Linear Approximation in Multivariable Calculus

Have you ever wanted to estimate a hard-to-calculate value without doing a full calculation? 📏 In multivariable calculus, linear approximation gives us a fast and reliable way to estimate the value of a function near a point where we already know the answer. This idea is closely connected to tangent planes, total differential, and later to the chain rule because all of them describe how a function changes when its inputs change a little.

What You Will Learn

By the end of this lesson, students, you will be able to:

explain what linear approximation means in two or more variables,
use the tangent plane to estimate values of a function near a point,
connect linear approximation to the total differential,
see why this idea matters in tangent planes and the chain rule,
use examples to estimate real quantities quickly and accurately.

The key idea is simple: if a surface is smooth and you zoom in close enough, it looks almost flat. That “almost flat” surface is the tangent plane, and the formula for that plane gives the linear approximation.

The Big Idea Behind Linear Approximation

In single-variable calculus, you may remember that a function $f(x)$ can be approximated near $x=a$ by its tangent line:

$$f(x) \approx f(a) + f'(a)(x-a).$$

This works because the graph of $f$ and its tangent line match closely near $a$.

In multivariable calculus, the same idea becomes a tangent plane approximation. For a function of two variables, $z=f(x,y)$, the linear approximation near a point $(a,b)$ is

$$f(x,y) \approx f(a,b) + f_x(a,b)(x-a) + f_y(a,b)(y-b).$$

This formula says that the change in $f$ is approximately the sum of the changes coming from each input variable. The partial derivatives $f_x(a,b)$ and $f_y(a,b)$ measure how steep the surface is in the $x$-direction and $y$-direction at the point $(a,b)$.

If $f$ is differentiable at $(a,b)$, then this approximation becomes especially accurate when $(x,y)$ is close to (a,b)`. 🌟

Tangent Planes and Why They Matter

A tangent plane is the best flat surface that matches a curved surface at a point. If the graph of $z=f(x,y)$ is like a hill, the tangent plane is the flat sheet that just touches the hill at one point and has the same local slope in the $x$ and $y$ directions.

The tangent plane to $z=f(x,y)$ at $(a,b,f(a,b))$ is

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

This is not just a geometric object; it is also the formula for linear approximation. That means tangent planes are the geometric version of the same idea.

Example 1: Approximating a Square Root

Let $f(x,y)=\sqrt{x+y}$. Find a linear approximation near $(3,1)$.

First, compute the function value:

$$f(3,1)=\sqrt{4}=2.$$

Now find the partial derivatives:

$$f_x(x,y)=\frac{1}{2\sqrt{x+y}}, \qquad f_y(x,y)=\frac{1}{2\sqrt{x+y}}.$$

At $(3,1)$, both partial derivatives are

$$f_x(3,1)=\frac{1}{4}, \qquad f_y(3,1)=\frac{1}{4}.$$

So the linear approximation is

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

If we want to estimate $\sqrt{4.1}$, we can use a nearby point such as $(3.1,1)$ or $(3,1.1)$, because $x+y$ is near $4$. For example,

$$f(3.1,1)=\sqrt{4.1} \approx 2+\frac{1}{4}(0.1)=2.025.$$

This is close to the actual value and much faster than a calculator-based exact evaluation if the goal is estimation. ✅

Total Differential: A Compact Way to Describe Change

The total differential is closely related to linear approximation. For a function $z=f(x,y)$, the differential is

$$dz=f_x(x,y)\,dx+f_y(x,y)\,dy.$$

Here, $dx$ and $dy$ are small changes in the inputs, and $dz$ is the predicted change in the output. If we start at $(a,b)$, then the approximation becomes

$$\Delta z \approx dz = f_x(a,b)\Delta x + f_y(a,b)\Delta y.$$

This formula is another way to express linear approximation. In fact,

$$f(a+\Delta x,b+\Delta y) \approx f(a,b) + f_x(a,b)\Delta x + f_y(a,b)\Delta y.$$

So the total differential and linear approximation are two sides of the same coin. The linear approximation gives an estimated value of the function, while the differential estimates the change in the function.

