Chain Rule in Multivariable Settings

students, imagine checking the temperature of a metal plate 🔥 while also moving across it and changing your path at the same time. The temperature depends on position, and position depends on time. This is exactly the kind of situation where the multivariable chain rule becomes useful. In this lesson, you will learn how to track change through multiple layers of variables, connect the chain rule to tangent planes and linear approximation, and use the idea in realistic problems.

What the Multivariable Chain Rule Means

The chain rule is a method for finding how one quantity changes when it depends on other quantities that are also changing. In single-variable calculus, if $y=f(u)$ and $u=g(t)$, then $\frac{dy}{dt}=\frac{dy}{du}\frac{du}{dt}$. In multivariable calculus, the idea is the same, but there may be several input variables changing at once.

Suppose a function $z=f(x,y)$ depends on two variables, and those variables depend on time: $x=x(t)$ and $y=y(t)$. Then $z$ depends on $t$ indirectly. The chain rule says

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

This formula shows that the total rate of change comes from all the paths by which $t$ affects $z$. If $x$ changes quickly, it contributes through $\frac{\partial z}{\partial x}$, and if $y$ changes quickly, it contributes through $\frac{\partial z}{\partial y}$.

A good way to think about it is like a video game character moving on a map 🎮. The character’s altitude depends on location, and location depends on the player’s controls. If both the east-west and north-south positions change, the altitude changes because of both directions together.

Key terminology

A dependent variable is the quantity you want to study, such as $z$.
An intermediate variable is a variable in between, such as $x$ or $y$.
An independent variable is the original input, such as $t$.
A partial derivative like $\frac{\partial z}{\partial x}$ measures change in $z$ with respect to one variable while keeping others fixed.
A total derivative like $\frac{dz}{dt}$ measures the full change in $z$ as every linked variable changes.

One-Variable Input, Many-Variable Function

Let’s start with a common setup. Suppose $z=f(x,y)$ and both $x$ and $y$ depend on $t$. Then the chain rule is

$\frac{dz}{dt}=f_x(x(t),y(t))\frac{dx}{dt}+f_y(x(t),y(t))\frac{dy}{dt}.$

Here $f_x$ and $f_y$ are partial derivatives of $f$. The notation reminds you that the derivatives are evaluated at the current values of $x(t)$ and $y(t)$.

Example 1: Temperature on a surface

Suppose the temperature is modeled by $T(x,y)=100-2x^2-y^2$, and a bug moves so that $x(t)=t$ and $y(t)=3t$. Find $\frac{dT}{dt}$.

First compute the partial derivatives:

$\frac{\partial T}{\partial x}=-4x,\qquad \frac{\partial T}{\partial y}=-2y.$

Also,

$\frac{dx}{dt}=1,\qquad \frac{dy}{dt}=3.$

Now apply the chain rule:

$\frac{dT}{dt}=(-4x)(1)+(-2y)(3).$

Substitute $x=t$ and $y=3t$:

$\frac{dT}{dt}=-4t-6(3t)=-22t.$

At $t=1$, the temperature is decreasing at rate $-22$ units per time. This tells you how the temperature changes along the bug’s path 🐞.

Multiple Intermediate Variables

Sometimes the dependency chain has more steps. Suppose $z=f(x,y)$, but now $x=g(s,t)$ and $y=h(s,t)$. Then $z$ depends on $s$ and $t$ through both $x$ and $y$. The chain rule becomes

\frac{\partial z}{\partial s}=\frac{\partial z}{\partial x}\frac{\partial x}{\partial s}+\frac{\partial z}{\partial y}\frac{\partial y}{\partial s}

and

\frac{\partial z}{\partial t}=\frac{\partial z}{\partial x}\frac{\partial x}{\partial t}+\frac{\partial z}{\partial y}\frac{\partial y}{\partial t}.

This is the same idea as before: each input variable influences $z$ through every available path.

Example 2: Two parameters control a surface value

Let $z=x^2y+y$, with $x=s^2+t$ and $y=s-t^2$. Find $\frac{\partial z}{\partial s}$.

First find the partial derivatives of $z$ with respect to $x$ and $y$:

$\frac{\partial z}{\partial x}=2xy,$

$\frac{\partial z}{\partial y}=x^2+1.$

Next compute:

$\frac{\partial x}{\partial s}=2s,\qquad \frac{\partial y}{\partial s}=1.$

Now apply the chain rule:

$\frac{\partial z}{\partial s}=(2xy)(2s)+(x^2+1)(1).$

Finally, substitute $x=s^2+t$ and $y=s-t^2$:

$\frac{\partial z}{\partial s}=4s(s^2+t)(s-t^2)+(s^2+t)^2+1.$

This expression tells how $z$ changes when $s$ changes while $t$ stays fixed.

