Higher-Order Partial Derivatives Explained Simply

Imagine the temperature inside a heated room depends on two things: how far east you stand (x) and how far north you stand (y). A first partial derivative, like ∂T/∂x, tells you how fast the temperature changes as you walk east while staying at the same north position. But what if you want to know how that rate of change itself changes — for example, whether the eastward warming speeds up or slows down as you keep walking? That is exactly what higher-order partial derivatives describe.

A higher-order partial derivative is simply a partial derivative taken more than once. If f(x, y) is a function of two variables, you first differentiate with respect to one variable and then differentiate the result again — possibly with respect to the same variable or the other one. These second (and third, fourth, …) derivatives are essential tools: they power the second-derivative test for maxima and minima, they appear in Taylor’s theorem for functions of several variables, and they sit at the heart of physics — the heat equation and the wave equation are both written with second-order partial derivatives.

Key Concepts

• First-order partials: fx = ∂f/∂x treats y as a constant; fy = ∂f/∂y treats x as a constant.
• Second-order partials: fxx = ∂²f/∂x² (differentiate twice in x), fyy = ∂²f/∂y² (differentiate twice in y).
• Mixed partials: fxy means differentiate first by x, then by y; fyx means differentiate first by y, then by x.
• Clairaut’s theorem: if fxy and fyx are continuous near a point, then fxy = fyx there — the order of differentiation does not matter.
• Higher orders: fxxy (x, then x, then y), fxyz for three variables, and so on — the same idea keeps going.

Formulas

• fxx = ∂/∂x (∂f/∂x); fyy = ∂/∂y (∂f/∂y)
• fxy = ∂/∂y (∂f/∂x); fyx = ∂/∂x (∂f/∂y)
• Clairaut: fxy = fyx whenever the mixed partials are continuous.
• Units: each differentiation divides the unit by the variable’s unit — if f is in metres and x in seconds, fxx has units m/s², an acceleration-like quantity.

Worked Example

Find all second-order partial derivatives of f(x, y) = x³y² + 3x²y − 2y.

1. First partials. fx = 3x²y² + 6xy (treat y as constant; −2y is constant in x, so its derivative is 0). fy = 2x³y + 3x² − 2.
2. fxx. Differentiate fx by x: ∂/∂x (3x²y² + 6xy) = 6xy² + 6y.
3. fxy. Differentiate fx by y: ∂/∂y (3x²y² + 6xy) = 6x²y + 6x.
4. fyx. Differentiate fy by x: ∂/∂x (2x³y + 3x² − 2) = 6x²y + 6x. This equals fxy, confirming Clairaut’s theorem.
5. fyy. Differentiate fy by y: ∂/∂y (2x³y + 3x² − 2) = 2x³.
6. Evaluate at (1, 1): fxy(1, 1) = 6 + 6 = 12, fxx(1, 1) = 6 + 6 = 12, fyy(1, 1) = 2.

Common Misconceptions

• “fxy and fyx are always equal.” Only when the mixed partials are continuous; strange counterexamples exist where the order changes the answer.
• “fxx means fx times fx.” No — it is differentiation applied twice, not squaring.
• “You must differentiate the original function from scratch for every second partial.” Work step by step instead: find fx once, then differentiate that result to get fxx and fxy.

Real-World Applications

• Heat equation: ∂T/∂t = k ∂²T/∂x² describes how temperature spreads through a metal rod.
• Wave equation: ∂²u/∂t² = c² ∂²u/∂x² governs vibrating strings, sound, and light.
• Optimization: the Hessian matrix of second partials classifies maxima, minima, and saddle points of surfaces.
• Economics: diminishing marginal returns are expressed through negative second partials of a production function.

Higher-order partial derivatives are just repeated differentiation — differentiate once, then differentiate the result. Keep track of which variable you differentiate at each step, remember that fxy means x first and then y, and use Clairaut’s theorem (mixed partials are equal when continuous) as a built-in check on your arithmetic.