Real-World Interpretation

Imagine a factory machine where output depends on temperature and pressure. If the engineer knows the current output and how sensitive output is to each variable, then small changes in temperature and pressure can be estimated using the differential. This is useful because it avoids repeated full calculations and helps predict the effect of small adjustments. 🏭

How to Build a Linear Approximation Step by Step

To find a linear approximation of $f(x,y)$ near $(a,b)$, follow these steps:

Find the point value $f(a,b)$.
Compute the partial derivatives $f_x(x,y)$ and $f_y(x,y)$.
Evaluate those partial derivatives at $(a,b)$.
Substitute into

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

Example 2: Approximating an Exponential Function

Let $f(x,y)=e^{x+2y}$ and approximate near $(0,0)$.

First,

$$f(0,0)=e^0=1.$$

Next, compute partial derivatives:

$$f_x(x,y)=e^{x+2y}, \qquad f_y(x,y)=2e^{x+2y}.$$

At $(0,0)$,

$$f_x(0,0)=1, \qquad f_y(0,0)=2.$$

So the linear approximation is

$$L(x,y)=1+x+2y.$$

Now estimate $e^{0.03+0.02}$:

$$e^{0.03+0.02}=f(0.03,0.01) \approx 1+0.03+2(0.01)=1.05.$$

This approximation works because $(0.03,0.01)$ is very close to $(0,0)$. The closer the point is to the center of approximation, the better the estimate usually is.

Why Linear Approximation Fits the Bigger Picture

Linear approximation is not an isolated topic. It sits right at the center of several important ideas in multivariable calculus.

Tangent planes give the geometry of the approximation.
Partial derivatives give the slopes in coordinate directions.
The total differential describes predicted change.
The chain rule later uses the same linear thinking when variables depend on other variables.

This connection matters because many real problems involve quantities that change indirectly. For example, if pressure depends on time and temperature depends on time, then the chain rule helps find how pressure changes over time. The reason the chain rule works so well is that small changes can be tracked using linear approximations.

In other words, linear approximation is one of the building blocks for understanding how complicated systems change. It turns a curved problem into a simpler flat one, at least locally. 📐

Common Mistakes to Avoid

A few mistakes show up often when students learn this topic:

confusing the approximation point $(a,b)$ with the new point $(x,y)$,
forgetting to evaluate partial derivatives at the base point,
using linear approximation far away from the point of tangency,
mixing up the value estimate $L(x,y)$ with the change estimate $dz$.

Remember: linear approximation is accurate near the point where it is built. If you move too far away, the surface may curve too much for the tangent plane to stay accurate.

Conclusion

Linear approximation is a powerful way to estimate values of multivariable functions near a known point. It uses the tangent plane formula

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

and it is closely connected to the total differential

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

students, this lesson shows how a curved surface can be treated like a flat plane when you are zoomed in close enough. That idea is essential for tangent planes, helpful for estimation, and foundational for the chain rule in multivariable calculus. Once you understand linear approximation, you have a strong tool for analyzing change in many variables.

Study Notes

Linear approximation estimates a multivariable function near a point using a linear function.
For $z=f(x,y)$ near $(a,b)$, the approximation is

$$f(x,y) \approx f(a,b) + f_x(a,b)(x-a) + f_y(a,b)(y-b).$$

The tangent plane to the graph of $z=f(x,y)$ has the same formula as the linear approximation.
The partial derivatives $f_x$ and $f_y$ measure how the function changes in the $x$ and $y$ directions.
The total differential is

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

and it estimates the change in the output.

Linear approximation works best when the new point is close to the base point.
This topic connects directly to tangent planes, differentials, and the chain rule.
The main purpose is to replace a hard curved problem with an easier flat approximation near one point.