Chain Rule and the Gradient

The multivariable chain rule becomes even more powerful when using vectors. If $f(x,y)$ changes with time through $x(t)$ and $y(t)$, then

$\frac{dz}{dt}=\nabla f(x(t),y(t))\cdot \langle x'(t),y'(t)\rangle,$

where

$\nabla$ f=$\left$\langle \frac{\partial f}{\partial x},\frac{\partial f}{\partial y}$\right$\rangle.

The gradient points in the direction of steepest increase, and the velocity vector $\langle x'(t),y'(t)\rangle$ tells the direction of motion. Their dot product measures how much the motion moves uphill or downhill.

This connection is important because it links the chain rule to geometry. If you move along a curve on a surface, the rate of change of the surface value depends on how your direction lines up with the gradient.

Example 3: Direction of fastest change

Suppose $f(x,y)=x^2+y^2$. Then

$\nabla f=\langle 2x,2y\rangle.$

At the point $(1,2)$, the gradient is

$\langle 2,4\rangle.$

If a particle moves with velocity $\langle -1,3\rangle$, then the rate of change of $f$ along its path is

$\nabla$ f(1,2)$\cdot$ \langle -1,3\rangle=\langle 2,4\rangle$\cdot$ \langle -1,3\rangle=-2+12=10.

So even though the particle moves partly left, the overall path makes $f$ increase.

Connection to Tangent Planes and Linear Approximation

The chain rule is closely related to tangent planes. Near a point $(a,b)$, a differentiable function $f(x,y)$ can be approximated by its linearization:

$L(x,y)=f(a,b)+f_x(a,b)(x-a)+f_y(a,b)(y-b).$

This formula gives the equation of the tangent plane to the surface $z=f(x,y)$ at the point $(a,b,f(a,b))$.

Why does this matter for the chain rule? Because the linear approximation explains why the total change of a function can be estimated by combining partial changes. The differential is

$dz=f_x\,dx+f_y\,dy.$

If $x$ and $y$ both depend on $t$, then dividing by $dt$ gives

$\frac{dz}{dt}=f_x\frac{dx}{dt}+f_y\frac{dy}{dt}.$

So the chain rule is the derivative version of the tangent plane idea. Both are about local linear behavior near a point.

Real-world interpretation

Imagine a hill where elevation depends on your east-west and north-south position ⛰️. The tangent plane is like a flat sheet touching the hill at one point. If you take a tiny step east or north, the elevation change is predicted by the slopes in those directions. If your position itself depends on time, then the chain rule tells you how fast your elevation changes as you walk.

How to Solve Chain Rule Problems

A reliable strategy helps avoid mistakes:

Identify the outer function, like $z=f(x,y)$.
List the intermediate variables, like $x=x(t)$ and $y=y(t)$.
Compute all needed partial derivatives.
Compute the derivatives of the intermediate variables.
Substitute into the chain rule and simplify.

For more complicated situations, draw a dependency diagram. For example, if $w=f(x,y,z)$ and $x,g,h$ all depend on $u$ and $v$, you can organize the paths from the final variable back to the original inputs. This makes it easier to count every route correctly.

A common mistake is forgetting one path. If $z$ depends on both $x$ and $y$, then both contributions must appear. Another mistake is mixing up partial derivatives and ordinary derivatives. Use $\frac{\partial}{\partial x}$ when treating other variables as fixed, and use $\frac{d}{dt}$ when the variable is changing along a path.

Conclusion

The chain rule in multivariable settings is a tool for following change through several connected variables. It tells you how to compute rates like $\frac{dz}{dt}$ when $z$ depends on variables such as $x$ and $y$, and those variables depend on time or on other parameters. students, this topic connects directly to tangent planes, because both ideas come from local linear approximation. The same partial derivatives that describe a tangent plane also describe how changes combine in the chain rule. With practice, you can use these ideas to analyze motion, temperature, height, pressure, and many other real-world quantities.

Study Notes

The multivariable chain rule tracks how a dependent quantity changes through intermediate variables.
For $z=f(x,y)$ with $x=x(t)$ and $y=y(t)$, the rule is $$\frac{dz}{dt}=f_x\frac{dx}{dt}+f_y\frac{dy}{dt}.$$
For $z=f(x,y)$ with $x=x(s,t)$ and $y=y(s,t)$, use separate formulas for $\frac{\partial z}{\partial s}$ and $\frac{\partial z}{\partial t}$.
The gradient is $\nabla f=\left\langle f_x,f_y\right\rangle$, and the chain rule can be written using a dot product.
The chain rule is connected to tangent planes because both rely on linear approximation near a point.
The differential $dz=f_x\,dx+f_y\,dy$ helps explain why the chain rule works.
Always include every path of dependence so no contribution is missed.
Real-world uses include motion on a surface, temperature change, and optimization problems.