Surface Integrals: Scalar Fields — Compute surface integrals of scalar fields - Multivariable Calculus

SECTION 1

Historical Context & Motivation

Long before anyone wrote down a formula for a surface integral, scientists needed ways to measure quantities spread across curved surfaces. Imagine trying to figure out how much heat escapes through a curved metal shield, or how much rain collects on a dome-shaped roof. These are exactly the kinds of problems that drove mathematicians to extend ordinary integration from flat regions to surfaces floating in three-dimensional space.

The idea grew naturally out of single-variable and double integrals. Once mathematicians like Leonhard Euler and Joseph-Louis Lagrange mastered integration over flat two-dimensional regions, the next challenge was to handle curved surfaces — like spheres, cylinders, and more exotic shapes. The formal tools were refined during the eighteenth and nineteenth centuries, fueled by problems in physics such as gravitational attraction, fluid flow, and electromagnetism.

1760s • Euler & Surface Geometry

Leonhard Euler develops the theory of surfaces in space, introducing parameterizations that describe curved surfaces using two variables — a crucial prerequisite for surface integrals.

1788 • Lagrange & Area Elements

Joseph-Louis Lagrange formalizes the idea of an infinitesimal area element on a surface, paving the way for integrating functions over non-flat regions.

1820s • Gauss & Flux Integrals

Carl Friedrich Gauss uses surface integrals in his work on gravitational and electrostatic fields, publishing his famous divergence theorem that links surface integrals to volume integrals.

1850s • Riemann & Formalization

Bernhard Riemann rigorously defines integration on curved manifolds, placing surface integrals on a firm theoretical foundation used in modern mathematics.

The central question these mathematicians wrestled with was: If you know the value of some quantity at every point of a curved surface, how do you add up all of those values to get one meaningful total? That question is exactly what the surface integral of a scalar field answers, and it remains one of the most important tools in physics, engineering, and applied mathematics today.

SECTION 2

Core Principles & Definitions

Before diving into calculations, you need a clear picture of the key ideas behind surface integrals of scalar fields. A scalar field is simply a function f(x, y, z) that assigns a single number — like temperature, density, or pressure — to every point in space. A surface is a two-dimensional shape sitting in three-dimensional space, described by a parametric representation that maps two parameters (usually called u and v) to points (x, y, z) on the surface.

Scalar Field

A function f(x, y, z) that returns a single number at each point. Think of it as a 'value map' — every location has a temperature, elevation, or charge density.

Parametric Surface

A surface described by r(u, v) = ⟨x(u,v), y(u,v), z(u,v)⟩. Two parameters sweep out every point on the surface, much like latitude and longitude cover the Earth.

Surface Element dS

A tiny patch of area on the surface. It accounts for how the surface stretches and curves, calculated using the cross product of partial derivatives: ‖rᵤ × rᵥ‖ du dv.

The Surface Integral

The double integral ∬_S f dS adds up the values of f over every tiny patch of the surface. It generalizes the single-variable integral ∫ f ds along a curve to two dimensions.

✦

KEY TAKEAWAY

Think of a surface integral like painting a curved wall. The amount of paint you need depends on two things: how thick you apply the paint at each spot (the scalar field f) and the actual area of each small patch of wall (the surface element dS). A flat wall and a bumpy wall of the same width and height have different actual areas, so the surface element handles that stretching for you.

SECTION 3

Visual Explanation

The diagram below shows a curved surface S sitting in three-dimensional space. The surface is divided into small patches, each with area ΔS. At the center of each patch, the scalar field f has some value. The surface integral sums f × ΔS over every tiny patch and takes the limit as the patches become infinitely small.

The surface S is broken into small patches (colored rectangles). Each patch has area ΔS and the scalar field f is evaluated at a sample point in each patch. The surface integral is the limit of the sum Σ f · ΔS as patches shrink to zero.

