) as the rate of change of ”density” at this point. Take
a voltage. The flux can be changed by changing the amount of the magnetic field but also by a a small cube and let it flow with the field. If it expands in volume, then the divergence is
changing the direction. If we turn around a magnet around the wire, we get an electric voltage. positive, the density becomes smaller.
This happens in a power-generator like an alternator in a car. In practical implementations,
the wire is turned inside a fixed magnet. Example. What is the flux of the vector field F~ (x, y, z) = h2x, 3z 2 + y, sin(x)i through the
solid G = [0, 3] × [0, 3] × [0, 3] \ ([0, 3] × [1, 2] × [1, 2] ∪ [1, 2] × [0, 3] × [1, 2] ∪ [0, 3] × [0, 3] × [1, 2])
which is a cube where three perpendicular cubic holes have been removed.
RRR RRR
Answer: Use the divergence theorem: div(F~ ) = 2 and so G
div(F~ ) dV = 2 G
dV =
2Vol(G) = 2(27 − 7) = 40. Note that the flux integral here would be over a complicated surface
over dozens of rectangular planar regions.

It is good to see why the divergence theorem is true: consider a small box [x, x + dx] × [y, y +
dy] × [z, z + dz]. The flux of F~ = hP, Q, Ri through the faces perpendicular to the x-axes is
[F~ (x + dx, y, z) · h1, 0, 0i + F
~ (x, y, z) · h−1, 0, 0i]dydz = P (x + dx, y, z) − P (x, y, z) = Px dxdydz.
Similarly, the flux through the y-boundaries is Py dydxdz and the flux through the two z-
boundaries is Pz dzdxdy. The total flux through the faces of the cube is (Px +Py +Pz ) dxdydz =
div(F~ ) dxdydz. A general solid can be approximated as a union of small cubes. The sum of the
fluxes over all the cubes is the sum of the fluxes through the faces which do not touch an other
box. Fluxes through adjacent sides cancel. The sum of all the infinitesimal fluxes of the cubes
is the flux through the boundary of the R R union.
R The sum of all the div(F~ ) dxdydz is a Riemann
sum approximation for the integral G
div( F~ ) dxdydz. In the limit, where dx, dy, dz goes
to zero, we obtain the divergence theorem.

Section 6.4: Divergence theorem

There are three integral theorems in three dimensions. Besides the fundamental theorem of
line integrals and Stokes theorem, there is the divergence theorem:
Divergence theorem. Let G be a solid region in space. Assume the boundary of the solid is
a surface S. Let F~ be a vector field. Then
Z Z Z Z Z
div(F~ ) dV = F~ · dS .
G S

The orientation of S is such that the normal vector ~ru × ~rv points outside the solid G.
Example. Find the flux of curl(F ) through a torus if F~ = hyz 2 , z + sin(x) + y, cos(x)i and the
torus has the parametrization
Example: Let F~ (x, y, z)R=R hx, ~
R y, zi and let S be sphere. The divergence of F is the constant
~
function div(F ) = 3 and ~
div(F ) dV = 3 · 4π/3 = 4π. The flux through the boundary is ~r(θ, φ) = h(2 + cos(φ)) cos(θ), (2 + cos(φ)) sin(θ), sin(φ)i .
RR RR G R π R 2π
S
~r · ~ru × ~rv dudv = S
|~r(u, v)|2 sin(v) dudv = 0 0 sin(v) dudv = 4π also. Answer: The answer is 0 because the divergence of curl(F ) is zero. By the divergence theorem,
RR the flux is zero.
What does it mean? Assume ρ is the density of a fluid with velocity field F~ , The flux S
F~ ·dS
of the fluid through the
RRRclosed surface S bounding a region G is the minus the rate of change Similarly as Green’s theorem allowed to calculate the area of a region by passing along the
of the mass −(d/dt) G
ρdV inside G. This is because the flux through the boundary boundary, the volume of a region can be computed as a flux integral: Take for example the
is
R Rthe
R amount of mass leavingR G R Rin unit time. But this flux is by Gauss theorem equal to vector field F~ (x, y, z) = hx, 0, 0i which has divergence 1. The flux of this vector field through
div(F~ ) dV . Therefore, ( d ρ + div(F~ ) dV = 0. Because this is true for all regions RR
G G dt the boundary of a solid region is equal to the volume of the solid: ~ = Vol(G).
hx, 0, 0i · dS
δG

9 10
 Fundamental theorem of line integrals
If C : ~r(t), t ∈ [a, b] is a curve and f is a function. Then
Z
~ = f (~r(b)) − f (~r(a))
∇f · dr
C

