[Paper Review] The Average Projected Area Theorem - Generalization to Higher Dimensions

Zachary Slepian|arXiv (Cornell University)|Sep 3, 2011

Experimental and Theoretical Physics Studies3 references3 citations

TL;DR

This paper generalizes Cauchy's 3D average projected area theorem to arbitrary dimensions using a geometric, physically intuitive approach. It derives a closed-form expression for the ratio $ k(d) $ of average projected area to surface area in $ d $-dimensional convex bodies, showing $ k(d) \propto 1/\sqrt{d} $ asymptotically, and connects this behavior to statistical mechanics via the $ 1/\sqrt{N} $ scaling of standard deviation.

ABSTRACT

In 3-d the average projected area of a convex solid is 1/4 the surface area, as Cauchy showed in the 19th century. In general, the ratio in n dimensions may be obtained from Cauchy's surface area formula, which is in turn a special case of Kubota's theorem. However, while these latter results are well-known to those working in integral geometry or the theory of convex bodies, the results are largely unknown to the physics community---so much so that even the 3-d result is sometimes said to have first been proven by an astronomer in the early 20th century! This is likely because the standard proofs in the mathematical literature are, by and large, couched in terms of concepts that are may not be familiar to many physicists. Therefore, in this work, we present a simple geometrical method of calculating the ratio of average projected area to surface area for convex bodies in arbitrary dimensions. We focus on a pedagogical, physically intuitive treatment that it is hoped will be useful to those in the physics community. We do discuss the mathematical background of the theorem as well, pointing those who may be interested to sources that offer the proofs that are standard in the fields of integral geometry and the theory of convex bodies. We also provide discussion of the applications of the theorem, especially noting that higher-dimensional ratios may be of use for constructing observational tests of string theory. Finally, we examine the limiting behavior of the ratio with the goal of offering intuition on its behavior by pointing out a suggestive connection with a well-known fact in statistics.

Motivation & Objective

• To provide a physically intuitive derivation of the average projected area ratio for convex bodies in arbitrary dimensions, filling a gap in accessibility for physicists.
• To generalize the 3D result—where the average projected area is $ 1/4 $ of the surface area—to $ n $-dimensional convex bodies using geometric reasoning.
• To derive a closed-form expression for the ratio $ k(d) $ of average projected area to surface area in $ d $ dimensions, including a recursion relation and explicit product formulas.
• To explore the limiting behavior of $ k(d) $ as $ d \to \infty $, showing $ k(d) \to 0 $ with asymptotic $ 1/\sqrt{d} $ scaling.
• To draw a conceptual analogy between the $ 1/\sqrt{d} $ scaling of $ k(d) $ and the $ 1/\sqrt{N} $ scaling of standard deviation in statistical mechanics.

Proposed method

• Uses a physical picture of light projection and shadow averaging to define the average projected area in $ d $-dimensional space, avoiding reliance on advanced measure theory.
• Applies Cauchy’s surface area formula and Kubota’s theorem as foundational results, but reformulates them via geometric intuition rather than Grassmannian or Haar measures.
• Derives the key ratio $ k(d) = \frac{1}{\sqrt{\pi}(d-1)M_d} $, where $ M_d = \frac{\Gamma((d-1)/2)}{\Gamma(d/2)} $, using hyperspherical volume and surface area identities.
• Establishes a recursion relation: $ k(d+1) = \frac{1}{2\pi d k(d)} $, anchored at $ k(2) = 1/\pi $ and $ k(3) = 1/4 $, enabling numerical computation.
• Provides explicit product formulas for $ k(d) $: $ k(d) = \frac{1}{2} \prod_{n=0}^{(d-3)/2} \frac{2n+1}{2n+2} $ for odd $ d $, and $ k(d) = \frac{1}{\pi} \prod_{n=0}^{(d-4)/2} \frac{2n+2}{2n+3} $ for even $ d $.
• Performs a series expansion of $ k(d) $ at large $ d $, yielding $ k(d) \approx \frac{1}{\sqrt{2\pi}} \left( d^{-1/2} + \frac{1}{4} d^{-3/2} + \cdots \right) $, confirming the $ 1/\sqrt{d} $ behavior.

Experimental results

Research questions

• RQ1 What is the ratio of average projected area to surface area for a convex body in $ d $-dimensional space, and how does it generalize from the 3D case?
• RQ2 How can the higher-dimensional projected area theorem be derived using physical intuition rather than advanced integral geometry?
• RQ3 What is the asymptotic behavior of the ratio $ k(d) $ as the number of dimensions $ d \to \infty $?
• RQ4 Is there a meaningful connection between the $ 1/\sqrt{d} $ scaling of $ k(d) $ and statistical scaling laws such as $ \sigma \propto 1/\sqrt{N} $?
• RQ5 Can a simple recursion or closed-form expression be derived for $ k(d) $, enabling exact or efficient computation?

Key findings

• The ratio $ k(d) $ of average projected area to surface area in $ d $-dimensional convex bodies is given by $ k(d) = \frac{1}{\sqrt{\pi}(d-1)M_d} $, where $ M_d = \frac{\Gamma((d-1)/2)}{\Gamma(d/2)} $.
• For $ d=3 $, the formula recovers the classical result $ k(3) = 1/4 $, and for $ d=2 $, $ k(2) = 1/\pi \approx 0.318 $.
• The ratio $ k(d) $ decreases monotonically with increasing $ d $, with values ranging from $ k(2) \approx 0.318 $ to $ k(32) \approx 0.071 $, as shown in Table 1.
• Asymptotically, $ k(d) \sim \frac{1}{\sqrt{2\pi}} d^{-1/2} $, with a series expansion confirming $ 1/\sqrt{d} $ scaling to high accuracy even at $ d=5 $.
• A recursion relation $ k(d+1) = \frac{1}{2\pi d k(d)} $, with $ k(3) = 1/4 $, allows exact computation of $ k(d) $ for all $ d \geq 3 $.
• The paper draws a suggestive analogy between the $ 1/\sqrt{d} $ scaling of $ k(d) $ and the $ 1/\sqrt{N} $ scaling of standard deviation in statistical mechanics, offering physical intuition for the result.