Page 127 - Vector Analysis
P. 127

§4.3 The Surface Integrals                                                                                             123

Since   JΨ  =    det  ([ ])    the  identity  above           implies  that    |JΨ|(?g2   ˝ Ψ)   =   ?g1,  so  we  conclude
                        DΨ ,

that

                               ż (f ˝ ψ1)?g1 dA = ż (f ˝ ψ2)?g2 dA .                                               (4.13)

                               V1 V2

Therefore, the surface integral of f over Σ is independent of the choice of parameterizations

of Σ. In particular, the surface area of a regular C 1-surface which can be parameterized by

a global parametrization is also independent of the choice of parameterizations.
    As noticed in Remark 4.45, the first fundamental form ?g associated with the parametriza-

tion tV, ψu can be viewed as the Jacobian of the map ψ. Therefore, we arrive at the con-
clusion that dS ‘‘ = ”?g dA. dS is called the surface element. Moreover, similar to the

reason provided in Remark 4.22, the surface integral of a positive continuous function f

over Σ, where f is considered as the mass density of the surface given by

                               f (x) = lim                        the mass of ψ(∆)
                                                   diam(∆)Ñ0  the surface area of ψ(∆)

                                       ψ´1 (x)P∆

is the total mass of the surface.
    Next, we study the surface area of general regular surfaces that cannot be parameterized

using a single pair tV, ψu. Let Σ Ď R3 be a regular surface, and tVi, ψiuiPI be a collection
of local parameterizations satisfying that for each p P Σ there exists i P I such that tVi, ψiu
is a local parametrization of Σ at p. If there exists a countable collection of non-negative
functions tζjujPJ defined on Σ such that

    1.  For  each     j  P J,  spt(ζj) ” the  closure         of  ␣x  P  Σ  ˇ  ζj (x)  ‰  0(  Ď  Vi  for  some i P I;
                                                                            ˇ

    2.  ř        ζj(x) = 1 for all  x P Σ,

            jPJ

then intuitively we can compute the surface area by

                                            żż

                                                              ÿ                                                    (4.14)

                                              dS =                     ζj dS ,

                                              Σ jPJ Σ

where the surface integral of ζj over Σ is defined by (4.12) since spt(ζj) Ď ψ(Vi) and ζj = 0

outside spt(ζj). In other words, each term on the right-hand side of (4.14) can be evaluated

by

                                    ż            ż            (ζj ˝ ψi)?gi dS .
                                       ζj dS =

                                    Σ Vi
   122   123   124   125   126   127   128   129   130   131   132