Partial Derivatives & Tangent Planes

Partial Derivatives and Tangent Planes

Partial Derivatives

For a function z = f(x, y), the partial derivative with respect to x, denoted as ∂f/∂x or f_x, measures how f changes as x varies while keeping y constant. Geometrically, this represents the slope of the surface in the direction of the x-axis.

Similarly, the partial derivative with respect to y, denoted as ∂f/∂y or f_y, measures the rate of change of f as y varies while keeping x constant. This represents the slope of the surface in the direction of the y-axis.

Tangent Planes

Just as a tangent line provides a linear approximation to a curve in 2D, a tangent plane provides a linear approximation to a surface in 3D. The tangent plane to the surface z = f(x, y) at a point (a, b, f(a, b)) is the plane that contains all the tangent lines to the surface at that point.

The equation of the tangent plane can be constructed using the partial derivatives at the point (a, b):

This equation shows that the tangent plane passes through the point (a, b, f(a, b)) and has slopes f_x(a, b) and f_y(a, b) in the x and y directions, respectively.

Key Concepts

• Linear Approximation: Near the point of tangency, the tangent plane closely approximates the surface. This is formalized by the concept of differentiability.
• Normal Vector: The vector formed by the partial derivatives and -1, specifically [f_x, f_y, -1], is orthogonal (perpendicular) to the tangent plane at the point (a, b).
• Gradients: The vector of partial derivatives, [f_x, f_y], is known as the gradient. It points in the direction of the steepest ascent on the surface.

Real-world Applications

• Optimization: Finding local maxima and minima on a surface often involves identifying points where the tangent plane is horizontal (i.e., both partial derivatives are zero).
• Physics: Tangent planes are used to approximate potential energy surfaces, helping analyze the stability of physical systems.
• Machine Learning: Gradient descent relies on calculating partial derivatives to find the minimum of a cost function in multi-dimensional parameter space.

Historical Context

The idea of using partial derivatives to study surfaces emerged as calculus was extended from functions of a single variable to functions of several variables in the 18th and 19th centuries. Mathematicians such as Clairaut, Euler, and Lagrange formalized many of the tools we now use to analyze multivariable functions, including the use of tangent planes as local linear approximations.

Tangent planes also became central in the development of differential geometry, where surfaces are studied using local linear objects (tangent spaces). The modern view of linearization in many dimensions generalizes this same idea far beyond simple graphs of functions z = f(x, y).

Related Concepts

• Gradient and Directional Derivatives: The gradient vector built from partial derivatives describes the direction of steepest ascent and is closely tied to the geometry of the tangent plane.
• Differentiability in Several Variables: The existence of a well-defined tangent plane is the geometric expression of differentiability for functions of multiple variables.
• Multivariable Taylor Approximations: Higher-order Taylor polynomials extend the first-order linear approximation given by the tangent plane to capture more subtle local behavior of the surface.
• Optimization and Critical Points: Methods for finding and classifying critical points on surfaces rely on partial derivatives and the structure of the tangent plane.