4.1 Errors, Corrections, and Pacing/Taping

4.1 Errors, Corrections, and Pacing/Taping

In civil engineering and geomatics operations, no physical measurement is exact. Every observed distance, angle, or elevation difference contains uncertainties inherent to physical sensors, observer perception, and environmental dynamics. Mastering the theory of errors and the mathematical mechanics of tape corrections is essential for both field practice and the Civil Engineering Licensure Examination (CELE).

---

1. Classification and Sources of Errors

A clear distinction exists between human blunders and legitimate scientific errors:

• Mistakes (Blunders): Gross inaccuracies caused by carelessness, misreading graduations, transposing digits in the field book (e.g., recording $54.28\text{ m}$ as $45.28\text{ m}$), or sighting the wrong target. Mistakes cannot be treated mathematically; when detected, the observation must be discarded and repeated.
• Systematic Errors (Cumulative Errors): Errors that follow definite physical and mathematical laws. Under identical conditions, they maintain a constant sign and predictable magnitude (e.g., thermal expansion of a steel tape, a tape manufactured $0.005\text{ m}$ too long). Systematic errors are cumulative—their magnitude compounds as the length of the line increases. They can be modeled, calculated, and mathematically eliminated.
• Accidental Errors (Random Errors): Unavoidable variations that remain after all blunders and known systematic errors have been eliminated. Caused by minute, unpredictable fluctuations in observer judgment, instrument sensitivity, and atmospheric conditions. Random errors follow the laws of probability and tend to be compensating ($E_{\text{total}} \propto \sqrt{n}$). They are adjusted using the method of least squares.

Sources of Errors

1. Instrumental Errors: Imperfections in instrument manufacture, graduation misalignment, worn friction clamps, eccentric verniers, or uncalibrated tape lengths.
2. Personal Errors: Limitations of human sensory perception—such as estimating fractional scale divisions, parallax in optical reticles, or imperfect alignment of a plumb bob string over a survey monument.
3. Natural Errors: Environmental disturbances beyond the surveyor's direct control, including temperature gradients, barometric pressure variations, wind deflection of plumb bobs, solar heating on one side of a transit, and atmospheric refraction.

---

2. Statistical Analysis of Observations & Most Probable Value

When a quantity is measured repeatedly under identical precision standards, random variations produce a distribution of values. Geomatics applies classical probability theory to extract the best estimate of the true magnitude.

Most Probable Value (MPV)

For a series of $n$ independent observations of equal precision, the Most Probable Value (MPV) is the arithmetic mean: $\bar{x} = \frac{\sum x_i}{n}$

Residuals and Standard Deviation

The residual $v_i$ of an individual measurement is its deviation from the MPV: $v_i = x_i - \bar{x}$ Notice that the algebraic sum of residuals for an arithmetic mean is identically zero: $\sum v_i = 0$.

The sample standard deviation $\sigma$ characterizes the dispersion or precision of a single observation: $\sigma = \sqrt{\frac{\sum v_i^2}{n - 1}}$

The standard error of the mean $\sigma_m$ defines the precision of the computed mean itself: $\sigma_m = \frac{\sigma}{\sqrt{n}} = \sqrt{\frac{\sum v_i^2}{n(n - 1)}}$

In classical Philippine board exam problems, the Probable Error (PE) represents the $50\%$ confidence interval ($50\%$ of residuals fall within $\pm PE$): $PE = 0.6745 \sigma = 0.6745 \sqrt{\frac{\sum v_i^2}{n - 1}}$ $PE_m = 0.6745 \sigma_m = \frac{PE}{\sqrt{n}} = 0.6745 \sqrt{\frac{\sum v_i^2}{n(n - 1)}}$

Weighted Measurements

When observations are made under unequal conditions, different instruments, or varying repetition counts, weights ($w_i$) are assigned to reflect relative reliability:

• Weights based on precision: Weights are inversely proportional to variances or the square of probable errors: $w_i \propto \frac{1}{\sigma_i^2} \propto \frac{1}{PE_i^2}$.
• Weights based on repetitions: If a line is measured $n_i$ times, the weight is directly proportional to repetitions: $w_i \propto n_i$.
• Weights based on route distance: For differential leveling lines of length $D_i$, random error compounds as $\sqrt{D_i}$, meaning variance is proportional to $D_i$. Consequently, weight is inversely proportional to distance: $w_i \propto \frac{1}{D_i}$.

