Visual Estimation of Partial Derivatives from Contour Maps

Core Definition

On a 2D contour map (level curve plot) representing a multivariable surface $z = f(x, y)$ , each curve connects points of equal constant height $f(x, y) = c$ .

The sign of $\frac{\partial f}{\partial x}$ is determined by taking a step in the positive $x$ -direction (to the right, holding $y$ fixed): crossing into higher contour values means $f_x > 0$ ; crossing into lower contour values means $f_x < 0$ .
The sign of $\frac{\partial f}{\partial y}$ is determined by taking a step in the positive $y$ -direction (upward, holding $x$ fixed): crossing into higher contour values means $f_y > 0$ ; crossing into lower contour values means $f_y < 0$ .
Tangency: If stepping in the coordinate direction is tangent to a level curve, the instantaneous rate of change is zero ( $f_x = 0$ or $f_y = 0$ ).
Magnitude (Steepness): Closely packed contour lines indicate a steep slope (large $|f_x|$ or $|f_y|$ ); widely spaced contour lines indicate a gradual slope (small magnitude).

Concept Explanation

Reading Signs: Step-and-Observe Technique

To evaluate partial derivatives without an explicit formula, use the Step-and-Observe Rule:

                       y ^
                         |          c = 30
                         |       /----------\
                         |      /   c = 20   \
                         |     |   /------\   |
                         |     |  | c = 10 |  |
                         |     |  |   *P   |  |  ---> Step right (+x): crosses 10 -> 20 -> 30  ==> f_x > 0
                         |     |  |        |  |
                         |     |   \------/   |
                         |      \------------/
                         +----------------------------> x

For $\frac{\partial f}{\partial x}(x_0, y_0)$ :
- Stand at the point $(x_0, y_0)$ on the contour map.
- Trace a horizontal ray to the right (increasing $x$ , maintaining $y = y_0$ ).
- If the contour labels increase along this ray, $\frac{\partial f}{\partial x} > 0$ .
- If the contour labels decrease, $\frac{\partial f}{\partial x} < 0$ .
- If the ray points along the contour line itself (tangent), $\frac{\partial f}{\partial x} = 0$ .
For $\frac{\partial f}{\partial y}(x_0, y_0)$ :
- Stand at the point $(x_0, y_0)$ .
- Trace a vertical ray upward (increasing $y$ , maintaining $x = x_0$ ).
- If the contour labels increase along this ray, $\frac{\partial f}{\partial y} > 0$ .
- If the contour labels decrease, $\frac{\partial f}{\partial y} < 0$ .
- If the ray points along the contour line itself (tangent), $\frac{\partial f}{\partial y} = 0$ .

Estimating Numerical Magnitudes (Average Rate of Change)

When level curve spacing $\Delta x$ or $\Delta y$ can be measured, approximate the partial derivative using finite difference quotients:

\frac{\partial f}{\partial x} \approx \frac{\Delta z}{\Delta x} = \frac{c_{\text{next}} - c_{\text{current}}}{\Delta x}

\frac{\partial f}{\partial y} \approx \frac{\Delta z}{\Delta y} = \frac{c_{\text{next}} - c_{\text{current}}}{\Delta y}

High Density (Tightly Spaced Lines): $\Delta x$ is small for a fixed change in elevation $\Delta z \implies \left|\frac{\Delta z}{\Delta x}\right|$ is large (cliff, steep gorge, rapid stress concentration).
Low Density (Widely Spaced Lines): $\Delta x$ is large for the same $\Delta z \implies \left|\frac{\Delta z}{\Delta x}\right|$ is small (plateau, flat valley floor).

Engineering Applications

Environmental & Water Resources Engineering: Topographic contour maps determine stormwater flow directions. Water flows perpendicular to level curves down the steepest descent gradient, while $\frac{\partial E}{\partial x} = 0$ along retention berms.
Materials & Mechanical Stress Auditing: Photoelastic fringe contour maps and FEA stress maps reveal critical stress concentrations where contour fringes crowd together near geometric notches.
Meteorology & Aviation: Atmospheric isobar maps (lines of constant pressure) indicate wind velocity; tightly spaced isobars signify intense pressure gradients $\frac{\partial P}{\partial x}$ that drive severe turbulence.

Example Problems

Example 1: Estimating Derivatives on Concentric Circular Contours

A circular hill has height modeled by level curves $x^2 + y^2 = c$ , with contour intervals of $10\text{ m}$ at $c \in \{10, 20, 30, 40\}$ . At the point $(0, 3)$ on the circle $x^2 + y^2 = 9 \approx 10$ :

Determine the signs of $\frac{\partial f}{\partial x}(0, 3)$ and $\frac{\partial f}{\partial y}(0, 3)$ .
Explain the geometric condition that causes one of these derivatives to vanish.

Solution

Analyze Directional Steps:
- In the $x$ -direction: At $(0, 3)$ , the horizontal line $y = 3$ is tangent to the circle $x^2 + y^2 = 9$ . A small step $\Delta x$ causes no first-order change in distance from the origin ( $r^2 = (\Delta x)^2 + 3^2 \approx 9 + \mathcal{O}((\Delta x)^2)$ ). Thus, the instantaneous slope is zero: $\frac{\partial f}{\partial x}(0, 3) = 0$
- In the $y$ -direction: Stepping upward from $y = 3$ to $y = 4$ increases the distance from the origin ( $r = y$ ), moving toward higher-valued outer contour rings ( $c = 20, 30$ ). Thus, the function is increasing and the partial derivative is positive: $\frac{\partial f}{\partial y}(0, 3) > 0$
Geometric Tangency: Because the level curve is horizontal at the top of the circle $(0, 3)$ , the coordinate direction $+x$ is collinear with the level curve tangent. Since function values are constant along level curves, directional variation along the tangent is identically zero.

Connections & References

Contour Foundations: Level Curves and Contour Maps
Algebraic Core: Partial Derivatives
Geometric Slicing: Geometric Interpretation of Partial Derivatives
Course Sessions: Session 09 - Partial Derivatives

Tony's Intertwingled Memex • Reference Manual