Multivariable Calculus: Functions of Several Variables
Introduction
The physical world is three-dimensional, and few quantities of interest depend on only a single variable. Temperature varies with position and time. Fluid velocity depends on location within the flow. The gravitational field extends throughout space, changing magnitude and direction from point to point. Multivariable calculus provides the mathematical framework for analyzing such systems, extending the ideas of single-variable calculus to functions of several variables.
The leap from one to many dimensions is not merely additive — it introduces qualitatively new concepts. A function of two variables defines a surface rather than a curve. Its rate of change depends on direction, not just on which side of a point you examine. Integrals become multidimensional, computing volumes rather than areas. Vector fields add the richness of directional behavior to every point in space.
Partial Derivatives
For a function of several variables, the partial derivative measures the rate of change with respect to one variable while holding all others constant. Computationally, partial differentiation treats the other variables as constants and differentiates normally. Conceptually, it answers the question: if I change only x, how does the function respond?
Computing and Interpreting Partial Derivatives
The partial derivative ∂f/∂x at a point (a,b) equals the limit of [f(a+h,b) − f(a,b)] / h as h approaches zero. This is exactly the ordinary derivative of the function f(x,b) with respect to x at x = a. Geometrically, it is the slope of the tangent line to the surface z = f(x,y) in the x-direction at the point.
Higher-order partial derivatives extend naturally. The second partial derivative ∂²f/∂x² measures curvature in the x-direction. Mixed partial derivatives like ∂²f/∂x∂y measure how the rate of change in one direction varies as you move in another direction. Clairaut’s theorem states that mixed partial derivatives are equal when the function is sufficiently smooth, a result that simplifies many calculations and theoretical arguments.
The Gradient Vector
The gradient ∇f = (∂f/∂x, ∂f/∂y, ∂f/∂z) collects all first partial derivatives into a vector. The gradient points in the direction of steepest ascent of the function, and its magnitude gives the rate of change in that direction. Perpendicular to level surfaces, the gradient provides the normal vector needed for tangent plane calculations.
The gradient is central to optimization. At a local maximum or minimum, the gradient is zero — there is no direction of ascent or descent. Finding these critical points involves solving the system of equations ∇f = 0, then classifying each point using the second derivative test. The Hessian matrix H of second partial derivatives determines whether a critical point is a local maximum, local minimum, or saddle point. These techniques underpin optimization theory.
Directional Derivatives
The directional derivative D_u f = ∇f · u measures the rate of change in the direction of the unit vector u. It reaches its maximum in the direction of the gradient and its minimum in the opposite direction. The directional derivative being zero in all directions tangent to a level surface characterizes the tangent plane.
The gradient being orthogonal to level surfaces explains why contour maps reveal steep terrain through closely spaced contour lines. Hikers reading topographic maps intuitively use this relationship: closely spaced contours indicate steep slopes where the gradient magnitude is large.
Multiple Integrals
Just as the single integral accumulates a function over an interval, the double integral accumulates a function over a region in the plane, and the triple integral over a volume in space. These integrals compute volumes, average values, total mass, moments of inertia, and other accumulated quantities.
Double Integrals over Rectangular Regions
The double integral of f over a rectangle is defined as the limit of double Riemann sums. Divide the rectangle into small subrectangles, multiply each area by the function value at a sample point, sum, and take the limit as the mesh becomes finer. Iterated integration evaluates the double integral by integrating with respect to one variable at a time, treating the other variable as constant during each step.
Fubini’s theorem guarantees that the order of integration does not matter for continuous functions over rectangular regions. This flexibility often simplifies computation: one order may produce elementary integrals while the other requires special functions.
Integration over General Regions
Regions in the plane are rarely perfect rectangles. Integration over general regions is handled by describing the region with inequalities and setting up appropriate limits of integration. Type I regions are bounded by vertical lines and functions of x. Type II regions are bounded by horizontal lines and functions of y. Choosing the correct type simplifies the resulting integrals dramatically.
Polar coordinates offer an alternative for regions with circular symmetry. The transformation x = r cosθ, y = r sinθ introduces the Jacobian factor r, making integrals over disks, annuli, and sectors straightforward. Similarly, cylindrical and spherical coordinates simplify triple integrals over cylinders and spheres.
