GELE Geodesy — The Geoid, Gravity and HeightsStudy Notes
Detailed study notes for GELE Geodesy — The Geoid, Gravity and Heights. These are the kind of notes you would take if you were reviewing with someone who has already scored well on the GELE: organised by what Professional Regulation Commission (PRC) — Board of Geodetic Engineering tests first, followed by the nice-to-knows, and ending with the traps to avoid.
Exam context
The Geodetic Engineer Licensure Examination is conducted by Professional Regulation Commission (PRC) — Board of Geodetic Engineering and is scheduled for September 2026. The Geodesy subtest is marked as "Core" in the official pattern, and The Geoid, Gravity and Heights appears in position 5th of 6 in the GELE Geodesy review rotation. Passing mark: 70% weighted average, no sub-test below 50%. Recent GELE 2026 papers have drawn roughly a meaningful share of questions from this subject.
The Geoid, Gravity and Heights - Study Notes
The relationship between the geoid, gravity, and height systems forms the foundation of modern geodetic practice in the Philippines and globally. This chapter addresses three fundamental surfaces that geodetic engineers must understand: the mathematical ellipsoid (WGS84/PRS92), the geoid (the true equipotential surface defining mean sea level), and the terrain where we work. Understanding the conversion between ellipsoidal heights (h) from GNSS receivers and orthometric heights (H) used in engineering is critical for survey work under Philippine standards (RA 4374 — Cadastral Law; PD 1529 — Property Registration Law; RA 8560 — Amending RA 4374). This knowledge directly supports infrastructure projects, land registration, and precise positioning across the Philippine archipelago, where geoid undulation varies significantly (approximately −12 m to +5 m) due to regional gravity anomalies.
Summary
The geoid, gravity, and height systems form the foundation of modern geodetic practice in the Philippines and globally. Three fundamental surfaces must be understood: the ellipsoid (smooth, mathematical), the geoid (gravity-defined, undulating), and the terrain (physical). The critical relationship h = H + N connects ellipsoidal height (h, from GNSS), orthometric height (H, true engineering elevation), and geoid undulation (N, geoid position relative to ellipsoid). GNSS receivers provide h directly; applying a geoid model determines N; thus H = h − N converts GNSS heights to usable elevations. Gravity varies with latitude (~52 mGal from equator to poles), elevation (free-air gradient ~3 mGal per 100 m), and local mass (topographic anomalies ±20 mGal). The geoid is an equipotential surface: where gravity is strong, geoid rises (positive N); where gravity is weak, geoid sinks (negative N). In the Philippines, geoid undulation ranges approximately −12 m to +5 m due to regional crustal variations. Modern geoid models (Philippine Geoid Model, EGM2020) are derived from satellite gravity data, surface gravity measurements, and levelling networks, with typical accuracy ±0.5–1.0 m. For high-precision projects, local refinement using gravity surveys and levelling can reduce uncertainty to ±0.1–0.2 m. Common pitfalls include confusing h and H, mishandling the sign of N, using outdated models, and assuming N is constant over extended surveys. Best practices for Filipino engineers include always applying a geoid model to GNSS data, documenting the model version in all reports, verifying against levelled benchmarks, and using official models (PGM) for government work under RA 8560. Practical applications span cadastral surveying, water resources, building design, and flood risk mapping—all of which depend critically on correct height systems. Understanding these concepts and their proper application is essential for professional practice, regulatory compliance, and project success in the Philippines.
Sections
In geodetic work, we operate with three distinct surfaces that must be clearly understood: **The Ellipsoid (Mathematical Surface):** The ellipsoid is a mathematical reference surface that approximates Earth's shape. It is smooth, rotating, and defined by two parameters: semi-major axis (a) and flattening (f). For WGS84, a = 6,378,137.0 m and f = 1/298.257223563. The Philippine Reference System (PRS92) is aligned to WGS84 and is the official reference datum for the Philippines. The ellipsoid provides the geometric foundation for positioning but does NOT represent the true gravitational equipotential surface. **The Geoid (Gravity-Defined Surface):** The geoid is defined as the equipotential surface of Earth's gravity field that best approximates global mean sea level (MSL). Unlike the smooth ellipsoid, the geoid undulates (varies in height) due to lateral variations in Earth's internal mass distribution. Where dense rock formations exist (e.g., mountain ranges, crustal roots), the geoid rises relative to the ellipsoid (positive undulation N > 0). Where mass deficits exist (e.g., ocean trenches, sedimentary basins), the geoid dips below the ellipsoid (negative undulation N < 0). In Philippine waters, the geoid undulation ranges from approximately −12 m (south) to +5 m (north) due to regional gravity variations associated with subduction zones and crustal structure. The geoid is the reference surface for orthometric (levelled) heights — the heights that engineers use for design elevations, drainage calculations, and infrastructure planning. **The Terrain (Physical Surface):** This is the actual ground surface where we place monuments, buildings, and infrastructure. Heights measured from levelling instruments refer upward perpendicular to the geoid (and plumb lines at each station), while GNSS-derived heights are orthogonal to the ellipsoid. The terrain elevation relative to MSL is always needed for engineering work, not the ellipsoidal height. **The Critical Relationship:** The fundamental relationship governing these three surfaces is: $$h = H + N$$ Where: - **h** = ellipsoidal height (measured perpendicular to the ellipsoid; what GNSS receivers report directly in WGS84 coordinates) - **H** = orthometric height (measured perpendicular to the geoid along the direction of gravity; the true engineering elevation above MSL) - **N** = geoid undulation or geoid height (the height of the geoid above the ellipsoid; negative when the geoid lies below the ellipsoid) Rearranging: **H = h − N** This formula is the bridge connecting GNSS survey data to engineering elevations. Every GNSS receiver in the Philippines reports WGS84 coordinates with ellipsoidal height h, but practical engineering requires orthometric height H. Without a reliable geoid model (such as GEOID12B, GEOID18, or the Philippine Geoid Model), you cannot convert GNSS heights to usable elevations.
