# Nabla Operator and its 3 Most Important Applications

## Table of contents

- 3 basic ways to apply Nabla to functions Here you will learn how to form the gradient, divergence and rotation of a function using the Nabla operator.
- Apply Nabla to Nabla Here you apply the Nabla operator to Nabla operator using the scalar and cross product.
- 5 useful ways to apply Nabla twice Here you will learn how to get divergence of the gradient, divergence of the rotation and so on when the Nabla operator is applied twice to a function.
- The 9 most useful rules with Nabla Here you will learn some important mathematical rules that you can use to simplify or rewrite expressions with Nabla.

Nabla operator \(\nabla\) is notationally similar to a vector and looks like this in the three-dimensional case when we express it with Cartesian coordinates:

**Nabla operator in 3d**

The three components of the Nabla operator are **partial derivatives** with respect to \(x\), \(y\) or \(z\). Derivatives that occur alone are called differential operators. You can apply a differential operator to a function. The result is the derivative of the function.

The Nabla operator unfolds its effect only if it is applied to a scalar or vector function. The result is a *multidimensional derivative of the function*.

## 3 basic ways to apply Nabla to functions

The Nabla operator can be applied to **scalar functions** \( f(x,y,z) \) as well as to **vector functions**. A 3d vector function has three components:

**Example of 3d vector function**

The components of a vector function are scalar functions like \( f(x,y,z) \). So you can think of a scalar function as a 1d vector function that has exactly one component. If a vector function, as in Eq. 2

, depends on the space coordinates \( x \), \(y\), and \(z\), then we call it a **vector field**. A component of the vector field, such as the first component \( F_{1}(x,y,z) \), is also sometimes written down with index '\(\text{x}\)' instead of index '1' to indicate that it is the component of the vector field that points along the \(x\) direction: \( F_{\text{x}}(x,y,z) \).

So a vector field is nothing else than a vector \( \boldsymbol{F} \) which can change its length and direction depending on its position in space.

You can manipulate a vector in different ways:

**Scalar multiplication**- you can multiply a vector by a real number, let's call it \( a \in \mathbb{R} \): \( \boldsymbol{F} \, a \). For example, the number \( a \) could be the scalar function \( f \) and \( \boldsymbol{F} \) could be the Nabla operator.You can form a

**scalar product**of the vector with another vector \( \boldsymbol{R} \): \( \boldsymbol{R} \cdot \boldsymbol{F} \). For example, the nabla operator could be this vector \( \boldsymbol{R} \).You can form a

**cross product**of the vector with another vector \( \boldsymbol{R} \): \( \boldsymbol{R} \times \boldsymbol{F} \). For example, the nabla operator could be this vector \( \boldsymbol{R} \).

### #1: Scalar multiplication with Nabla

Let's calculate what we would get if we apply the Nabla operator \( \nabla \) to a scalar function \( f(x,y,z) \) that depends on three variables.

**Scalar multiplication with Nabla = Gradient**

In the second step, we just used the definition of the Nabla operator, and in the last step, we pulled the scalar function (a number, so to speak) into the vector and omitted the dependence of \(x\), \(y\), and \(z\) to make the result more compact.

The result of Nabla applied to \( f \) is called a **gradient** and obviously represents a three-dimensional vector field with three components:

The

*1st component*contains the slope \( \frac{\partial f}{\partial x} \) in \( x \) direction.The

*2nd component*contains the slope \( \frac{\partial f}{\partial y} \) in \( y \) direction.The

*3rd component*contains the slope \( \frac{\partial f}{\partial z} \) in \( z \) direction.

Note: Such scalar multiplication as in 3

is not commutative, so mathematically speaking, the nabla operator is not a proper vector.

Of course, you may also use a two-dimensional Nabla operator, which has only two (and not three) components. You could then calculate a two-dimensional gradient as follows:

**Gradient in two dimensions**

And the Nabla operator in one dimension is simply a partial derivative \( \frac{\partial f}{\partial x} \).

### #2: Scalar product with Nabla

Another way to combine the Nabla operator this time with a vector field \(\boldsymbol{F}(x,y,z)\) as in 2

is to form a scalar product:

**Scalar product with Nabla = divergence**

&~=~ \frac{\partial F_{\text x}}{\partial x} + \frac{\partial F_{\text y}}{\partial y} + \frac{\partial F_{\text z}}{\partial z} \end{align} $$

In the scalar product, you apply the derivatives to the scalar functions (vector field components) componentwise:

Form the derivative of the first component \( F_{\text x}(x,y,z) \) to \(x\).