Notice how the grid lines on the surface are not evenly spaced — this reflects the fact that the surface curves and stretches in different directions. The surface element dS automatically accounts for this stretching. On a steeply tilted part of the surface, the patches have more actual area than they appear to in a flat parameter grid, and the cross-product magnitude captures this distortion exactly.

SECTION 4

Mathematical Framework

To actually compute a surface integral, you need three ingredients: a parametric description of the surface, the scalar field f(x, y, z), and the surface element dS that measures how much area each tiny patch really has. Let's formalize each piece.

PARAMETRIC SURFACE

r(u, v) = ⟨x(u,v), y(u,v), z(u,v)⟩, (u,v) ∈ D

r is the position vector tracing out the surface as u and v range over the parameter domain D in the uv-plane.

PARTIAL DERIVATIVES

rᵤ = ∂r/∂u = ⟨∂x/∂u, ∂y/∂u, ∂z/∂u⟩ rᵥ = ∂r/∂v = ⟨∂x/∂v, ∂y/∂v, ∂z/∂v⟩

These two vectors are tangent to the surface at each point. Their cross product gives a vector perpendicular to the surface whose magnitude equals the area of the infinitesimal parallelogram.

SURFACE ELEMENT

dS = ‖rᵤ × rᵥ‖ du dv

The magnitude of the cross product ‖rᵤ × rᵥ‖ tells you how much the surface stretches at each point. Multiply by du dv (the tiny rectangle in parameter space) to get the actual area dS on the surface.

SURFACE INTEGRAL OF A SCALAR FIELD

∬_S f dS = ∬_D f(r(u,v)) ‖rᵤ × rᵥ‖ du dv

Replace f(x,y,z) with f(r(u,v)) by substituting the parametric equations, then integrate over the parameter domain D. This converts a surface integral into an ordinary double integral you already know how to evaluate.

💡

Special Case: z = g(x, y)

When the surface is given as z = g(x, y), you can use x and y directly as parameters. The surface element simplifies to dS = √(1 + (∂g/∂x)² + (∂g/∂y)²) dx dy. This is the form you will use most often in introductory problems.

SECTION 5

Deriving the Surface Element — Step by Step

The heart of any surface integral computation is finding ‖rᵤ × rᵥ‖. Let's look at this process visually and then walk through the algebra for the important special case where the surface is given explicitly as z = g(x, y).

Left: a tiny rectangle du × dv in the parameter domain. Right: the parametrization r(u,v) maps it to a curved patch on the surface, spanned by the tangent vectors rᵤ du and rᵥ dv. Below: the explicit computation for z = g(x, y).

Here's the key insight: in the parameter domain, every tiny patch is a flat rectangle with area du × dv. But on the actual surface, that patch gets stretched and tilted. The cross product rᵤ × rᵥ produces a vector whose magnitude equals the area of the parallelogram spanned by the two tangent vectors. That magnitude, ‖rᵤ × rᵥ‖, is the area magnification factor that converts flat parameter area into true surface area.

ℹ

Quick Check

If the surface is flat and horizontal (z = constant), then gₓ = 0 and gᵧ = 0, so dS = √(1 + 0 + 0) dx dy = dx dy. The surface integral reduces to an ordinary double integral — exactly what you'd expect!

SECTION 6

Worked Example

Let's compute the surface integral of the scalar field f(x, y, z) = x² + y² over the paraboloid z = x² + y² for 0 ≤ x² + y² ≤ 1 (the part of the paraboloid below z = 1). This is a classic introductory problem that uses all the tools we've developed.