Heading
1. The Three Fundamental Surfaces in Geodesy
Examples
Example 1A: Converting GNSS Height to Orthometric Height
Problem
A GNSS survey in Quezon City, Metro Manila yields an ellipsoidal height h = 52.30 m (WGS84). The Philippine Geoid Model at that location gives N = −30.10 m. Calculate the orthometric height H (engineering elevation above MSL).
Solution
Using the fundamental relationship: H = h − N H = 52.30 − (−30.10) H = 52.30 + 30.10 H = 82.40 m above mean sea level Interpretation: Although the GNSS ellipsoidal height is only 52.30 m, the actual elevation above MSL is 82.40 m because the geoid in this region lies 30.10 m BELOW the WGS84 ellipsoid. This 30 m difference is due to a gravity deficit (negative gravity anomaly) associated with the basin structure of the Central Luzon Plain.
Example 1B: Determining Geoid Undulation from Levelling and GNSS
Problem
A geodetic benchmark in Mindanao has been precisely levelled to an orthometric height H = 112.20 m. A subsequent GNSS survey of the same point yields an ellipsoidal height h = 124.50 m (WGS84). Determine the local geoid undulation N.
Solution
Rearranging the height equation: N = h − H N = 124.50 − 112.20 N = +12.30 m Interpretation: The geoid at this Mindanao location is 12.30 m ABOVE the WGS84 ellipsoid. This positive undulation reflects a regional gravity excess (positive gravity anomaly) possibly associated with denser crustal material or an uplifted basement. Such points with accurately determined N values become ground-truth stations for geoid modeling and improve the national geoid model.
Example 1C: Why GNSS Height Alone Cannot Represent Elevation
Problem
A GNSS receiver at a coastal site reads h = 40 m, yet the site is known to be at or very near mean sea level. Without consulting a geoid model, explain why h ≠ H.
Solution
GNSS measures ellipsoidal height h perpendicular to the WGS84 reference ellipsoid. It does NOT measure height relative to the geoid (MSL). At this particular coastal location, suppose the geoid model shows N ≈ +40 m (the geoid is 40 m above the ellipsoid). Then: H = h − N = 40 − 40 = 0 m The point truly is at (or very close to) MSL, H ≈ 0 m. The apparent GNSS height of 40 m is an illusion created by the local geoid geometry. In regions where N is large and positive (like some island areas in the Philippines), GNSS heights can appear much higher than true elevations. Conversely, where N is large and negative (like some other regions), GNSS heights underestimate true elevations. A geoid model is absolutely essential for GNSS-to-elevation conversion.
Key Points
- The ellipsoid is a smooth mathematical surface; the geoid is gravity-defined and undulates due to mass variations.
- The fundamental height equation h = H + N links ellipsoidal, orthometric, and geoid heights.
- GNSS provides h (ellipsoidal); engineering requires H (orthometric); a geoid model bridges them via N.
- Negative N occurs where the geoid is below the ellipsoid; positive N where the geoid is above.
- PRS92/WGS84 is the Philippine geodetic reference frame; orthometric heights are referenced to mean sea level.
- Philippine geoid undulation ranges approximately −12 m to +5 m due to regional crustal and density variations.