The Weighted Most Probable Value is: $\bar{x}_w = \frac{\sum w_i x_i}{\sum w_i}$

---

3. Pacing and Pace Factor Calibration

Pacing is a rapid, non-instrumental distance estimation method utilized during preliminary site reconnaissance and traverse error detection.

• Pace: The distance covered in one normal forward step, measured heel-to-heel or toe-to-toe.
• Stride: A double pace (two consecutive steps), measured from the heel of one foot to the next placement of the same foot ($1\text{ stride} = 2\text{ paces}$).

Pace Factor Calibration Protocol

1. Establish a straight, horizontal baseline of known length $D$ (typically $50.00\text{ m}$ to $100.00\text{ m}$) using a calibrated steel tape.
2. Walk the baseline at a natural, consistent stride across at least 4 to 6 trials.
3. Record the number of paces for each trial ($N_1, N_2, \dots, N_k$) and determine the mean pace count $\bar{N} = \frac{\sum N_i}{k}$.
4. Calculate the individual's Pace Factor (PF) in meters per pace: $\text{PF} = \frac{\text{Known Baseline Distance } D}{\text{Mean Number of Paces } \bar{N}} \quad (\text{m/pace})$
5. To determine the horizontal length of an unknown line, count paces $N_{\text{unknown}}$ and apply: $D_{\text{unknown}} = N_{\text{unknown}} \times \text{PF}$

---

4. Taping Corrections: Physical Principles and Derivations

Steel tapes are manufactured and calibrated under standardized laboratory conditions: a standard temperature ($T_0 = 20^\circ\text{C}$ or $68^\circ\text{F}$), standard tension ($P_0 = 50\text{ N}$ or $10\text{ lb}$ to $15\text{ lb}$), and supported horizontally throughout their entire length.

When deployed in the field under non-standard conditions, mathematical corrections must be calculated and applied.

The Cardinal Rule: Measuring vs. Laying Out (T-L-A-M-S-L)

Every taping scenario falls into one of two fundamental categories:

1. Measuring an Unknown Distance: Monuments already exist in the field. The tape is laid between them. If the tape is Too Long, each graduated interval spans more ground than labeled, so fewer tape lengths fit into the distance; the raw reading is too small. Hence, ADD the correction to find the true distance.
2. Laying Out a Specified Distance: The desired distance is fixed on design drawings (e.g., establishing a $30.000\text{ m}$ column center). If the tape is Too Long, measuring out to the $30.000\text{ m}$ mark would lay out too much ground. Hence, SUBTRACT the correction on the tape graduations to place the field marker accurately.

---

1. Absolute Length / Calibration Correction ($C_L$)

If a nominal $30\text{-m}$ tape actually measures $L_{\text{actual}} = 30.008\text{ m}$, the error per tape length is $c_L = 30.008 - 30.000 = +0.008\text{ m}$ (Tape is too long). For an observed line length $L$: $C_L = c_L \left( \frac{L}{L_{\text{nominal}}} \right)$

2. Temperature Correction ($C_t$)

Steel expands with rising temperature and contracts with cooling. The correction is modeled by thermal expansion mechanics: $C_t = \alpha L (T - T_0)$ Where:

• $\alpha$ = coefficient of linear expansion for steel ($11.6 \times 10^{-6}\text{ /}^\circ\text{C}$ or $0.0000116\text{ /}^\circ\text{C}$)
• $L$ = measured or target length (m)
• $T$ = field temperature during observation ($^\circ\text{C}$)
• $T_0$ = standardization temperature ($20^\circ\text{C}$)

Sign Convention: If $T > T_0$, $C_t$ is positive (tape expands $\rightarrow$ tape too long). If $T < T_0$, $C_t$ is negative (tape contracts $\rightarrow$ tape too short).