Surface Integral of f = x² + y² over z = x² + y²

1
Step 1 — Identify the surface and scalar field

Surface: z = g(x, y) = x² + y², with 0 ≤ x² + y² ≤ 1. Scalar field: f(x, y, z) = x² + y². Since z = x² + y² on the surface, we can also write f = z on S, but we'll keep it as x² + y² for the parametric approach.
2
Step 2 — Compute the partial derivatives of g

gₓ = ∂g/∂x = 2x and gᵧ = ∂g/∂y = 2y. These measure how steeply the surface tilts in the x- and y-directions.

gₓ = 2x, gᵧ = 2y
3
Step 3 — Build the surface element dS

Using the formula dS = √(1 + gₓ² + gᵧ²) dx dy, we substitute: dS = √(1 + (2x)² + (2y)²) dx dy = √(1 + 4x² + 4y²) dx dy.

dS = √(1 + 4x² + 4y²) dx dy
4
Step 4 — Set up the double integral

∬_S f dS = ∬_D (x² + y²) √(1 + 4x² + 4y²) dx dy, where D is the disk x² + y² ≤ 1 in the xy-plane. Since both f and dS depend on x² + y², polar coordinates will simplify this greatly.
5
Step 5 — Convert to polar coordinates

Let x = r cos θ, y = r sin θ. Then x² + y² = r², and dx dy = r dr dθ. The domain D becomes 0 ≤ r ≤ 1, 0 ≤ θ ≤ 2π. The integral becomes: ∫₀²π ∫₀¹ r² · √(1 + 4r²) · r dr dθ = ∫₀²π dθ · ∫₀¹ r³√(1 + 4r²) dr = 2π ∫₀¹ r³√(1 + 4r²) dr.

2π ∫₀¹ r³√(1 + 4r²) dr
6
Step 6 — Evaluate the radial integral by substitution

Let u = 1 + 4r², so du = 8r dr, which gives r dr = du/8. Also r² = (u − 1)/4. When r = 0, u = 1; when r = 1, u = 5. Then r³√(1 + 4r²) dr = r² · √u · r dr = ((u−1)/4) · √u · (du/8) = (1/32)(u − 1)√u du = (1/32)( $u^{3/2}$ − $u^{1/2}$ ) du. Integrating: (1/32)[ $2u^{5/2}$ /5 − $2u^{3/2}$ /3] from 1 to 5.
7
Step 7 — Compute the final answer

At u = 5: 2( $5^{5/2}$ )/5 − 2( $5^{3/2}$ )/3 = 2(25√5)/5 − 2(5√5)/3 = 10√5 − (10√5)/3 = (20√5)/3. At u = 1: 2/5 − 2/3 = (6 − 10)/15 = −4/15. So the bracket is (20√5)/3 − (−4/15) = (20√5)/3 + 4/15 = (100√5 + 4)/15. Multiply by 2π/32 = π/16: ∬_S f dS = (π/16) × (100√5 + 4)/15 = π(100√5 + 4)/240 = π(25√5 + 1)/60.

∬_S f dS = π(25√5 + 1)/60 ≈ 2.996

SECTION 7

Strengths, Limitations & Comparisons

Surface integrals of scalar fields are one member of a family of integration techniques. Understanding how they compare to other types of integrals helps you choose the right tool for a given problem.

Integral Type	Domain	What It Computes
Line integral (scalar)	A curve C in 2D or 3D	Total of a scalar field along a path — e.g., total mass of a wire
Double integral	A flat region D in the xy-plane	Sum of f(x,y) over a flat region — e.g., area, mass of a flat plate
Surface integral (scalar)	A surface S in 3D	Sum of f(x,y,z) over a curved surface — e.g., total heat, mass of a shell
Surface integral (vector/flux)	A surface S in 3D	Net flow of a vector field through the surface — used in Gauss's law

Comparison of integral types

✦

KEY TAKEAWAY

A surface integral of a scalar field is the natural 'next step' beyond a double integral: instead of integrating over a flat region, you integrate over a curved surface. The extra factor ‖rᵤ × rᵥ‖ is the price you pay for curvature — it stretches your area element so that every patch on the surface is measured correctly, just as √(1 + (dy/dx)²) stretches ds for arc length in single-variable calculus.

One important limitation to keep in mind: computing ‖rᵤ × rᵥ‖ can get algebraically messy. For surfaces that are not easily described by z = g(x,y), you may need a full parametric representation, which makes the cross-product computation more involved. In practice, many problems are designed so that the integrand simplifies nicely — so always look for symmetry and smart coordinate choices (like polar or spherical) before diving into brute-force algebra.