Form the derivative of the second component \( F_{\text y}(x,y,z) \) to \(y\).

Form the derivative of the third component \( F_{\text z}(x,y,z) \) to \(z\).

Add the three derivatives together.

The result of the scalar product of Nabla with \( \boldsymbol{F} \) is called **divergence** and represents a three-dimensional scalar function \( f(x,y,z) \). In the case of gradient, a vector function \( \boldsymbol{F} \) was obtained from a scalar function \(f\). In the case case of divergence, we make a scalar function out of a vector function. So exactly the other way around!

If you apply the nabla operator \( \nabla \) to a vector function \(\boldsymbol{F}\) using the scalar product, then the result \( \nabla \cdot \boldsymbol{F} \) is called *divergence of* \(\boldsymbol{F}\).

The following three-dimensional vector field is given:

**Example: 3d vector field**

Form the scalar product of the Nabla operator with the vector field \( \boldsymbol{F} \):

**Example calculation: divergence of a 3d vector field**

&~=~ 6x^2 + z \end{align} $$

### #3: Cross product with Nabla

As with the scalar product 6

, you apply the nabla operator to a vector function \( \boldsymbol{F}(x,y,z) \) to form the cross product:

**Cross product with Nabla = rotation**

&~=~ \begin{bmatrix} \frac{\partial F_z}{\partial y} - \frac{\partial F_y}{\partial z} \\ \frac{\partial F_x}{\partial z} - \frac{\partial F_z}{\partial x} \\ \frac{\partial F_y}{\partial x} - \frac{\partial F_x}{\partial y} \end{bmatrix} \end{align} $$

The result of the cross product 9

is a vector again! Nabla leaves here the vector function \(\boldsymbol{F}\) as vector function.

If you apply the nabla operator \(\nabla\) to a vector function \(\boldsymbol{F}\) using the cross product, the resulting vector \(\nabla \times \boldsymbol{F} \) is called the *rotation of* \(\boldsymbol{F}\).

Consider again the vector field as in Eq. 7

:

**Example vector field for the rotation of the vector field**$$ \begin{align} \boldsymbol{F}(x,y,z) ~=~ \begin{bmatrix} 2x^3 \\ zy \\ 5xy \end{bmatrix} \end{align} $$

Apply the nabla operator to \( \boldsymbol{F} \) by forming the cross product:

**Example calculation: Rotation of a 3d vector field**

## Apply Nabla to Nabla

Of course you can also form the scalar and cross product of Nabla with Nabla. The scalar product gives a new operator \( \nabla \cdot \nabla \), which we call the **Laplace operator**:

**Laplace operator: scalar product of two Nabla operators**

&~=~ \frac{\partial^2}{\partial x^2} + \frac{\partial^2}{\partial y^2} + \frac{\partial^2}{\partial z^2} \\\\

&~=:~ \nabla^2 \end{align} $$

The Laplace operator is the sum of the *second* derivatives with respect to \(x\), \(y\) and \(z\). The scalar product \( \nabla \cdot \nabla \) is written as \( \nabla^2 \) for short (sometimes also as \(\Delta\)).

The cross product of two Nabla operators is not interesting because it always gives the zero vector. Why? Because the partial derivatives are commutative and thus cancel each other out inside the Nabla-Nabla cross product:

**Cross product of Nabla operator with itself is zero**

So if you apply the Nabla cross product 12

to any vector function: \( (\nabla \times \nabla) ~\cdot~ \boldsymbol{F} = \boldsymbol{0} \) or \( (\nabla \times \nabla) ~\times~ \boldsymbol{F} = \boldsymbol{0}\) then you always get the zero vector.

**Note!** Associativity does not apply:

**Examples: Cross product with nabla is not associative**$$ \begin{align} (\nabla \times \nabla) ~\cdot~ \boldsymbol{F} ~\neq~ \nabla \times (\nabla ~\cdot~ \boldsymbol{F}) \\\\

0 ~=~ (\nabla \times \nabla) ~\times~ f ~\neq~ \nabla \times (\nabla \times f) \end{align} $$

## 5 useful ways to apply Nabla twice

In electrodynamics, fluid dynamics, and other areas of physics, there are equations in which Nabla operator is applied twice to a vector field or scalar field. There are exactly five useful different possibilities to apply Nabla twice:

### 1st possibility: divergence of the gradient

If you apply the Laplace operator 11

to a scalar function \( f(x,y,z) \), then you get the **divergence of the gradient of \( f\)**:

**Formula: Divergence of the gradient**

&~=~ \frac{\partial^2 f}{\partial x^2} + \frac{\partial^2 f}{\partial y^2} + \frac{\partial^2 f}{\partial z^2} \end{align} $$

By the way, associativity applies here, so parentheses are redundant: \( (\nabla \cdot \nabla) \, f ~=~ \nabla \cdot (\nabla \, f) ~=~ \nabla \cdot \nabla \, f\).

