Divergence Theorem
Chapter 2 — Vector Calculus
The divergence theorem (also called Gauss's theorem, not to be confused with Gauss's law) converts a volume integral of a divergence into a surface integral of flux:
\[\int_V (\vec{\nabla}\cdot\vec{F})\,dV = \oint_{\partial V} \vec{F}\cdot d\vec{A}.\]
Here \(V\) is any well-behaved volume, \(\partial V\) is its closed bounding surface, and \(d\vec{A}\) is an outward-pointing area element.
Why it works (the tiling argument)
Chop the volume \(V\) into many tiny sub-volumes \(V_i\). The divergence is defined so that, for each tiny piece,
\[(\vec{\nabla}\cdot\vec{F})_i\,V_i \approx \oint_{\partial V_i} \vec{F}\cdot d\vec{A}.\]
Now add up all the sub-volumes. On the left you get a Riemann sum for \(\int_V(\vec{\nabla}\cdot\vec{F})\,dV\). On the right, every internal face is shared between two neighboring cells with oppositely-oriented outward normals — those contributions cancel in pairs. Only the external faces survive, and they assemble into \(\oint_{\partial V}\vec{F}\cdot d\vec{A}\). Taking the limit gives the theorem.
How to read it
The left side is a volume integral of a local scalar (divergence). The right side is a surface integral of the field itself (flux). The theorem says these are always equal — the total "source content" inside a volume equals the net outflow through its boundary.
You can use the theorem in either direction:
- Volume → surface: if you know \(\vec{\nabla}\cdot\vec{F}\) and want a flux, integrate over the enclosed volume. (Example: Gauss's law for a uniform \(\rho\) inside a sphere.)
- Surface → volume: if you're given a flux integral that's hard to compute directly, convert it into \(\int_V(\vec{\nabla}\cdot\vec{F})\,dV\).
Worked example
Let \(\vec{F} = \vec{r} = (x,y,z)\) and take \(V\) to be a sphere of radius \(R\) centered at the origin.
Volume side: we already know \(\vec{\nabla}\cdot\vec{r} = 3\), so
\[\int_V 3\,dV = 3\cdot\frac{4}{3}\pi R^3 = 4\pi R^3.\]
Surface side: on the sphere \(\vec{r}\cdot d\vec{A} = r\,dA = R\,dA\), so
\[\oint_{\partial V}\vec{r}\cdot d\vec{A} = R\cdot(4\pi R^2) = 4\pi R^3.\]
They match.
Relation to Gauss's law
Applying the divergence theorem to the electric field gives
\[\oint_{\partial V}\vec{E}\cdot d\vec{A} = \int_V (\vec{\nabla}\cdot\vec{E})\,dV.\]
The left side is \(Q_{\rm enc}/\epsilon_0\) by the integral Gauss's law. The right side is the volume integral of whatever local quantity equals the charge density divided by \(\epsilon_0\). Equating integrands (valid because \(V\) is arbitrary) gives the differential form of Gauss's law, \(\vec{\nabla}\cdot\vec{E} = \rho/\epsilon_0\).
Practice Problems
Hint
Solution
\(\vec{\nabla}\cdot\vec{F} = 1\), so the volume integral is just the volume: \(\int_V 1\,dV = 1\).
Answer: \(1\).
Hint
Solution
\(\vec{\nabla}\cdot\vec{F}=0\) for any constant \(\vec{F}\). The divergence theorem then says \(\oint\vec{F}\cdot d\vec{A}=\int_V 0\,dV = 0\).
Answer: flux of a uniform field through any closed surface is zero.
Hint
Solution
\(\int_V 3r^2\,dV = 3\int_0^R r^2\cdot 4\pi r^2\,dr = 12\pi\cdot R^5/5 = \frac{12\pi R^5}{5}\).
Answer: \(\frac{12\pi R^5}{5}\).
Hint
Solution
\(\oint\vec{F}\cdot d\vec{A} = \int_V 2\,dV = 2V_0\).
Answer: \(2V_0\).
Hint
Solution
The divergence theorem requires \(\vec{F}\) to be smooth inside \(V\). For the Coulomb field, the "missing" flux of \(4\pi k\) is concentrated in a delta-function divergence at the origin: treating \(\vec{\nabla}\cdot(\hat{r}/r^2) = 4\pi\delta^3(\vec{r})\) makes the theorem work. Physically, this delta function is the point source — it's exactly \(\rho/\epsilon_0\) for a point charge.
Answer: the divergence isn't zero at the origin; it contains a delta function representing the point source.