The gradient vector is a central idea in multivariable calculus because it tells how a quantity changes as you move through space. For a scalar field f(x,y), each point in the plane has one output value, such as temperature, height, or pressure. The gradient vector points in the direction where the function increases most rapidly.
This makes it useful for analyzing surfaces, maps, optimization problems, and physical fields.
Understanding Calculus: The Gradient Vector
Each component of the gradient comes from a partial derivative. To find the first component, imagine moving a tiny distance east or west while keeping the north south position fixed. The resulting rate tells how sensitive the output is to horizontal movement.
The second component measures sensitivity to vertical movement. A large positive component means that moving in that coordinate direction raises the value quickly. A negative component means the value falls in that direction.
The size of each component depends on the units used. A temperature change per metre has different units from a height change per kilometre, so sensible unit choices matter when interpreting a gradient.
Contour maps give a useful picture. A contour joins points with the same output value, like a height contour on a hiking map. Walking along one contour does not change the height at first.
Therefore, the gradient cannot point along the contour. It points across the contours. Closely spaced contours show rapid change over a short distance, so the gradient has a larger magnitude there.
Widely spaced contours show a gentler change. This helps students connect a symbolic calculation to a map. It also explains why the gradient points uphill rather than simply toward the visually highest point on a whole map.
Real movement is often constrained to a chosen direction, such as a road, a pipe, or the direction of wind. The change in a scalar field along that path depends on how much the path direction lines up with the gradient. Movement directly uphill gives the full local rate of increase.
Movement across a contour gives no immediate change. Movement partly uphill gives part of the rate, while movement downhill produces a negative rate.
This idea appears in physics when a temperature sensor moves through air, in geography when a walker follows a trail, and in engineering when pressure changes along a channel. The rule uses a dot product, which acts like a measure of alignment between two directions.
A zero gradient needs careful thought. It can occur at the top of a hill, the bottom of a valley, or a saddle point. A saddle can rise in one direction while falling in another, even though the first derivative information is zero at the exact centre.
Further tests are needed to classify such points. Students often use second derivatives or inspect values near the point. This matters in optimization, where a computer may search for a lowest cost or best design.
A zero gradient can signal a possible answer, but it does not prove that the answer is the greatest or least value. Always check the surrounding surface and any constraints on the variables.
Key Facts
- For f(x,y), the gradient is grad f = <partial f/partial x, partial f/partial y>.
- For f(x,y,z), the gradient is grad f = <f_x, f_y, f_z>.
- The directional derivative in the unit direction u is D_u f = grad f dot u.
- The maximum directional derivative equals |grad f| and occurs when u points in the direction of grad f.
- The gradient is perpendicular to level curves f(x,y) = c and perpendicular to level surfaces f(x,y,z) = c.
- If grad f = <0,0> at a point, there is no single direction of steepest increase from the first derivative alone.
Vocabulary
- Gradient vector
- The vector made from the partial derivatives of a scalar function, pointing in the direction of greatest increase.
- Scalar field
- A function that assigns a single numerical value to each point in space, such as temperature or elevation.
- Level curve
- A curve in the input plane where a function f(x,y) has the same constant value.
- Directional derivative
- The rate of change of a function as you move from a point in a chosen direction.
- Magnitude
- The length of a vector, which for the gradient gives the maximum rate of increase of the function.
Common Mistakes to Avoid
- Treating the gradient as a number instead of a vector is wrong because it has both direction and magnitude.
- Forgetting to use a unit vector in D_u f = grad f dot u is wrong because the result then includes the length of the direction vector, not just the rate in that direction.
- Assuming the gradient points along a level curve is wrong because moving along a level curve keeps f constant, so the gradient is perpendicular to it.
- Thinking a larger function value always means a larger gradient is wrong because the gradient measures rate of change, not the value of the function itself.
Practice Questions
- 1 Find grad f at (2,3) for f(x,y) = x^2 y + 4y. Then find its magnitude.
- 2 For f(x,y) = 3x^2 - 2xy + y^2, compute grad f at (1,2) and find the directional derivative in the unit direction u = <3/5,4/5>.
- 3 A contour map shows level curves packed closely together on the left side and spread far apart on the right side. Explain where the gradient magnitude is larger and describe the direction of the gradient relative to the contours.