Find The Average Height Of A Hemisphere Above The Disk

You're staring at a hemisphere sitting on a table. The flat circular base — the disk — touches the surface everywhere at once. Here's the thing — the curved dome rises up, smooth and symmetric. Simple geometry, right?

But here's the question that trips up calculus students every semester: what's the average height of that dome above the disk?*

Not the maximum height (that's just the radius). Consider this: the average* — if you could measure the height at every single point on the base and take the mean. That said, not the height at the center. It's a classic multivariable calculus problem, and the answer surprises people who haven't worked through it.

Let's walk through it properly. Also, no hand-waving. On the flip side, no "it can be shown that. " Just the reasoning, step by step, the way I'd explain it over coffee.

What Is Average Height Anyway?

Before we touch an integral, let's be precise about what we're asking.

We have a solid hemisphere of radius R. Its flat face lies in the xy-plane, centered at the origin. So naturally, the curved surface is the top half of the sphere x² + y² + z² = R²*, with z ≥ 0*. The base — the "disk" — is the region D defined by x² + y² ≤ R²* in the plane z = 0*.

At any point (x, y) in that disk, the height of the hemisphere above it is simply the z-coordinate on the surface:

z = √(R² − x² − y²)

The average value of a function over a region is the integral of that function over the region, divided by the area of the region. So:

Average height = (1 / Area(D)) ∬_D z dA

Area of the disk is πR². That part's easy. The work lives in the double integral.

Why This Matters (Beyond the Homework Problem)

You might wonder: who cares about the average height of a hemisphere?

Fair question. This isn't just a textbook exercise. The same principle — averaging a height function over a base region — shows up in:

• Physics: finding the center of mass of a hemispherical shell or solid
• Engineering: calculating average pressure on a curved surface, or mean clearance in a bearing
• Statistics: expected value of a random variable distributed over a disk
• Computer graphics: sampling techniques for hemisphere lighting (cosine-weighted hemisphere sampling uses exactly this geometry)

And conceptually? It's a gateway. Once you're comfortable setting up and evaluating this integral, you can handle average values over any region — ellipses, triangles, weird bounded domains. The hemisphere is just the cleanest, most symmetric starting point.

How to Actually Compute It

Setting up in Cartesian coordinates (don't do this)

You could* write the integral as:

∬_D √(R² − x² − y²) dx dy

with limits x from −R to R, y from −√(R² − x²) to √(R² − x²).

Technically correct. The inner integral involves √(R² − x² − y²) with respect to y, which begs for a trig substitution. Practically painful. Then the outer integral gets messy fast.

There's a better way.

Polar coordinates — the natural choice

The disk D is a circle. Worth adding: the function √(R² − x² − y²) depends only on distance from the origin. This screams polar coordinates.

Let:

• x = r cos θ*
• y = r sin θ*
• dA = r dr dθ*
• x² + y² = r²*

The height becomes z = √(R² − r²).

The region D in polar terms: r goes from 0 to R, θ goes from 0 to 2π.

The integral transforms cleanly:

∬_D √(R² − r²) * r dr dθ = ∫₀²π ∫₀ᴿ √(R² − r²) * r dr dθ

Notice the extra r from the Jacobian. That's not optional — it's what makes the substitution work.

Evaluating the inner integral

Focus on: ∫₀ᴿ r √(R² − r²) dr

This is a classic u-substitution. Let:

u = R² − r²*
du = −2r dr* → r dr = −du/2*

When r = 0*, u = R²*.
When r = R*, u = 0*.

So the integral becomes:

∫_{u=R²}^{0} √u * (−du/2) = (1/2) ∫₀^{R²} u^{1/2} du = (1/2) * [ (2/3) u^{3/2} ]₀^{R²} = (1/3) (R²)^{3/2} = (1/3) R³

That's it. The inner integral evaluates to R³/3.

The outer integral

Now plug that back in:

∫₀²π (R³/3) dθ = (R³/3) * [θ]₀²π = (2π/3) R³

If you found this helpful, you might also enjoy how to calculate the percentage by mass or what is the value of x apex 2.2 3.

So the double integral ∬_D z dA = (2π/3) R³.

Divide by the area

Area of the disk = πR².

Average height = (2π/3 R³) / (πR²) = 2R/3.

That's the answer. The average height of a hemisphere above its base disk is two-thirds of the radius.

Wait — Does That Make Sense?

Let's sanity-check.

The maximum height is R (at the center). If height were distributed uniformly, the average would be R/2. The minimum is 0 (at the rim). But it's not uniform — there's more area* near the rim where the height is small, and less area* near the center where the height is large.

