How Are the Dot Product and Divergence Related?

The dot product and divergence are connected through the symbol $\nabla \cdot \mathbf{F}$. In Cartesian coordinates, this notation tells you to differentiate the components of a vector field and add the derivatives. The result, called the divergence, is a scalar field.

The notation resembles an ordinary dot product, but $\nabla$ is not an ordinary vector of numbers. It is a differential operator. Understanding that distinction explains both the formula and what divergence measures.

The Cartesian formula

For a vector field in three dimensions,

$$\mathbf{F}(x,y,z) = \bigl(F_x(x,y,z), F_y(x,y,z), F_z(x,y,z)\bigr),$$

the divergence is defined in Cartesian coordinates by

$$\nabla \cdot \mathbf{F} = \frac{\partial F_x}{\partial x} + \frac{\partial F_y}{\partial y} + \frac{\partial F_z}{\partial z}.$$

The operator $\nabla$ (also called del or nabla) is written in these coordinates as

$$\nabla = \left(\frac{\partial}{\partial x},\frac{\partial}{\partial y},\frac{\partial}{\partial z}\right).$$

The dot notation indicates that the first differential operator acts on the first component of $\mathbf{F}$, the second on the second, and the third on the third. In two dimensions, omit the $z$ term; in $n$ dimensions, add one matching partial derivative for each coordinate.

Worked example

Consider

$$\mathbf{F}(x,y,z) = (x^2y,\ yz,\ xz^2).$$

Differentiate each component with respect to its corresponding coordinate:

$$\begin{aligned} \nabla \cdot \mathbf{F} &= \frac{\partial (x^2y)}{\partial x} + \frac{\partial (yz)}{\partial y} + \frac{\partial (xz^2)}{\partial z} \\ &= 2xy + z + 2xz. \end{aligned}$$

So the divergence is the scalar field $2xy+z+2xz$. At the point $(1,2,3)$, its value is

$$2(1)(2)+3+2(1)(3)=13.$$

This value describes the field’s local tendency to flow outward near that point. It is not the vector $\mathbf{F}(1,2,3)$ itself; that vector is $(2,6,9)$.

Why call it a dot product?

For ordinary vectors $\mathbf{a}=(a_1,a_2,a_3)$ and $\mathbf{b}=(b_1,b_2,b_3)$, their dot product is $a_1b_1+a_2b_2+a_3b_3$. The divergence formula has the same component-by-component pattern, but the entries of $\nabla$ are derivative operators. When they act on the components of $\mathbf{F}$, they produce derivatives rather than products of numbers.

Thus, $\nabla \cdot \mathbf{F}$ is often described as the dot product of the differential operator $\nabla$ with the vector field. This is useful notation, but it should not obscure the fact that differentiation is happening.

What divergence measures

Imagine a fluid velocity field. If more fluid flows out of a tiny region than flows in, the divergence there is positive. If more flows in than out, it is negative. If inflow and outflow balance locally, the divergence is zero. More precisely, divergence measures net outward flux per unit volume in the limit as the region shrinks to a point.

The same interpretation applies to other vector fields, such as an electric field, although the physical meaning of the field and its units differ. Divergence is a scalar because this local source-or-sink measure has a value at each point, not a direction.

Common mistakes

• Treating $\nabla$ as a fixed numerical vector. Its entries are derivative operators, and they act on the field that follows.
• Differentiating the wrong component. For Cartesian divergence, differentiate $F_x$ with respect to $x$, $F_y$ with respect to $y$, and so on.
• Expecting a vector result. The sum of the matching partial derivatives is a scalar.
• Confusing divergence with $\mathbf{F}\cdot\nabla$. The expression $\mathbf{F}\cdot\nabla$ is a differential operator that acts on a scalar function, producing a directional-derivative-type expression. It is not the divergence of $\mathbf{F}$.
• Using the Cartesian formula in other coordinates without adjustment. In cylindrical or spherical coordinates, divergence has additional scale factors. The simple component formula above applies directly to Cartesian coordinates.

For example, if $f$ is a scalar function, then $\nabla f$ is its gradient, a vector field. Taking the divergence of that gradient gives the Laplacian: $\nabla\cdot(\nabla f)=\nabla^2 f$. This is another use of the same operator notation, with the derivatives acting on the expression to their right.