3.3 Differential and Integral Calculus Applied to Curves, Surfaces, and Surveying Computations

Key Takeaways

  • Differential calculus enables rate-of-change modeling for curve tangents, highway vertical curve slopes, and error sensitivity analysis in geodetic measurements.
  • Partial derivatives (∂f / ∂x_i) form the foundation of variance propagation, evaluating total standard error using the General Law of Error Propagation.
  • Differential optimization (setting f'(x) = 0) identifies high/low points of crest/sag parabolic vertical curves and optimal network geometry.
  • Arc length integration (L = ∫ sqrt(1 + (dy/dx)²) dx) computes precise spatial distances along vertical parabolic curves, clothoid transition spirals, and ellipsoidal meridian arcs.
  • Integral calculus evaluates cross-sectional areas and earthwork volumes using Average End-Area and Prismoidal formulas with prismoidal corrections (C_p).
Last updated: July 2026

3.3 Differential and Integral Calculus Applied to Curves, Surfaces, and Surveying Computations

Calculus provides the essential mathematical framework for analyzing continuous spatial variation in geodetic engineering. From evaluating rate-of-change sensitivity in error propagation to integrating area under terrain profiles and computing earthwork volumes, differential and integral calculus are indispensable tools for professional practice.


Differential Calculus and Error Propagation Analysis

Differential calculus analyzes rates of change, tangent slopes of curves, and sensitivity of computed quantities to measurement errors.

Derivatives of Key Functions in Surveying

  • Trigonometric Functions: ddx(sinx)=cosx,ddx(cosx)=sinx,ddx(tanx)=sec2x\frac{d}{dx}(\sin x) = \cos x, \quad \frac{d}{dx}(\cos x) = -\sin x, \quad \frac{d}{dx}(\tan x) = \sec^2 x
  • Transcendental Functions: ddx(ex)=ex,ddx(lnx)=1x\frac{d}{dx}(e^x) = e^x, \quad \frac{d}{dx}(\ln x) = \frac{1}{x}

General Law of Variance / Error Propagation

When a calculated quantity $Z$ is a function of independent measured variables $X_1, X_2, \dots, X_n$ with standard errors $\sigma_{X_1}, \sigma_{X_2}, \dots, \sigma_{X_n}$, the total variance $\sigma_Z^2$ is determined using partial derivatives:

σZ2=(ZX1)2σX12+(ZX2)2σX22++(ZXn)2σXn2\sigma_Z^2 = \left(\frac{\partial Z}{\partial X_1}\right)^2 \sigma_{X_1}^2 + \left(\frac{\partial Z}{\partial X_2}\right)^2 \sigma_{X_2}^2 + \dots + \left(\frac{\partial Z}{\partial X_n}\right)^2 \sigma_{X_n}^2

Worked Example: Error Propagation in Rectangular Area

Problem: The area of a rectangular survey parcel is $A = W \times L$. Measured dimensions are $L = 100.00 \text{ m} \pm 0.03 \text{ m}$ and $W = 50.00 \text{ m} \pm 0.02 \text{ m}$. Compute the standard error $\sigma_A$ in the calculated area.

Solution:

  1. Compute partial derivatives: AL=W=50.00,AW=L=100.00\frac{\partial A}{\partial L} = W = 50.00, \quad \frac{\partial A}{\partial W} = L = 100.00
  2. Apply the variance propagation formula: σA2=(AL)2σL2+(AW)2σW2=(WσL)2+(LσW)2\sigma_A^2 = \left(\frac{\partial A}{\partial L}\right)^2 \sigma_L^2 + \left(\frac{\partial A}{\partial W}\right)^2 \sigma_W^2 = (W \sigma_L)^2 + (L \sigma_W)^2
  3. Substitute values: WσL=50.00×0.03=1.50 m2W \sigma_L = 50.00 \times 0.03 = 1.50 \text{ m}^2 LσW=100.00×0.02=2.00 m2L \sigma_W = 100.00 \times 0.02 = 2.00 \text{ m}^2 σA2=(1.50)2+(2.00)2=2.25+4.00=6.25\sigma_A^2 = (1.50)^2 + (2.00)^2 = 2.25 + 4.00 = 6.25
  4. Take square root to get standard error: σA=6.25=2.50 m2\sigma_A = \sqrt{6.25} = 2.50 \text{ m}^2

Maxima and Minima Applications in Parabolic Vertical Curve Design

Vertical curves in highway alignment transition between grade tangents using vertical axis parabolas: $y = a x^2 + b x + c$.

