The Moving Delta Identity

The Key Formula

Advection of spikes

The time rate of change of a delta spike (at a fixed observation point ) equals minus the spatial gradient of the spike, weighted by the spike’s velocity.

---

Why This Matters

This identity is the entire bridge from particle dynamics to continuum field equations. When you have microscopic density and momentum , the moving delta identity lets you convert time derivatives into spatial divergences—exactly what you need for PDEs.

---

Three Ways to Justify It

(1) Chain Rule Picture — The Physicist’s Proof

Setup: Let be the vector from the particle to the observation point.

Key observation: The delta is a function of , not explicitly of or separately:

Apply chain rule:

Compute the time derivative of :

(The gradient has no time dependence; only the position changes.)

Substitute:

The intuition: It’s just the chain rule. The only time-dependence in the delta is through the particle position , so picks up the rate at which that argument is changing.

---

(2) Distribution / Test-Function Proof — The Rigorous Way

For a Dirac delta to be well-defined, we test it against smooth functions. For any smooth function :

Apply the definition of :

Use the chain rule on the RHS:

Use integration by parts on (assuming no boundary terms):

Compare: Both sides give the same result for any test function . Therefore, as distributions:

The key insight: A delta’s ‘value’ is defined by how it acts on test functions. When you test both sides against , you get the same answer—so the distributions are equal.

---

(3) One-Dimensional Sanity Check

Consider a 1D moving spike: with (constant velocity).

Sketch it:

t = 0:        t = Δt:       t = 2Δt:
    ↓             ↓             ↓
    |             |             |
    x = 0         x = vΔt       x = 2vΔt

What you observe at a fixed point :

• If the spike hasn’t arrived yet (): and .
• Just before the spike arrives: the slope of is negative (falling off to the left), so . And you’re about to see the spike, so . [OK]
• Just after the spike passes: the slope is positive (rising to the right, but the spike has moved past), so . And the spike is gone, so . [OK]

The sign checks out: .

The physical picture: At a fixed , the spike moves past you. The time change you see is opposite to the spatial slope it leaves behind. A negative slope means the spike is moving away (future: negative change). A positive slope means it’s approaching (future: positive change).

---

Common Pitfalls & Clarifications

Pitfall 1: “Why the minus sign?”

The question: Why isn’t it just ?

The answer: Because . The argument of the delta is position relative to the particle, so when the particle moves forward, the argument moves backward. The minus sign accounts for this reversal.

---

Pitfall 2: “Shouldn’t act on too?”

The question: When you write , why doesn’t the gradient hit the velocity?

The answer: is the Eulerian gradient—it differentiates w.r.t. the field coordinate , not w.r.t. the particle position . The velocity is a Lagrangian quantity (it depends on and the particle index , not on the field point ). So . See the dyadic note for more on this distinction.

---

Pitfall 3: “This is just kinematics, right?”

Yes! This identity has nothing to do with physics—it’s pure mathematics. It holds for any moving spike, whether the particle is a gas molecule, a dust grain, or a labeled point in a flow. The physics comes in when you specify what equation the particle obeys (Newton’s law, Langevin equation, etc.).

---

Using the Identity: Example — Mass Continuity

Given:

Compute :

Factor out the gradient (it’s linear and acts only on ):

Result:

This is the microscopic continuity equation. No approximation—it’s exact for point particles. The moving delta identity is the entire algebra.

---

Summary: Comparison of the Three Approaches

Approach	Benefits	Trade-offs
Chain rule (physicists)	Intuitive, fast, gets the sign right	Glosses over distribution theory
Test function (rigorous)	Mathematically airtight, generalizes to broader contexts	More abstract, harder to visualize
1D sketch (pedagogical)	Great for intuition, builds confidence	Only works in 1D; requires careful visualization

For deep understanding: Work through all three approaches. The chain rule provides speed and intuition, the test function approach ensures mathematical rigor, and the 1D sketch offers a concrete gut check.

---