Multivariable Calculus Theorems

Why This Matters

These theorems aren't just abstract formulas. They're the connective tissue of vector calculus, linking line integrals, surface integrals, volume integrals, and derivatives into a coherent framework. You're being tested on your ability to recognize when to apply each theorem, how they relate to one another, and why certain conditions (like conservative fields or continuous partial derivatives) make computations dramatically simpler. The big picture: these results let you convert hard integrals into easier ones and reveal deep relationships between local behavior (derivatives, divergence, curl) and global behavior (integrals over boundaries).

Don't just memorize the theorem statements. Know what type of problem each theorem solves and how they form a hierarchy. Green's Theorem is a special case of Stokes' Theorem, which connects to the Divergence Theorem through the broader framework of differential forms. When you see an integral, ask yourself: Can I convert this to a simpler domain? Is this field conservative? That conceptual reflex is what separates strong exam performance from mere formula recall.

---

Fundamental Theorems Connecting Integrals and Derivatives

These theorems share a common structure: they relate an integral over a region to an integral over its boundary, reducing dimensional complexity and revealing when path independence applies.

Gradient Theorem (Fundamental Theorem of Line Integrals)

If $\mathbf{F} = \nabla f$ (meaning $\mathbf{F}$ is a conservative field with potential function $f$), then the line integral depends only on the endpoints:

$\int_C \mathbf{F} \cdot d\mathbf{r} = f(\mathbf{b}) - f(\mathbf{a})$

• Path independence is the key payoff. You skip parameterization entirely and just evaluate the potential function at the two endpoints.
• Conservative field test: If $\nabla \times \mathbf{F} = \mathbf{0}$ on a simply connected domain, then $\mathbf{F}$ has a potential function. The simply connected condition matters: on domains with holes, a zero curl field can still fail to be conservative.

Green's Theorem

Green's Theorem converts between a line integral around a closed curve $C$ and a double integral over the enclosed planar region $R$. It has two forms:

• Circulation-curl form: $\oint_C \mathbf{F} \cdot d\mathbf{r} = \iint_R \left( \frac{\partial Q}{\partial x} - \frac{\partial P}{\partial y} \right) dA$, where $\mathbf{F} = \langle P, Q \rangle$ and $C$ is traversed counterclockwise (positive orientation).
• Flux-divergence form: $\oint_C \mathbf{F} \cdot \mathbf{n} \, ds = \iint_R \left( \frac{\partial P}{\partial x} + \frac{\partial Q}{\partial y} \right) dA$, which relates outward flux across $C$ to the divergence over $R$.
• Area computation trick: Setting $P = -y$ and $Q = x$ gives $\text{Area} = \frac{1}{2} \oint_C (x \, dy - y \, dx)$. This lets you compute enclosed area from a line integral, which is useful when the boundary has a nice parameterization but the region doesn't.

Stokes' Theorem

Stokes' Theorem relates the surface integral of curl over an oriented surface $S$ to the line integral around its boundary curve $C = \partial S$:

$\iint_S (\nabla \times \mathbf{F}) \cdot d\mathbf{S} = \oint_C \mathbf{F} \cdot d\mathbf{r}$

• This generalizes Green's Theorem to arbitrary oriented surfaces in $\mathbb{R}^3$. Green's is just the special case where $S$ is a flat region in the $xy$-plane.
• Orientation must be consistent: the boundary curve's direction and the surface normal must agree via the right-hand rule. Getting this wrong flips the sign.
• Curl interpretation: $\nabla \times \mathbf{F}$ measures local rotation of the field. If $\nabla \times \mathbf{F} = \mathbf{0}$ everywhere on a simply connected domain, the field is conservative.

Divergence Theorem (Gauss's Theorem)

The Divergence Theorem converts a flux integral over a closed surface $S$ into a volume integral over the enclosed region $V$:

$\iiint_V \nabla \cdot \mathbf{F} \, dV = \oiint_S \mathbf{F} \cdot d\mathbf{S}$

• Use this when computing flux directly over $S$ is painful but the divergence $\nabla \cdot \mathbf{F}$ is simple (e.g., a constant or polynomial).
• Physical interpretation: divergence measures source/sink strength at a point. The theorem says the net outward flux through the boundary equals the total source strength inside the volume.
• The surface must be closed (no boundary edges). If it's not closed, you can't apply this theorem directly.

---

Differentiation Rules for Multivariable Functions

These theorems establish how derivatives behave when functions depend on multiple variables, ensuring consistency and enabling chain-rule computations.

Chain Rule for Multivariable Functions

When you have a composite function where the intermediate variables themselves depend on other variables, the multivariable chain rule tells you how to differentiate through the layers.

For $z = f(x, y)$ where $x = x(t)$ and $y = y(t)$:

$\frac{dz}{dt} = \frac{\partial f}{\partial x}\frac{dx}{dt} + \frac{\partial f}{\partial y}\frac{dy}{dt}$

• Tree diagram method: Draw the dependency structure. Each path from the output variable down to the variable you're differentiating with respect to contributes one term. You multiply along each path and sum all paths.
• Jacobian generalization: For vector-valued functions, the chain rule becomes matrix multiplication: $D(\mathbf{g} \circ \mathbf{f}) = D\mathbf{g} \cdot D\mathbf{f}$, where $D\mathbf{f}$ is the Jacobian matrix of $\mathbf{f}$.

