Divergence and Curl

Divergence and curl describe two different local features of a vector field. Divergence measures whether the field is spreading out or squeezing in near a point, while curl measures whether it tends to make a small object rotate.

If you remember one contrast, make it this: divergence is about local outflow, and curl is about local spin.

Divergence measures local outflow or inflow

For a 3D vector field

$$\mathbf{F} = (P, Q, R),$$

the divergence is

$$\nabla \cdot \mathbf{F} = \frac{\partial P}{\partial x} + \frac{\partial Q}{\partial y} + \frac{\partial R}{\partial z}.$$

This adds the rate of change of each component in its own direction. If the result is positive at a point, the field is locally acting more like an outward flow there. If it is negative, the field is locally acting more like an inward flow there.

That flow picture is most useful when the vector field is differentiable near the point and actually represents something like velocity.

Curl measures local rotation

For the same 3D field, the curl is

$$\nabla \times \mathbf{F} = \left( \frac{\partial R}{\partial y} - \frac{\partial Q}{\partial z}, \frac{\partial P}{\partial z} - \frac{\partial R}{\partial x}, \frac{\partial Q}{\partial x} - \frac{\partial P}{\partial y} \right).$$

Curl measures local rotation. A nonzero curl means the field has a tendency to make a tiny paddle wheel spin.

In a 2D field $\mathbf{F} = (P, Q)$, many courses use

$$\frac{\partial Q}{\partial x} - \frac{\partial P}{\partial y}$$

as "the curl." Strictly speaking, this is the $z$-component of the 3D curl when the field lies in the plane.

Divergence vs. curl in one worked example

The clearest comparison is to put a pure spreading field next to a pure rotating field.

First, consider

$$\mathbf{F}(x,y) = (x,y).$$

This field points away from the origin, and the arrows get longer as you move farther out. Its divergence is

$$\nabla \cdot \mathbf{F} = \frac{\partial x}{\partial x} + \frac{\partial y}{\partial y} = 1 + 1 = 2.$$

Its 2D curl value is

$$\frac{\partial Q}{\partial x} - \frac{\partial P}{\partial y} = \frac{\partial y}{\partial x} - \frac{\partial x}{\partial y} = 0 - 0 = 0.$$

So this field has positive divergence and no curl. It behaves like pure local spreading without spin.

Now compare that with

$$\mathbf{G}(x,y) = (-y,x).$$

This field circles around the origin. Its divergence is

$$\nabla \cdot \mathbf{G} = \frac{\partial (-y)}{\partial x} + \frac{\partial x}{\partial y} = 0 + 0 = 0.$$

Its 2D curl value is

$$\frac{\partial Q}{\partial x} - \frac{\partial P}{\partial y} = \frac{\partial x}{\partial x} - \frac{\partial (-y)}{\partial y} = 1 - (-1) = 2.$$

So this field has zero divergence but nonzero curl. It behaves like local rotation without net spreading.

That is the main contrast:

$$\mathbf{F}(x,y) = (x,y) \quad \text{spreads out,}$$

while

$$\mathbf{G}(x,y) = (-y,x) \quad \text{swirls around.}$$

If a problem asks what each quantity detects, this example already gives the answer: divergence notices the first field, and curl notices the second.

Common mistakes with divergence and curl

1. Treating divergence and curl as the same kind of measurement. They answer different questions.
2. Forgetting that curl in 2D is often presented as a scalar shortcut, not the full 3D vector.
3. Assuming positive divergence means the vectors are large. Divergence depends on how the field changes, not just on arrow length.
4. Assuming zero divergence means the field is zero. A field can be nonzero everywhere and still have zero divergence.
5. Using the flow interpretation without checking the model. "Source," "sink," and "rotation" are physical intuitions, not automatic facts in every context.

Where divergence and curl are used

Divergence and curl appear in vector calculus, fluid flow, and electromagnetism because they separate two useful local behaviors: expansion and rotation.

In fluid models, divergence can describe local compression or expansion of the flow, while curl can describe local spinning. In electromagnetism, both appear in Maxwell's equations, where they connect field behavior to charge, current, and changing fields.

More broadly, they help you read a vector field instead of only plotting arrows.

A quick mental picture that usually helps

Imagine placing two tiny tools into a field:

1. A tiny balloon tests whether the field tends to expand or compress around a point. That is the divergence idea.
2. A tiny paddle wheel tests whether the field tends to twist it. That is the curl idea.

These are pictures, not definitions, but they are useful pictures when the field is smooth and represents something flow-like.

Try a similar problem

Take the field

$$\mathbf{H}(x,y) = (2x,-2y).$$

Compute its divergence and its 2D curl value. Then decide whether the field behaves more like local spreading, local rotation, both, or neither.

If you want one more check, try $\mathbf{K}(x,y) = (x,-x)$ and see whether the divergence, the curl, or both change.