SECTION 8

Connection to Advanced Theory

Surface integrals of scalar fields are the foundation for more advanced topics you'll encounter in higher-level mathematics and physics. Two of the most important extensions are surface integrals of vector fields (flux integrals) and the major theorems of vector calculus — Stokes' theorem and the divergence theorem.

This Lesson	Advanced Extension
Scalar field f(x,y,z) on a surface	Vector field F(x,y,z) on a surface → flux integral ∬ F · dS
Surface element dS = ‖rᵤ × rᵥ‖ du dv	Oriented surface element dS = (rᵤ × rᵥ) du dv (a vector, not just its magnitude)
Compute total of a quantity on a surface	Divergence theorem links ∬ F · dS to ∭ ∇·F dV — surface flux equals volume divergence
Parametrize a surface with two variables	Differential forms generalize surfaces and integrals to any dimension (manifolds)

From scalar surface integrals to advanced vector calculus

The scalar surface integral you learned in this lesson is not just an end in itself — it's a stepping stone. Once you're comfortable evaluating ∬_S f dS, you're well prepared to tackle flux integrals, and from there, the powerful divergence and Stokes' theorems that unify multivariable calculus. These theorems are the backbone of electromagnetism, fluid dynamics, and modern physics.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

Explain in your own words why the factor ‖rᵤ × rᵥ‖ is needed in a surface integral. What would go wrong if you simply integrated f(x,y,z) over the parameter domain D without it?

Show answer

The factor ‖rᵤ × rᵥ‖ accounts for how much the surface stretches compared to the flat parameter domain. Without it, you'd be measuring area in the uv-plane, not on the actual surface. For a steeply tilted or highly curved surface, the true surface area can be much larger than the area of D. Omitting ‖rᵤ × rᵥ‖ would undercount the contribution of regions where the surface is stretched, giving you the wrong total.

Question 2: BASIC CALCULATION

Compute ∬_S 1 dS for the surface z = 3x + 4y over the region D: 0 ≤ x ≤ 1, 0 ≤ y ≤ 2. (This gives the surface area of S.)

Show answer

Since z = g(x,y) = 3x + 4y, we have gₓ = 3 and gᵧ = 4. The surface element is dS = √(1 + 9 + 16) dx dy = √26 dx dy. Since f = 1, the integral becomes ∬_D √26 dx dy = √26 × (area of D) = √26 × (1)(2) = 2√26 ≈ 10.198. The surface area of this tilted plane over the given rectangle is 2√26 square units.

Question 3: INTERMEDIATE

Evaluate ∬_S (x² + y²) dS where S is the hemisphere z = √(1 − x² − y²) (the upper half of the unit sphere).

Show answer

Let g(x,y) = √(1 − x² − y²). Then gₓ = −x/√(1−x²−y²) and gᵧ = −y/√(1−x²−y²). So 1 + gₓ² + gᵧ² = 1 + (x²+y²)/(1−x²−y²) = 1/(1−x²−y²). Thus dS = 1/√(1−x²−y²) dx dy. The integral becomes ∬_D (x²+y²)/√(1−x²−y²) dx dy over the disk x²+y² ≤ 1. In polar: ∫₀²π ∫₀¹ r²/√(1−r²) · r dr dθ = 2π ∫₀¹ r³/√(1−r²) dr. Using the substitution u = 1−r², this evaluates to 2π · 4/3 · (1/something)... Carrying through: let u = 1−r², du = −2r dr, r² = 1−u. The integral becomes 2π ∫₁⁰ (1−u)/√u · (−du/2) = π ∫₀¹ (

u^{−1/2}

−

u^{1/2}

) du = π[2√u − (2/3)

u^{3/2}

]₀¹ = π(2 − 2/3) = 4π/3.