Consequences. R
1) For closed curves, the line integral C ∇f · dr ~ is zero.
2) Gradient fields are path independent: if F~ = ∇f , then the line integral between two
points P and Q does not depend on the path connecting the two points.
3) The theorem holds in any dimension. In one dimension, it reduces to the fundamental
Rb
theorem of calculus a f ′ (x) dx = f (b) − f (a)
4) The theorem justifies the name conservative for gradient vector fields.
5) The term ”potential” was coined by George Green who lived from 1783-1841.
Example. Let f (x, y, z) = x2 + y 4 + z. Find the line integral of the vector field
F~ (x, y, z) = ∇f (x, y, z) along the path ~r(t) = hcos(5t), sin(2t), t2 i from t = 0 to t = 2π.
Example: How heavy are we, at distance r from the center of the earth?
Answer: The law of gravity can be formulated as div(F~ ) = 4πρ, where ρ is the mass density. Solution. ~r(0) = h1, 0, 0i and ~r(2π) = h1, 0, 4π 2i and f (~r(0)) = 1 and f (~r(2π)) = 1 + 4π 2 . The
R
We assume that the earth is a ball of radius R. By rotational symmetry, the gravitational fundamental theorem of line integral gives C ∇f dr ~ = f (r(2π)) − f (r(0)) = 4π 2 .
force is normal to the surface: F~ (~x) = F~ (r)~x/||~x||. The flux of F~ through a ball of radius r is
RR RRR
~ = 4πr 2 F~ (r). By the divergence theorem, this is 4πMr = 4π
F~ (x) · dS ρ(x) dV ,
Sr Br
where Mr is the mass of the material inside Sr . We have (4π)2 ρr 3 /3 = 4πr 2 F~ (r) for r < R and Green’s theorem. If R is a region in the plane with boundary C and F~ = hP, Qi is a vector
(4π)2 ρR3 /3 = 4πr 2 F~ (r) for r ≥ R. Inside the earth, the gravitational force F~ (r) = 4πρr/3. field, then
Outside the earth, it satisfies F~ (r) = M/r 2 with M = 4πR3 ρ/3. Z Z Z
curl(F~ ) dxdy = ~ .
F~ · dr
R C

Remarks.
1
1) It can be used to switch from double integrals to one dimensional integrals.
0.8
2) The curve is oriented in such a way that the region is to the left.
0.6
3) The boundary of the curve can consist of piecewise smooth pieces.
Rb
0.4
4) If C : t 7→ ~r(t) = hx(t), y(t)i, the line integral is a hP (x(t), y(t)), Q(x(t), y(t))i ·
0.2
hx (t), y (t)i dt.
′ ′

0.5 1 1.5 2 5) Green’s theorem was found by George Green (1793-1841) in 1827 and by Mikhail Ostro-
gradski (1801-1862).
6) If curl(F~ ) = 0 in a simply connected region, then the line integral along a closed curve is
zero. If two curves connect two points then the line integral along those curves agrees.
We are now at the end of the course. Lets have an overview over the integral theorems and 7) Taking F~ (x, y) = h−y, 0i or F~ (x, y) = h0, xi gives area formulas.
give a typical example in each case.
Example. Find the line integral of the vector field F~ (x, y) = hx4 + sin(x) + y, x + y 3 i along
The fundamental theorem for line integrals, Green’s theorem, StokesR theoremR and divergence the path ~r(t) = hcos(t), 5 sin(t) + log(1 + sin(t))i, where t runs from t = 0 to t = π.
theorem are all incarnation of one single theorem. It has the form A dF = δA F , where dF
is a exterior derivative of F and where δA is the boundary of A. They all generalize the Solution. curl(F~ ) = 0 implies that the line integral depends only on the end points
fundamental theorem of calculus. (0, 1), (0, −1) of the path. Take the simpler path ~r(t) = h−t, 0i, −1 ≤ t ≤ 1, which has velocity
R1
~r ′ (t) = h−1, 0i. The line integral is −1 ht4 − sin(t), −ti · h−1, 0i dt = −t5 /5|1−1 = −2/5.

Remark We could also find a potential f (x, y) = x5 /5 − cos(x) + xy + y 5/4. It has the
property that grad(f ) = F . Again, we get f (0, −1) − f (0, 1) = −1/5 − 1/5 = −2/5.