### 2nd possibility: divergence of rotation

For this you need of course a vector function \( \boldsymbol{F} \), because the rotation is defined only for a vector function. You have already calculated the rotation of \( \boldsymbol{F} \) in 9

. FJust form the scalar product with the result 9

of the rotation:

**Formula: Divergence of rotation**

&~+~ \frac{\partial}{\partial z}\left( \frac{\partial F_y}{\partial x} ~-~ \frac{\partial F_x}{\partial y} \right) ~=~ 0 \end{align} $$

As you can see from the result: Divergence of rotation is always zero. A physical example is the magnetic field \( \boldsymbol{B} = \nabla \times \boldsymbol{A} \) (with \( \boldsymbol{A} \) as vector potential). Divergence of the magnetic field vanishes: \( \nabla \cdot \boldsymbol{B} = 0 \).

### 3rd possibility: rotation of the gradient

Here you first form the gradient of a scalar function \( f \) as in 3

and then you form the cross product of the Nabla operator with the resulting vector \( \nabla \, f \) as in 9

:

**Formula: Rotation of the gradient**

&~=~ \begin{bmatrix} \frac{\partial}{\partial y}\frac{\partial f}{\partial z} - \frac{\partial}{\partial z}\frac{\partial f}{\partial y} \\ \frac{\partial}{\partial z}\frac{\partial f}{\partial x} - \frac{\partial}{\partial x}\frac{\partial f}{\partial z} \\ \frac{\partial}{\partial x}\frac{\partial f}{\partial y} - \frac{\partial}{\partial y}\frac{\partial f}{\partial x} \end{bmatrix} ~=~ 0 \end{align} $$

The rotation of the gradient is always zero. A physical example: The *electrostatic* field can be written as a gradient of a scalar potential \(V\): \( \boldsymbol{E} = \nabla \, V \). Thanks to our result we can conclude that the rotation of the E-field vanishes: \( \nabla \times \boldsymbol{E} = 0 \). We say: Electrostatic E-fields are therefore vortex-free!

### 4th possibility: rotation of the rotation

You can apply the Nabla operator twice as a cross product to a vector function (see Eq. 8

for how to do this):

**Formula: Rotation of the rotation**

&~=~ \begin{bmatrix} \frac{\partial}{\partial y}\frac{\partial F_y}{\partial x} - \frac{\partial^2 F_x}{\partial y^2} - \frac{\partial^2 F_x}{\partial z^2} + \frac{\partial}{\partial z}\frac{\partial F_z}{\partial x} \\ \frac{\partial}{\partial z}\frac{\partial F_z}{\partial y} - \frac{\partial^2 F_y}{\partial z^2} - \frac{\partial^2 F_y}{\partial x^2} + \frac{\partial}{\partial x}\frac{\partial F_x}{\partial y} \\ \frac{\partial}{\partial x}\frac{\partial F_x}{\partial z} - \frac{\partial^2 F_z}{\partial x^2} - \frac{\partial^2 F_z}{\partial y^2} + \frac{\partial}{\partial y}\frac{\partial F_y}{\partial z} \end{bmatrix} \end{align} $$

### 5th possibility: gradient of divergence

The last useful way is to form the divergence \( \nabla \cdot \boldsymbol{F} \) first. The result is a scalar function. And then apply the nabla operator to this scalar function:

**Formula: Gradient of divergence**

&~=~ \begin{bmatrix} \frac{\partial^2 F_x}{\partial x^2} + \frac{\partial }{\partial x}\frac{\partial F_y}{\partial y} + \frac{\partial }{\partial x}\frac{\partial F_z}{\partial z} \\ \frac{\partial }{\partial y}\frac{\partial F_x}{\partial x} + \frac{\partial^2 F_y}{\partial y^2} + \frac{\partial }{\partial y}\frac{\partial F_z}{\partial z} \\ \frac{\partial }{\partial z}\frac{\partial F_x}{\partial x} + \frac{\partial }{\partial z}\frac{\partial F_y}{\partial y} + \frac{\partial^2 F_z}{\partial z^2} \end{bmatrix} \end{align} $$

All the other conceivable cases: "*Rotation of divergence*", "*Gradient of rotation*", "*Divergence of divergence*" and "*Gradient of gradient*" are not defined and do not occur in physics at all.

