Why a Sphere’s Surface Area Is Four Times Its Shadow

The area of a sphere is 4πR², four times the area of a circle of the same radius. Grant Sanderson asks for a geometric explanation of that relationship, not merely a formula.

A flat sheet cannot wrap a sphere without distortion because their curvatures differ. Two arguments nevertheless connect the sphere to circles, followed by a general fact about convex shapes.

The first approach compares the sphere with the side of its enclosing cylinder. Unwrapped, the cylinder’s side is a rectangle of width 2πR and height 2R, with area 4πR².

The key is to understand why the areas match. Approximate the sphere with tiny surface patches and project each horizontally outward to the cylinder.

Projection stretches a patch along its latitude but compresses it vertically because the sphere is tilted. The two effects cancel in the limiting small-patch calculation.

Let d be the distance from a patch to the vertical axis. Similar triangles show that its width scales by R/d as it reaches the cylinder.

A cross-section reveals the second factor. The tangent is perpendicular to the radius; similar right triangles show that the vertical height scales by d/R.

Multiplying R/d by d/R gives one, so infinitesimal area is preserved. The polar endpoints require the limiting interpretation rather than division by zero.

In this particular animation the projected rectangles can also look like rotated originals, but this is a consequence of the chosen parameterization, not a general property of projection.

Finer surface patches make the approximations converge to the smooth surface area. This limiting argument is a geometric introduction to the idea behind calculus.

To see four circles inside the unwrapped cylinder, imagine unrolling a circle’s concentric rings into horizontal strips. Their lengths grow linearly with radius, forming a triangle.

That triangle has base 2πR and height R. Four such triangles fit the cylinder rectangle, relating four circle areas to the sphere.

The second argument compares thin spherical rings directly with their shadows on a plane. Sanderson presents it as a guided exercise instead of supplying every step.

Parameterize a ring by the angle θ from the vertical axis. Its circumference is 2πR sin θ, and for a small angular width dθ its surface area is approximately 2πR² sin θ dθ.

The shadow ring has area approximately 2πR² sin θ cos θ dθ. The identity 2 sin θ cos θ = sin 2θ connects it to a ring at twice the angle.

As θ covers the northern hemisphere, 2θ covers the whole sphere; accounting for the doubled angular interval gives a total shadow area equal to one quarter of the sphere’s area.

Finally, the sphere illustrates a broader result: for a convex three-dimensional body, the average area of its orthogonal shadow over all orientations equals one quarter of its surface area. The video leaves its proof for another time.