Parabolic Curve Equations

If $g_1$ is initial tangent grade (%), $g_2$ is final tangent grade (%), and $L$ is total horizontal length of curve (meters):

y(x)=yPVC+g1x+(g2g12L)x2y(x) = y_{\text{PVC}} + g_1 x + \left(\frac{g_2 - g_1}{2 L}\right) x^2

where $x$ is horizontal distance measured from the Point of Vertical Curvature (PVC).

Finding High or Low Turning Point

To locate the maximum elevation on a crest curve or minimum elevation on a sag curve, set the first derivative (slope) to zero:

dydx=g1+(g2g1L)x=0\frac{dy}{dx} = g_1 + \left(\frac{g_2 - g_1}{L}\right) x = 0 xturn=g1Lg2g1=g1Lg1g2x_{\text{turn}} = \frac{-g_1 L}{g_2 - g_1} = \frac{g_1 L}{g_1 - g_2}

Worked Example: High Point of a Crest Vertical Curve

Problem: A vertical crest curve has an initial grade $g_1 = +3.0%$, final grade $g_2 = -2.0%$, and length $L = 200 \text{ meters}$. Find the distance $x$ from PVC to the highest point on the curve.

Solution:

  1. Identify grades: $g_1 = +0.03$, $g_2 = -0.02$, length $L = 200 \text{ m}$.
  2. Substitute into the turning point formula: x=g1Lg1g2=3.0×2003.0(2.0)=6005.0=120 metersx = \frac{g_1 L}{g_1 - g_2} = \frac{3.0 \times 200}{3.0 - (-2.0)} = \frac{600}{5.0} = 120 \text{ meters}

Arc Length Integration of Curves and Ellipsoidal Meridian Arcs

Arc length integration calculates exact spatial distances along curved paths.

Standard Arc Length Integral

The arc length $S$ of a curve $y = f(x)$ between $x = a$ and $x = b$ is given by:

S=ab1+(dydx)2dxS = \int_{a}^{b} \sqrt{1 + \left(\frac{dy}{dx}\right)^2} dx

Meridian Arc Length on a Reference Ellipsoid

The distance along a meridian on an ellipsoid from equator to latitude $\phi$ requires integrating the meridian radius of curvature $M(\phi)$:

M(ϕ)=a(1e2)(1e2sin2ϕ)3/2M(\phi) = \frac{a(1 - e^2)}{(1 - e^2 \sin^2 \phi)^{3/2}} S(ϕ)=0ϕM(ϕ)dϕ=a(1e2)0ϕ(1e2sin2ϕ)3/2dϕS(\phi) = \int_{0}^{\phi} M(\phi) d\phi = a(1 - e^2) \int_{0}^{\phi} (1 - e^2 \sin^2 \phi)^{-3/2} d\phi

Using binomial expansion, this integral is evaluated as a series:

S(ϕ)=A0ϕA2sin(2ϕ)+A4sin(4ϕ)A6sin(6ϕ)+S(\phi) = A_0 \phi - A_2 \sin(2\phi) + A_4 \sin(4\phi) - A_6 \sin(6\phi) + \dots


Area Integration under Curves and Numerical Integration

Determining cross-sectional area under irregular terrain boundaries uses numerical integration techniques.

