GELE Adjustment Computations (Least Squares) — Adjustment of Level Nets and TraversesSummary

When a level loop returns to its starting benchmark after closing through intermediate points, the difference between the calculated final elevation and the known starting elevation is the level misclosure e (in metres). This error arises from instrument calibration drift, human reading errors, and refraction effects. For example, if a loop starting at elevation 100.000 m returns with a calculated elevation of 100.012 m, the misclosure e = +0.012 m. The sign indicates whether the loop was leveled too high (positive) or too low (negative). In Philippine practice, allowable misclosure is defined by the formula: e_allowable = ±K√(distance in km), where K depends on the precision class (typically K = 5–10 mm for most engineering surveys).

The misclosure is distributed back through each section of the level loop in proportion to that section's length or number of setups. The correction to section i is: c_i = −e × (L_i / ∑L), where L_i is the length of section i and ∑L is the total loop length. The negative sign ensures corrections oppose the misclosure. For a loop with misclosure e = +0.012 m and four sections of lengths 1, 2, 3, and 2 km (total 8 km): c_1 = −0.012 × (1/8) = −0.0015 m; c_2 = −0.012 × (2/8) = −0.0030 m; c_3 = −0.012 × (3/8) = −0.0045 m; c_4 = −0.012 × (2/8) = −0.0030 m. The sum of all corrections equals −e, completely cancelling the misclosure.

In a closed traverse, the sum of all latitude components should equal zero (∑Lat = 0), and the sum of all departure components should equal zero (∑Dep = 0). When they do not, the traverse has an error of closure (EC). Geometrically, EC is the straight-line distance from the calculated closing point to the true starting point. It is calculated as: EC = √[(∑Lat)² + (∑Dep)²]. For example, if a 1000 m perimeter traverse has ∑Lat = +0.30 m and ∑Dep = −0.40 m, then EC = √(0.30² + 0.40²) = √0.25 = 0.50 m. The error of closure indicates the total magnitude of misclose; it combines both angular and distance errors.

Relative precision is the ratio of error of closure to the total perimeter of the traverse, expressed as a fraction 1/n. It quantifies the survey's quality independent of size. Relative Precision = EC / Perimeter. Using the previous example: RP = 0.50 / 1000 = 0.0005 = 1/2000. This is stated as '1 in 2000' and means, on average, 1 m of error per 2000 m of traverse. Philippine standards typically require: 1/1000 to 1/500 for general engineering surveys; 1/3000 or better for precise geodetic work (tied to PRS92 control). A survey with RP = 1/2000 would be acceptable for most engineering projects.

The Bowditch rule is the standard method for adjusting closed traverses when angle measurements are reliable but distance measurements may contain proportional errors. It distributes latitude and departure misclosures in proportion to each line's length. The correction to latitude of line i is: c_Lat,i = −(∑Lat) × (L_i / ∑L); similarly for departure: c_Dep,i = −(∑Dep) × (L_i / ∑L). The rule assumes errors increase with distance, so longer legs receive larger corrections. Example: In a 1000 m traverse with ∑Lat = +0.30 m and ∑Dep = −0.40 m, a 250 m line receives: c_Lat = −(0.30) × (250/1000) = −0.075 m; c_Dep = −(−0.40) × (250/1000) = +0.100 m. After correction, each line's adjusted latitude and departure are recalculated using: Lat_adj = Lat_original + c_Lat; Dep_adj = Dep_original + c_Dep.

The transit rule is an alternative adjustment method used when angle measurements are less reliable than distances. Unlike the Bowditch rule (proportional to length), the transit rule distributes misclosures in proportion to the magnitudes of latitude and departure components themselves. Correction to latitude of line i: c_Lat,i = −(∑Lat) × (|Lat_i| / ∑|Lat|); correction to departure: c_Dep,i = −(∑Dep) × (|Dep_i| / ∑|Dep|). This method assumes angular errors are proportional to the projections of the sides, making it better for angle-measurement dominated error sources. The transit rule is less common in Philippine practice than Bowditch but is tested on advanced PRC questions.

Philippine surveying practice (aligned with international standards and RA 4374) sets permissible level misclosures based on the total leveling distance. The standard formula is: Allowable Misclosure = K√D, where D is the total loop distance in kilometres and K is a constant depending on the precision class. Typical values: K = 5 mm/km for third-order leveling (general surveys); K = 3 mm/km for second-order (engineering and property surveys); K = 2 mm/km for first-order (geodetic control networks). For a 10 km level loop at third-order: Allowable = 5 × √10 = 15.8 mm. If the actual misclosure is less than this tolerance, the loop is accepted; if greater, the survey must be re-run or re-checked.

Once corrections are calculated and applied, final adjusted elevations (for level nets) and adjusted coordinates (for traverses) are computed. For leveling: Elevation_adjusted = Elevation_observed + Correction. For traverses: each line's adjusted latitude and departure are calculated, then cumulative coordinates are recomputed starting from the known initial point. Example: If the first section of a level loop has an observed elevation gain of 2.345 m and correction c_1 = −0.0015 m, the adjusted elevation gain is 2.345 + (−0.0015) = 2.3435 m. These adjusted values are then used for all downstream calculations and are the official values entered into survey records and GIS databases.

The misclosure e is the input to the adjustment process. Its magnitude (in mm or cm) determines the size of all corrections. A larger misclosure results in larger corrections distributed across sections or legs. If the misclosure is small (within tolerance), the adjusted values differ only slightly from observed values. This relationship is fundamental: adjustment is reactive to observed misclosure, not proactive.