3. Tension / Pull Correction ($C_p$)

Under Hooke's Law for elastic axial deformation: $C_p = \frac{(P - P_0) L}{A E}$ Where:

• $P$ = applied tensile pull in the field (N)
• $P_0$ = standard pull (N)
• $L$ = measured length (m)
• $A$ = cross-sectional area of the tape ($\text{m}^2$ or $\text{mm}^2$)
• $E$ = modulus of elasticity of steel (typically $2.0 \times 10^{11}\text{ N/m}^2 = 200\text{ GPa} = 2.0 \times 10^5\text{ MPa}$)

Sign Convention: If $P > P_0$, the tape stretches ($C_p > 0$, tape too long). If $P < P_0$, the tape shortens ($C_p < 0$, tape too short).

4. Sag Correction ($C_s$)

When a tape is supported only at its ends (or at intermittent intervals) rather than along a continuous flat surface, gravity pulls the tape into a catenary curve. The horizontal chord distance between end supports is strictly shorter than the curved tape length. $C_s = \frac{w^2 L^3}{24 P^2} = \frac{W^2 L}{24 P^2}$ Where:

• $w$ = linear weight of the tape per unit length ($\text{N/m}$ or $\text{kg/m} \times 9.81\text{ m/s}^2$)
• $W = w L$ = total weight of the unsupported tape span (N)
• $L$ = unsupported span length (m)
• $P$ = applied field tension (N)

Sag Correction for Multiple Unsupported Spans

If a tape of total length $L$ is supported at $n$ equal spans of length $l = L/n$: $C_{s, \text{total}} = n \left[ \frac{w^2 (L/n)^3}{24 P^2} \right] = \frac{w^2 L^3}{24 n^2 P^2} = \frac{W^2 L}{24 n^2 P^2}$ Notice that adding intermediate supports drastically reduces sag: supporting the tape at the midpoint ($n = 2$) reduces total sag correction to $\frac{1}{4}$ of the single-span value.

5. Normal Tension ($P_n$)

The tension $P_n$ at which the positive elongation due to pull exactly offsets the negative shortening due to sag ($C_p = C_s$): $\frac{(P_n - P_0)L}{AE} = \frac{W^2 L}{24 P_n^2} \implies P_n^2 (P_n - P_0) = \frac{W^2 A E}{24}$ Because $P_n$ appears in cubic form, solving for normal tension requires iterative numerical methods or direct substitution in board examinations.

6. Slope Correction ($C_h$)

Measurements taken along an inclined slope of length $s$ with a vertical elevation difference $h$ must be reduced to the horizontal projection $d$: $d = \sqrt{s^2 - h^2} = s \cos \theta$ Using binomial series expansion ($h \ll s$): $C_h = s - d = s - \sqrt{s^2 - h^2} \approx \frac{h^2}{2s} + \frac{h^4}{8s^3}$ For slopes less than $10\%$, the first term provides sub-millimeter accuracy: $C_h \approx \frac{h^2}{2s}$ Horizontal distance is: $d = s - C_h$.

---

5. Comprehensive Worked Examples

Worked Example 1: Full Measurement Corrections

Problem: A $50.000\text{-m}$ steel tape was standardized at $20^\circ\text{C}$ and $50\text{ N}$ pull, supported throughout its length, having an actual length of $50.005\text{ m}$. The tape cross-sectional area is $3.20\text{ mm}^2$, linear weight is $0.025\text{ kg/m}$ ($w = 0.025 \times 9.8066 = 0.2452\text{ N/m}$), and $E = 2.0 \times 10^5\text{ MPa}$. A line was measured on flat terrain as exactly $150.000\text{ m}$ (three full tape spans) with the tape supported at the ends of each $50\text{-m}$ span only, under an applied pull of $90\text{ N}$ at an average field temperature of $34^\circ\text{C}$. Calculate the true horizontal length of the measured line.