## The 9 most useful rules with Nabla

With the knowledge you have just acquired, you can derive the following mathematical rules with the Nabla operator. You will encounter these rules in electrodynamics, for example, because they can be used to simplify and to transform certain expressions.

**Distributivity in multiplication**.**Rule: Distributivity for multiplication**Formula anchor $$ \begin{align} \nabla \, (f + g ) ~=~ \nabla f ~+~ \nabla g \end{align} $$**Distributivity with scalar product****Rule: Distributivity with scalar product**Formula anchor $$ \begin{align} \nabla \cdot (\boldsymbol{F} + \boldsymbol{R} ) ~=~ \nabla \cdot \boldsymbol{F} ~+~ \nabla \cdot \boldsymbol{R} \end{align} $$**Distributivity with cross product****Rule: Distributivity with cross product**Formula anchor $$ \begin{align} \nabla \times (\boldsymbol{F} + \boldsymbol{R} ) ~=~ \nabla \times \boldsymbol{F} ~+~ \nabla \times \boldsymbol{R} \end{align} $$**Product rule for scalar functions****Rule: Product rule for scalar functions**Formula anchor $$ \begin{align} \nabla (f \, g) ~=~ f \, \nabla g ~+~ g \nabla f \end{align} $$**Product rule for scalar and vector function****Rule: Product rule for scalar and vector function**Formula anchor $$ \begin{align} \nabla \, \cdot \, \left( f \, \boldsymbol{R} \right) ~=~ f \, \left( \nabla \cdot \boldsymbol{R} \right) ~+~ \boldsymbol{R} \cdot \left( \nabla f \right) \end{align} $$**Triple product with Nabla****Rule: Triple product with Nabla**Formula anchor $$ \begin{align} \nabla \cdot \left( \boldsymbol{F} \times \boldsymbol{R} \right) ~=~ \boldsymbol{R} \cdot \left( \nabla \times \boldsymbol{F} \right) ~-~ \boldsymbol{F} \cdot \left( \nabla \times \boldsymbol{R} \right) \end{align} $$**Double cross product with Nabla****Rule: Double cross product with Nabla**Formula anchor $$ \begin{align} \nabla \times \left( \boldsymbol{F} \times \boldsymbol{R} \right) &~=~ \left( \boldsymbol{R} \cdot \nabla \right) \, \boldsymbol{F} ~-~ \left( \boldsymbol{F} \cdot \nabla \right) \, \boldsymbol{R} \\\\

&~+~ \boldsymbol{F} \, \left( \nabla \cdot \boldsymbol{R} \right) ~-~ \boldsymbol{R} \, \left( \nabla \cdot \boldsymbol{F} \right) \end{align} $$**Gradient of a scalar product**.**Rule: Divergence of a scalar product**Formula anchor $$ \begin{align} \nabla \, \left( \boldsymbol{F} \cdot \boldsymbol{R} \right) &~=~ \boldsymbol{F} \times \left( \nabla \times \boldsymbol{R} \right) ~+~ \boldsymbol{R} \times \left( \nabla \times \boldsymbol{F} \right) \\\\

&~+~ \left( \boldsymbol{R} \cdot \nabla \right) \, \boldsymbol{F} ~+~ \left( \boldsymbol{F} \cdot \nabla \right) \, \boldsymbol{R} \end{align} $$**Rotation of a scaled vector field****Rule: Rotation of a scaled vector field**Formula anchor $$ \begin{align} \nabla \times \left( f \, \boldsymbol{R} \right) ~=~ f \, \left( \nabla \times \boldsymbol{R} \right) ~-~ \boldsymbol{R} \times \left( \nabla f \right) \end{align} $$

For example, if you swap the function \( \boldsymbol{F} \) with Nabla in the scalar product 6

, you get a **new operator**:

**Scalar product with nabla and vector field interchanged**

Just like the nabla operator, you can apply the operator 27

to any function \( f \): \( (\boldsymbol{F} \cdot \nabla)\,f\).

Now you have learned what the nabla operator is and how to apply it to scalar and vector functions. In the next lesson, we'll take a closer look at the gradient \( \nabla \, f \) and learn about the directional derivative.