The Bowditch rule and level-loop adjustment both use a weighted proportional scheme: weights are proportional to distance (or setup count). This is a practical approximation to least-squares adjustment when measurement errors are proportional to distance. In modern adjustment theory (Chapter 2), weights are inverse-variances; for proportional distance, this simplifies to w_i ∝ 1/σ_i², where σ_i is the measurement error of leg i. If error grows with distance, then w_i ∝ 1/L_i.

EC (in metres) tells absolute magnitude; relative precision (as 1/n) tells relative quality. A 0.50 m error in a 1000 m traverse (RP = 1/2000) is acceptable; the same 0.50 m error in a 500 m traverse (RP = 1/1000) is marginal. Both metrics are needed: EC identifies absolute problems; RP identifies whether the survey is consistent with its intended use and RA 4374 standards.

The allowable misclosure e_allowable = K√D (in mm) and relative precision thresholds (1/3000, 1/2000, 1/1000, etc.) are codified in RA 4374 and adopted from ISO/IEC standards. They directly connect field work to legal acceptability. If a survey exceeds tolerance, it is non-compliant and may not be registered with the Land Registration Authority (LRA) or used for property claims under PD 1529. This relationship bridges practical surveying and legal requirements.

These two methods embody different error models. The Bowditch rule is optimal when random distance errors dominate (systematic error in tape, refraction in EDM, or operator bias in pace). The transit rule is optimal when random angle errors dominate (theodolite pointing errors, backsight/foresight asymmetry). In reality, both exist; the choice depends on which source is believed to be larger. Modern surveys often use least-squares (Chapter 2) to weight both error sources simultaneously.

The goal of any adjustment (Bowditch, proportional-distance, or least-squares) is to find adjusted observations such that ∑Lat = 0, ∑Dep = 0 (or final elevation = starting elevation) and the total squared correction is minimized. Simple proportional methods achieve this by distributing the misclosure in a deterministic way. Least-squares methods (Chapter 2) do the same with weighted statistics, yielding formally optimal estimates.

Under CA 141 and RA 8560, all property boundaries and cadastral data entered into the Philippine Geospatial Reference System (PGRS) and the Land Information Management System (LIMS) must be based on adjusted surveys that meet RA 4374 standards and reference the PRS92 datum. Raw, unadjusted observations are not legally acceptable for title registration. This link between adjustment and legal registration is a recurrent theme in the PRC licensure exam.

A contractor establishes a reference benchmark for a 10-storey building project using a closed level loop around the site perimeter (2.5 km total). The misclosure is +0.018 m. Using proportional-distance adjustment with four equal sections (0.625 km each), each section receives a correction of c = −0.018 × (0.625/2.5) = −0.0045 m. The adjusted elevation of each intermediate point is computed, and these become the official benchmarks for construction layout and plumb-bob verification. This application is essential for ensuring building verticality and preventing costly rework during construction.

A licensed geodetic engineer surveyor completes a 850 m perimeter traverse of a residential lot in Makati City using a digital theodolite and steel tape. The traverse closes with ∑Lat = −0.24 m, ∑Dep = +0.18 m, giving EC = 0.30 m and RP = 0.30/850 ≈ 1/2833. Since 1/2833 exceeds the 1/3000 standard for precision, the engineer applies the Bowditch rule to adjust all four corners. The adjusted coordinates are then submitted to the Land Registration Authority (LRA) for property title issuance under CA 141 and PD 1529. Without proper adjustment meeting RA 4374 standards, the LRA would reject the survey.

A provincial government contracts a geodetic survey team to establish control points every 500 m along a 25 km national highway (DPWH project). A primary traverse is run and closes with EC = 0.85 m over 25 km (RP ≈ 1/29,400, excellent). The engineer applies Bowditch correction and distributes the error proportionally across 50 legs. This ensures that all subsequent intermediate points staked from these controls are referenced to a closure-adjusted network, maintaining horizontal accuracy throughout construction staking and grade control.

A consultant establishes a level network connecting an existing PRS92 vertical control point (BM 1001) to a dam site 15 km away. The forward and backward leveling loops each close within tolerance, but the two routes differ by 0.035 m. Adjustment resolves this discrepancy using proportional-distance correction, yielding a consensus elevation for the dam that satisfies both closure constraints. This prevents future disputes over spillway design elevation and ensures compatibility with WGS84 ellipsoidal heights used in modern GNSS surveys.

A large-scale cadastral survey for a mining concession in Mindanao involves a 40 km traverse enclosing 500 hectares. After closure, EC = 1.2 m (RP ≈ 1/33,333), just within acceptable limits. Using Bowditch adjustment, latitudes and departures of all 80 legs are corrected. The adjusted coordinate grid is then used for staking exploration drill holes, mineral reserve delineation, and environmental monitoring plots. Proper adjustment ensures that mineral estimates and area calculations are legally defensible and consistent with Philippine mining regulations (RA 7942).

During a university fieldwork practicum, a student submits a closed 2 km level loop with six sections and a measured misclosure of −0.025 m. The instructor immediately calculates e_allowable = 5√2 ≈ 7.1 mm. Since the student's error (25 mm) far exceeds this, the student is required to return to the field and re-measure. Understanding proportional-distance correction helps the instructor teach the student why the re-measurement is necessary and how to distribute the error once acceptable closure is achieved.

A surveying consultant's quality assurance team reviews monthly traverse work. They create a checklist: (1) Compute EC and RP for each traverse; (2) Verify RP meets the contract specification (e.g., 1/2000 for engineering surveys); (3) If acceptable, apply Bowditch correction and generate adjusted coordinates; (4) If not acceptable, flag the survey for re-measurement. This systematic process, based on the principles of this chapter, ensures all products released to clients meet RA 4374 standards and are legally defensible.