Solution:

1. Absolute Length Correction ($C_L$): $c_L = 50.005 - 50.000 = +0.005\text{ m (Tape is too long)}$ $C_L = +0.005 \times \left( \frac{150.000}{50.000} \right) = +0.0150\text{ m}$

2. Temperature Correction ($C_t$): $C_t = \alpha L (T - T_0) = (11.6 \times 10^{-6}\text{ /}^\circ\text{C})(150.000\text{ m})(34 - 20)^\circ\text{C}$ $C_t = (11.6 \times 10^{-6})(150)(14) = +0.0244\text{ m}$

3. Pull/Tension Correction ($C_p$): $A = 3.20\text{ mm}^2 = 3.20 \times 10^{-6}\text{ m}^2, \quad E = 2.0 \times 10^{11}\text{ N/m}^2$ $A E = 3.20 \times 10^{-6} \times 2.0 \times 10^{11} = 640,000\text{ N}$ $C_p = \frac{(P - P_0)L}{AE} = \frac{(90 - 50)(150.000)}{640,000} = \frac{40 \times 150}{640,000} = +0.0094\text{ m}$

4. Sag Correction ($C_s$): The line was measured in 3 separate $50\text{-m}$ spans ($n = 3$, $l = 50.000\text{ m}$): $W_{\text{span}} = w \cdot l = 0.2452\text{ N/m} \times 50\text{ m} = 12.26\text{ N}$ $C_{s, \text{span}} = \frac{W^2 l}{24 P^2} = \frac{(12.26)^2 \times 50.000}{24 \times (90)^2} = \frac{150.3076 \times 50}{194,400} = 0.03866\text{ m}$ $C_{s, \text{total}} = 3 \times 0.03866\text{ m} = 0.1160\text{ m}$ (Sag is subtractive in measurement!)

5. True Distance Determination: $L_{\text{true}} = L_{\text{meas}} + C_L + C_t + C_p - C_s$ $L_{\text{true}} = 150.0000 + 0.0150 + 0.0244 + 0.0094 - 0.1160 = 149.9328\text{ m} \approx 149.933\text{ m}$

---

Worked Example 2: Layout / Setting Out Under Field Conditions

Problem: An engineer must lay out a building foundation edge measuring exactly $90.000\text{ m}$ using a steel tape that is standardized as $30.000\text{ m}$ at $20^\circ\text{C}$ and $60\text{ N}$ pull throughout. In the field, the temperature is $10^\circ\text{C}$, the pull is maintained at $60\text{ N}$, and the tape will be supported throughout. The tape calibration certificate states that under standard conditions, its actual length is $29.994\text{ m}$. What distance must be laid out on the tape to establish the exact $90.000\text{ m}$ building footprint?

Solution:

1. Tape Condition Analysis: $c_L = 29.994 - 30.000 = -0.006\text{ m (Tape is too short)}$ $C_L = -0.006 \times \left( \frac{90.000}{30.000} \right) = -0.0180\text{ m}$

2. Temperature Correction ($C_t$): $C_t = (11.6 \times 10^{-6})(90.000)(10 - 20) = -0.0104\text{ m}$

3. Combined Net Error of the Tape: $C_{\text{net}} = C_L + C_t = -0.0180 + (-0.0104) = -0.0284\text{ m}$ The tape is net TOO SHORT by $0.0284\text{ m}$.

4. Application to Layout: Recall the rule: Tape Too Short $\rightarrow$ ADD to Lay Out! $L_{\text{layout}} = L_{\text{target}} - C_{\text{net}} = 90.000 - (-0.0284) = 90.0284\text{ m} \approx 90.028\text{ m}$ Verification: Because the tape contracts and is manufactured short, placing markers at $90.0284\text{ m}$ on this shrunken tape results in an actual ground distance of exactly $90.0284 - 0.0284 = 90.000\text{ m}$.