11 12
Stokes theorem. If S is a surface in space with boundary C and F~ is a vector field, then there are d theorems. We have here seen the situation in dimension 2 and 3, but it would
Z Z Z continue. The fundamental theorem of line integrals generalizes directly to higher dimensions:
curl(F~ ) · dS = ~
F~ · dr just define line integrals in d dimensions. Also the divergence theorem (aka Gauss theorem)
S C generalizes directly since an n-dimensional integral in n dimensions is obvious to do as is the
concept of divergence. The generalization of curl and flux is more suble, since in 4 dimensions
Remarks. already, the curl of a vector field is a 6 dimensional object. It is a d(d − 1)/2 dimensional object
1) Stokes theorem allows to derive Greens theorem: if F~ is z-independent and the surface S is in general.
contained in the xy-plane, one obtains the result of Green.
2) The orientation of C is such that if you walk along C and have your head in the direction dimension 1D objects 2D objects 3D objects 4D objects
of the normal vector ~ru × ~rv , then the surface to your left.
3) Stokes theorem was found by André Ampère (1775-1836) in 1825 and rediscovered by 1 FTC - - -
George Stokes (1819-1903). 2 FTL Green - -
4) The flux of the curl of a vector field does not depend on the surface S, only on the boundary
of S. This is analogue to the fact that the line integral of a gradient field only depends on the 3 FTL Stokes Gauss -
end points of the curve. 4 FTL Stokes 1 Stokes 2 Gauss
5) The flux of the curl through a closed surface like the sphere is zero: the boundary of such a
surface is empty. To formulate calculus in dimensions larger than 3, one needs also to generalize the derivatives
grad, curl and div. Look at the Pascal triangle: in one dimensions, there is a derivative
Example. Compute the line integral of F~ (x, y, z) = hx3 + xy, y, zi along the polygonal path
f (x) → f ′ (x) from scalar to scalar functions. It corresponds to the entry 1 − 1 in the Pascal
C connecting the points (0, 0, 0), (2, 0, 0), (2, 1, 0), (0, 1, 0).
triangle. The next entry 1 − 2 − 1 corresponds to differentiation in two dimensions, where we
have the gradient f → ∇f mapping a scalar function to a vector field with 2 components. And
Solution. The path C bounds a surface S : ~r(u, v) = hu, v, 0i parameterized by R =
we have a curl, F → curl(F ) which corresponds to the transition 2 − 1. The situation in three
[0, 2] × [0, 1]. By Stokes theorem, the line integral is equal to the flux of curl(F~ )(x, y, z) = dimensions is captured by the entry 1 − 3 − 3 − 1 in the Pascal triangle. The first derivative
h0, 0, −xi through S. The normal vector of S is ~ru × ~rv = h1, 0, 0i × h0, 1, 0i = h0, 0, 1i so that
RR R R R R 1 − 3 is the gradient. The second derivative 3 − 3 is the curl and the third derivative 3 − 1
~ = 2 1 h0, 0, −ui · h0, 0, 1i dudv = 2 1 −u dudv = −2.
curl(F~ ) dS
S 0 0 0 0 is the divergence. In four dimensions, we would have to look at 1 − 4 − 6 − 4 − 1. The first
derivative 1 − 4 is still the gradient. Then we have a first curl, which maps a vector field with
4 components into an object with 6 components. Then there is a second curl, which maps an
Divergence theorem If S is the boundary of a region B in space with boundary S and F~ is object with 6 components back to a vector field, we would have to look at 1 − 4 − 6 − 4 − 1. The
a vector field, then first derivative 1 − 4 is still the gradient. Then we have a first curl 4 − 6, which maps a vector
Z Z Z Z Z field with 4 components into an object with 6 components. Then there is a second curl, which
div(F~ ) dV = ~ .
F~ · dS maps such so called 2 forms back to vector fields 6 − 4 and then there is the divergence 4 − 1
B S which maps a vector field back to a scalar field. When setting up calculus in higher dimensions,
one talks about differential forms instead of scalar fields vector fields. If F~ is a one form,
Remarks. then curl(F~ ) is a 2-form, which happens in three dimensions to be a vector field again. The
1) The divergence theorem is also called Gauss theorem. divergence of a 2-form is a 3 form, which is a scalar field again. The general formalism defines
2) It can be helpful to determine the flux of vector fields through surfaces. a derivative d called exterior derivative on differential forms as well as integration of such k
3) It was discovered in 1764 by Joseph Louis Lagrange (1736-1813), later it was rediscovered forms on k dimensional objects. There is a boundary operation δ which maps a k-dimensional
by Carl Friedrich Gauss (1777-1855) and by George Green. object into a k − 1 dimensional object. This boundary operation is dual to differentiation and
4) For divergence free vector fields F~ , the flux through a closed surface is zero. Such fields F~ they both satisfy the same relation dd(F ) = 0 and δδG = 0. Differentiation and integration
are also called incompressible or source free. are linked by the general Stokes theorem

Example. Compute the flux of the vector field F~ (x, y, z) = h−x, y, z 2 i through the boundary R R
S of the rectangular box [0, 3] × [−1, 2] × [1, 2].
Solution. By Gauss theorem, the flux is equal to the triple integral of div(F ) = 2z over the δG F = G dF
R3R2 R2
box: 0 −1 1 2z dxdydz = (3 − 0)(2 − (−1))(4 − 1) = 27. which becomes a single theorem generalizing the fundamental theorem of calculus. It is the
fundamental theorem of multivariable calculus.
Finally, lets mention how all these theorems fit together. It turns out that in d-dimensions

13 14