Wait. That would pull the average down* below R/2. But we got 2R/3, which is above* R

Resolving the Apparent Paradox

At first glance the result ( \displaystyle \frac{2R}{3} ) may feel counter‑intuitive. After all, the height varies from (R) at the centre to (0) at the rim, and a naïve “mid‑point” guess would be (R/2). The key is that the area element itself depends on the radius. In polar coordinates the infinitesimal patch of surface is (r,dr,d\theta); patches near the edge carry a larger weight simply because they occupy more physical space.

When we integrate ( \sqrt{R^{2}-r^{2}} ) multiplied by this (r) factor, the contribution of the outer annuli is amplified even though their heights are smaller. The net effect is that the weighted average ends up above the simple arithmetic mean (R/2). Basically, the hemisphere “spreads” its height over a disproportionately large circumference, and that extra circumference compensates for the lower values of (z).

A quick sanity check with a unit hemisphere ((R=1)) illustrates the point:

• The exact average height is (2/3 \approx 0.Plus, 667). - If we sampled the height uniformly over the disc (ignoring the (r) weighting), the average would indeed be (0.5).
  That's why - By sampling according to the true area distribution, we obtain (0. 667), which matches the integral result.

So the computation is self‑consistent; the apparent contradiction disappears once the geometry of the domain is taken into account.

Why This Computation Matters

The hemisphere integral is more than a neat calculus exercise; it is a prototype for a whole class of average‑value problems that appear across mathematics, physics, and computer graphics.

• Geometric probability – When we pick a random point uniformly on a sphere and ask for the expected height of the enclosing hemisphere above the equatorial plane, the same (2R/3) emerges.
• Rendering and shading – In ray‑tracing, the colour contributed by a diffuse surface illuminated by an isotropic light source can be expressed as an integral over the hemisphere of cos θ dΩ. Evaluating that integral reduces to the same algebraic steps we just performed, yielding the familiar factor ( \pi R^{2} ) or ( 2\pi R^{2}/3 ) depending on the normalization.
• Mechanics – The first moment of a thin hemispherical shell about its flat face is exactly the integral we computed, and it shows up in the analysis of rotating bodies and pressure distributions.

Because the hemisphere is the simplest symmetric region whose boundary can be described by a single quadratic equation, mastering its average value equips you with a template that can be adapted to ellipsoids, spherical caps, or even irregularly bounded domains by changing the limits

of integration and the integrand accordingly. Here's a good example: if the domain is a spherical cap of height (h), the radial limit becomes a function of (h), and the resulting average height shifts predictably as (h \to 0) (approaching a flat disc) or (h \to R) (approaching the full hemisphere).

A Broader Perspective

What makes this particular integral so instructive is that it exposes a recurring theme in applied mathematics: the domain of integration is not just a passive backdrop but an active participant in shaping the result. Many students initially compute averages by naively integrating the function alone, forgetting that the "measure" — the (r,dr,d\theta) in polar coordinates, the (dV) in triple integrals, or the probability density in statistics — encodes the geometry of the problem. Ignoring that measure is like trying to compute the centre of mass of an irregular object while pretending it has uniform density.

The hemisphere example also serves as a gentle gateway to more sophisticated ideas. The same integral, viewed through the lens of differential geometry, becomes an integral over a two‑dimensional Riemannian manifold with the round metric (ds^{2} = R^{2}(d\theta^{2} + \sin^{2}\theta,d\phi^{2})). The Jacobian (\sin\theta) that appears in spherical coordinates plays the exact same role as the factor (r) in our polar computation — it weights each patch of the surface according to its true size. Recognising this structural parallel prepares the reader for topics such as integration on manifolds, Stokes' theorem in higher dimensions, and the divergence theorem, all of which generalise the simple idea we used here.

Concluding Remarks

We began with a question that seems almost trivial — what is the average height of a hemisphere?* — and uncovered a rich interplay between geometry, integration, and physical intuition. Now, the answer (2R/3) is elegant not only in its simplicity but in the way it emerges from a careful accounting of how area grows with radius. Along the way, we saw how a common pitfall (treating the domain as a flat disc) leads to an incorrect result, and how the correct treatment reveals the deep connection between the shape of a region and the quantities we compute over it.

This single example illustrates principles that extend far beyond the hemisphere: the importance of choosing the right coordinate system, the role of the Jacobian in preserving geometric truth, and the power of dimensional analysis as a sanity check. Whether you are computing the centre of mass of a dome, the flux of a vector field through a spherical surface, or the expected value of a random variable distributed over a curved space, the lessons drawn here remain the same — respect the geometry, set up the integral faithfully, and let the mathematics do the rest.