Clairaut's Theorem (Symmetry of Mixed Partials)

If both mixed partial derivatives $\frac{\partial^2 f}{\partial x \partial y}$ and $\frac{\partial^2 f}{\partial y \partial x}$ exist and are continuous near a point, then they're equal at that point:

$\frac{\partial^2 f}{\partial x \partial y} = \frac{\partial^2 f}{\partial y \partial x}$

• The continuity requirement is not optional. Counterexamples exist (the classic one involves a piecewise function with $f(0,0) = 0$) where the mixed partials exist but differ because they aren't continuous.
• Practical use: This simplifies Hessian matrix construction (the Hessian is symmetric for smooth functions) and serves as a quick check on your partial derivative computations.

---

Approximation and Local Behavior

These results let you approximate complicated functions near a point and understand implicit relationships between variables.

Taylor's Theorem for Multivariable Functions

Just as single-variable Taylor series approximate a function near a point using derivatives, the multivariable version uses partial derivatives and the Hessian matrix:

$f(\mathbf{x}_0 + \mathbf{h}) \approx f(\mathbf{x}_0) + \nabla f(\mathbf{x}_0) \cdot \mathbf{h} + \frac{1}{2}\mathbf{h}^T H(\mathbf{x}_0) \, \mathbf{h} + \cdots$

• The first-order terms (gradient) give the best linear approximation (the tangent plane). The second-order terms involve the Hessian matrix $H$, whose entries are the second partial derivatives $H_{ij} = \frac{\partial^2 f}{\partial x_i \partial x_j}$.
• The quadratic term $\frac{1}{2}\mathbf{h}^T H \mathbf{h}$ is critical for classifying critical points via the second derivative test: the eigenvalues of $H$ (or its determinant and trace in 2D) tell you whether you have a local min, local max, or saddle point.
• Error bounds depend on higher-order derivatives over the region, which matters for numerical analysis applications.

Implicit Function Theorem

When you have a constraint equation $F(x, y) = 0$ and you want to know whether you can solve for $y$ as a function of $x$ near some point, this theorem gives the answer.

• Existence guarantee: If $F(a, b) = 0$ and $\frac{\partial F}{\partial y}(a, b) \neq 0$, then there exists a smooth function $y = g(x)$ defined near $x = a$ satisfying $F(x, g(x)) = 0$.
• Derivative formula: You can find $\frac{dy}{dx}$ without ever solving for $y$ explicitly:

$\frac{dy}{dx} = -\frac{\partial F/\partial x}{\partial F/\partial y}$

• Higher dimensions: For a system $\mathbf{F}(\mathbf{x}, \mathbf{y}) = \mathbf{0}$, the condition becomes that the Jacobian matrix of $\mathbf{F}$ with respect to the dependent variables $\mathbf{y}$ must be non-singular (invertible) at the point.

---

Optimization Under Constraints

This theorem provides the standard method for constrained optimization, appearing constantly in applications and exams.

Lagrange Multiplier Theorem

When you need to optimize $f(\mathbf{x})$ subject to a constraint $g(\mathbf{x}) = c$, the method of Lagrange multipliers says that at any constrained extremum (where $\nabla g \neq \mathbf{0}$):

$\nabla f = \lambda \nabla g$

• Geometric interpretation: At the optimal point, the level curves of $f$ are tangent to the constraint surface. If they weren't tangent, you could move along the constraint and still increase (or decrease) $f$.
• Solving the system: You get $n + 1$ equations (the $n$ component equations from $\nabla f = \lambda \nabla g$ plus the constraint $g(\mathbf{x}) = c$) in $n + 1$ unknowns (the $n$ coordinates plus $\lambda$).
• Multiple constraints: With $k$ constraints, use $k$ multipliers: $\nabla f = \lambda_1 \nabla g_1 + \lambda_2 \nabla g_2 + \cdots + \lambda_k \nabla g_k$.
• The multiplier $\lambda$ itself has meaning: it measures the sensitivity of the optimal value to small changes in the constraint constant $c$.

---

Quick Reference Table

---

Self-Check Questions

1. Which two theorems both relate circulation around a boundary to a "curl-type" integral, and what distinguishes their domains of application?

2. If you're given a vector field and asked whether a line integral is path-independent, which theorem justifies your answer, and what condition must you verify?

3. Compare the Divergence Theorem and Stokes' Theorem: what type of integral does each convert, and what differential operator appears in each?

4. An exam gives you $F(x, y, z) = 0$ and asks for $\frac{\partial z}{\partial x}$. Which theorem applies, and what must be nonzero for it to work?

5. You need to maximize $f(x, y)$ subject to $g(x, y) = k$. Write the system of equations you must solve, and explain geometrically why $\nabla f$ and $\nabla g$ must be parallel at the solution.