Change of Variables
The Jacobian determinant generalizes the factor r from polar coordinates to arbitrary coordinate transformations. When transforming from (x,y) to (u,v), the factor |∂(x,y)/∂(u,v)| accounts for the local stretching or compression of area. The same principle extends to three dimensions.
Change of variables transforms difficult integrals into manageable ones. The transformation to spherical coordinates for integrating over a sphere introduces the Jacobian factor ρ² sinφ, making triple integrals over spherical regions routine.
Vector Calculus
Vector calculus extends differentiation and integration to vector fields — functions that assign a vector to each point in space. Velocity fields in fluid dynamics, force fields in electromagnetism, and gradient fields from potential functions are all vector fields.
Line Integrals and Surface Integrals
Line integrals accumulate a vector field along a curve. The work done by a force field moving a particle along a path is the line integral of the force. Line integrals of gradient fields depend only on the endpoints, not the path taken — conservative vector fields are precisely those that are gradients of scalar potential functions.
Surface integrals accumulate a vector field across a surface, measuring the flux — the net flow of a quantity through the surface. The flux of an electric field through a closed surface relates to the enclosed charge (Gauss’s law), and the flux of a magnetic field through any closed surface is zero (magnetic monopoles do not exist).
The Divergence and Curl
The divergence ∇·F measures the outward flow of a vector field from a point — the net rate at which the quantity is expanding or contracting. Positive divergence indicates a source; negative divergence indicates a sink. The curl ∇×F measures the rotation or circulation of a vector field around a point. Zero curl indicates irrotational flow.
The divergence theorem relates the flux through a closed surface to the triple integral of divergence within the volume. Stokes’ theorem relates the line integral around a closed curve to the surface integral of curl across any surface bounded by that curve. These theorems unify the calculus of vector fields and provide the mathematical foundation for Maxwell’s equations of electromagnetism. Understanding vector calculus is essential for vector calculus applications in physics and engineering.
Path Independence and Conservative Fields
A vector field is conservative if its line integral is path independent — the work done depends only on the endpoints. For a conservative field, there exists a potential function φ such that F = ∇φ. The condition for conservatism is that the curl is zero everywhere (in simply connected regions).
Conservative fields have several equivalent characterizations: the line integral around any closed loop is zero, the field is the gradient of some potential function, and the curl vanishes everywhere. These properties make conservative fields particularly tractable, as their line integrals reduce to evaluating the potential at the endpoints.
The Laplacian
The Laplacian ∇²f = ∇·∇f = ∂²f/∂x² + ∂²f/∂y² + ∂²f/∂z² combines the gradient and divergence operators. It appears in the heat equation describing diffusion, the wave equation describing propagation, and Laplace’s equation describing steady-state potentials. Harmonic functions satisfying Laplace’s equation are everywhere smooth and take on their extreme values only on boundaries.
Applications in Science and Engineering
Multivariable calculus is indispensable across technical disciplines. In fluid dynamics, the Navier-Stokes equations describe velocity and pressure fields in terms of partial derivatives, divergence, and curl. In electromagnetism, Maxwell’s equations relate electric and magnetic fields through divergence and curl relationships. In quantum mechanics, the Schrödinger equation involves the Laplacian operator acting on the wavefunction.
Structural engineers use double integrals to compute the center of mass and moment of inertia of irregular shapes. Climate scientists model atmospheric circulation using partial differential equations derived from conservation laws. Economists use partial derivatives to analyze how changes in multiple inputs affect production output and profit.
The method of Lagrange multipliers optimizes functions subject to equality constraints, finding extrema where the gradient of the objective is parallel to the gradient of the constraint. This technique solves constrained optimization problems in economics (utility maximization subject to budget constraints), engineering (design optimization subject to material limits), and physics (principle of least action).
What is the difference between a partial derivative and an ordinary derivative? A partial derivative measures change with respect to one variable while holding all other variables constant. An ordinary derivative applies to functions of a single variable.
How does the gradient relate to level surfaces? The gradient at a point is perpendicular to the level surface passing through that point. It points in the direction of greatest increase of the function.
What is the physical meaning of divergence? Divergence measures the net outward flow of a vector field per unit volume at a point. Positive divergence indicates a source; negative divergence indicates a sink.
What does the divergence theorem state? The divergence theorem equates the flux of a vector field through a closed surface to the triple integral of its divergence over the enclosed volume.
Calculus Basics Guide — Vector Calculus — Optimization Theory