Question 4: APPLIED

A thin metal dome has the shape z = 4 − x² − y² for z ≥ 0. The surface mass density at each point is σ(x,y,z) = z kg/m². Find the total mass of the dome.

Show answer

The total mass is M = ∬_S z dS. The dome z = 4 − x² − y² meets z = 0 when x²+y² = 4, so D is the disk r ≤ 2. Here gₓ = −2x, gᵧ = −2y, so dS = √(1+4x²+4y²) dx dy = √(1+4r²) · r dr dθ. On the surface, z = 4 − r². Thus M = ∫₀²π ∫₀² (4−r²)√(1+4r²) · r dr dθ = 2π ∫₀² (4−r²) r√(1+4r²) dr. Using the substitution u = 1+4r² (du = 8r dr, r² = (u−1)/4), when r = 0, u = 1 and when r = 2, u = 17. Then 4 − r² = 4 − (u−1)/4 = (17−u)/4. The integral becomes 2π ∫₁¹⁷ ((17−u)/4) · √u · (du/8) = (π/16) ∫₁¹⁷ (

17u^{1/2}

−

u^{3/2}

) du = (π/16)[17·(2/3)

u^{3/2}

− (2/5)

u^{5/2}

]₁¹⁷ = (π/16)[(34/3)(

17^{3/2}

) − (2/5)(

17^{5/2}

) − 34/3 + 2/5]. Computing:

17^{3/2}

= 17√17,

17^{5/2}

= 289√17. So = (π/16)[(34·17√17)/3 − (578√17)/5 − 34/3 + 2/5] = (π/16)[578√17/3 − 578√17/5 − 164/15] = (π/16)[578√17 · 2/15 − 164/15] = (π/240)[1156√17 − 164] ≈ (π/240)(4764.6 − 164) ≈ 60.2 kg.

Question 5: CRITICAL THINKING

Suppose S₁ is the cone z = √(x²+y²) and S₂ is the plane z = (x+y)/√2 + C, both restricted to the disk x²+y² ≤ 1. Both surfaces have constant ‖rᵤ × rᵥ‖ over D. Without computing the integrals, explain how you can determine which surface gives a larger value of ∬_S 1 dS (surface area). What does this tell you about the relationship between tilt and surface area?

Show answer

For the cone z = √(x²+y²), gₓ = x/√(x²+y²), gᵧ = y/√(x²+y²), so gₓ²+gᵧ² = 1, giving dS = √2 dx dy. For the tilted plane z = (x+y)/√2 + C, gₓ = 1/√2, gᵧ = 1/√2, so gₓ²+gᵧ² = 1, giving dS = √2 dx dy as well. Both surface elements are identical and constant! The surface areas are equal: ∬ 1 dS = √2 × (area of D) = π√2 for both. This illustrates a deep point: what matters for the area magnification factor is not the shape of the surface but the magnitude of its tilt (gradient). Two very different surfaces — one curved, one flat — can have the same surface area over a given domain if the total tilt magnitude is the same everywhere. The moral is that surface area depends on how steeply the surface departs from horizontal, not on whether it curves.

SUMMARY

Lesson Summary

Summary

A surface integral of a scalar field extends double integration from flat regions to curved surfaces in three-dimensional space. Given a parametric surface r(u, v), the integral is computed as ∬_D f(r(u,v)) ‖rᵤ × rᵥ‖ du dv, where the surface element dS = ‖rᵤ × rᵥ‖ du dv measures the true area of each infinitesimal patch on the surface. For surfaces given as z = g(x,y), this simplifies to dS = √(1 + gₓ² + gᵧ²) dx dy.

The computation strategy follows three steps: (1) parametrize the surface, (2) compute the cross product of the partial derivatives and take its magnitude, and (3) set up and evaluate the resulting double integral over the parameter domain. Always look for symmetry — polar coordinates are your best friend when the surface and integrand both depend on x² + y². This topic is the gateway to flux integrals, the divergence theorem, and Stokes' theorem — the cornerstones of vector calculus.