Numerical Integration Profiles:
Trapezoidal Rule: Linear segments between offset ordinates
Simpson's 1/3 Rule: Parabolic arcs through triplets of ordinates

Trapezoidal Rule

Assumes linear variation between offset ordinates spaced at uniform interval $d$:

Area=d2[y0+yn+2i=1n1yi]\text{Area} = \frac{d}{2} \left[ y_0 + y_n + 2 \sum_{i=1}^{n-1} y_i \right]

Simpson's 1/3 Rule

Assumes parabolic arcs through sets of three consecutive ordinates (requires an even number of intervals $n$):

Area=d3[y0+yn+4i=oddyi+2i=evenyi]\text{Area} = \frac{d}{3} \left[ y_0 + y_n + 4 \sum_{i=\text{odd}} y_i + 2 \sum_{i=\text{even}} y_i \right]

Comparison of Area Integration Methods

Integration MethodGeometric AssumptionRequired ConditionPrecision Level
Definite IntegralExact continuous function $f(x)$Known mathematical equationExact analytical
Trapezoidal RuleStraight lines between pointsAny number of intervalsGood for straight profiles
Simpson's 1/3 RuleParabolic arcs through 3 pointsEven number of intervals ($n$)Higher precision for curved boundaries

Volume of Earthwork: End-Area and Prismoidal Formulas

Volume calculations determine cut and fill quantities for highways, railways, and site grading.

Average End-Area Formula

Volumeea=(A1+A22)L\text{Volume}_{\text{ea}} = \left(\frac{A_1 + A_2}{2}\right) L

where $A_1$ and $A_2$ are cross-sectional areas at two stations separated by distance $L$.

Prismoidal Formula

Provides exact volume for regular prismoids:

Volumep=L6(A1+4Am+A2)\text{Volume}_p = \frac{L}{6} (A_1 + 4 A_m + A_2)

where $A_m$ is the cross-sectional area at mid-length ($L/2$), calculated from averaged mid-dimensions.

Prismoidal Correction ($C_p$)

Volumep=VolumeeaCp\text{Volume}_p = \text{Volume}_{\text{ea}} - C_p Cp=L12(c1c2)(w1w2)C_p = \frac{L}{12} (c_1 - c_2)(w_1 - w_2)

where $c_1, c_2$ are center cuts/fills and $w_1, w_2$ are total cross-section widths.

Worked Example: Prismoidal Volume Calculation

Problem: Two highway cross sections separated by $L = 100 \text{ m}$ have end areas $A_1 = 450 \text{ m}^2$ and $A_2 = 650 \text{ m}^2$. The mid-section area computed from average dimensions is $A_m = 540 \text{ m}^2$. Compute the prismoidal volume $V_p$.

Solution:

  1. Apply the prismoidal formula: Vp=L6(A1+4Am+A2)V_p = \frac{L}{6} (A_1 + 4 A_m + A_2)
  2. Substitute values: Vp=1006[450+4(540)+650]V_p = \frac{100}{6} [450 + 4(540) + 650] Vp=1006[450+2160+650]=1006[3260]=326000654,333.33 m3V_p = \frac{100}{6} [450 + 2160 + 650] = \frac{100}{6} [3260] = \frac{326000}{6} \approx 54,333.33 \text{ m}^3
Test Your Knowledge

Two highway cross sections separated by L = 100 meters have end areas A1 = 450 m² and A2 = 650 m². If the mid-section area computed from average end dimensions is Am = 540 m², what is the prismoidal volume Vp of earthwork?

A
B
C
D
Test Your Knowledge

The area of a rectangular survey field is given by A = W * L. If measured length is L = 100.00 m ± 0.03 m and measured width is W = 50.00 m ± 0.02 m, what is the standard error σ_A in the calculated area?

A
B
C
D
Test Your Knowledge

A vertical crest curve has an initial grade g1 = +3.0%, a final grade g2 = -2.0%, and a length L = 200 meters. Measured from the PVC, at what distance x does the highest point of the vertical curve occur?

A
B
C
D
Test Your Knowledge

Using Simpson's 1/3 Rule with 4 equal intervals of width d = 10 meters, what is the area bounded by a road centerline and offset ordinates y0 = 3.0 m, y1 = 6.0 m, y2 = 5.0 m, y3 = 7.0 m, and y4 = 3.0 m?

A
B
C
D