Gravity is the fundamental force that defines the geoid and the levelling system used in engineering. Understanding gravity variations is essential for precise height systems and geoid computation. **What is Gravity?** Gravity g is the local acceleration due to Earth's gravitational field plus the centrifugal acceleration due to Earth's rotation. It is a vector quantity pointing downward (in the direction of the plumb line), with SI units of m/s² (or commonly expressed in mGal, where 1 Gal = 10⁻² m/s²). **Factors Affecting Gravity:** 1. **Latitude Effect (Primary):** Gravity varies significantly with latitude due to two factors: - Earth's oblate shape: The equatorial radius is ~21 km larger than the polar radius, placing equatorial points farther from Earth's center where gravity is weaker. - Centrifugal acceleration: Centrifugal acceleration is maximum at the equator and zero at the poles, further reducing apparent gravity at the equator. As a result: - At the equator: g ≈ 9.780 m/s² - At the poles: g ≈ 9.832 m/s² - Variation: ~0.052 m/s² or 52 mGal over Earth's surface The Philippines, lying between latitudes 5°N and 21°N, experiences gravity values approximately 9.785 to 9.810 m/s², with typical values around 9.79–9.80 m/s² for most survey work. 2. **Elevation Effect (Free-Air):** As elevation increases, distance to Earth's center increases, reducing gravity. The free-air gravity gradient is approximately −0.3086 mGal/m (or −3.086 mGal per 100 m of elevation). Example: A point at 1000 m elevation experiences gravity ~308.6 mGal less than a reference point at sea level (assuming no other local mass variations). 3. **Topographic/Bouguer Effect:** Local mass variations (mountains, valleys, dense rock formations, ore bodies) create gravity anomalies. A dense granite mountain produces positive gravity anomalies (+5 to +20 mGal locally); a deep sedimentary basin produces negative anomalies (−5 to −30 mGal). **The Gravity Field and the Geoid:** The geoid is, by definition, an equipotential surface of Earth's gravity field. At every point on the geoid, the gravitational potential W is constant. Where gravity is strong (positive anomaly), the geoid is pulled upward (N > 0); where gravity is weak (negative anomaly), the geoid sags downward (N < 0). The relationship is approximate but fundamental: $$\Delta N \approx -\frac{\Delta g}{\overline{g}}$$ where Δg is the gravity anomaly and ḡ is the normal gravity (approximately 9.81 m/s²). This shows that negative gravity anomalies produce negative geoid undulations — a key principle in geoid modelling. **Gravity Measurements and Geopotential Numbers:** For precise orthometric heights, spirit levelling alone is insufficient because plumb lines (directions of gravity) are not everywhere parallel — they converge toward the center of mass. The rigorous definition of orthometric height uses **geopotential numbers** (C), obtained by integrating levelling data with gravity observations: $$C = \int_0^H g \, dH$$ where the integral is taken along the plumb line from sea level to the point. The dynamic (or orthometric) height then is: $$H_{dyn} = \frac{C}{\overline{g}_{ref}}$$ where ḡ_ref is a reference gravity value (typically the normal gravity at 45° latitude: 9.80665 m/s²). This method is more rigorous than simple spirit levelling and is the standard for first-order levelling networks in developed countries, including the Philippine levelling network under PD 1529 (Geotechnical Engineering) and RA 8560. **Practical Implication for Philippine Engineers:** When using GNSS-derived elevations in the Philippines, the conversion H = h − N depends on both the ellipsoid definition (WGS84/PRS92) and the geoid model, which ultimately reflects the gravity field. For projects requiring high precision (±0.1 m or better), actual gravity observations at critical points improve the geoid estimate and reduce height conversion errors.
Heading
2. Gravity and Its Role in Defining the Geoid
Examples
Example 2A: Free-Air Gravity Correction
Problem
A gravity measurement at the city of Davao (latitude 7.1°N, elevation 15 m above sea level) reads g_obs = 9.7863 m/s². A reference gravity station at the same latitude but at sea level (elevation 0 m) was measured as g_ref = 9.7891 m/s². Verify the free-air correction and determine if the elevation effect is consistent with the standard gradient of −3.086 mGal/m.
Solution
The difference in measured gravity: Δg_obs = 9.7891 − 9.7863 = 0.0028 m/s² = 28 mGal Expected free-air correction for Δh = 15 m: Δg_FA = −3.086 mGal/m × 15 m = −46.29 mGal = −0.004629 m/s² Expected g_Davao = g_ref + Δg_FA = 9.7891 − 0.004629 = 9.7845 m/s² Observed g_Davao = 9.7863 m/s² Difference from expected = 9.7863 − 9.7845 = +0.0018 m/s² = +18 mGal Interpretation: The measured gravity is ~18 mGal higher than predicted by free-air correction alone. This residual likely reflects a local positive gravity anomaly (Bouguer anomaly) in the Davao region, possibly from denser crustal material. This anomaly directly translates to the local geoid undulation: using ΔN ≈ −Δg/ḡ, a +18 mGal anomaly produces approximately +1.8 m of geoid height change.
Example 2B: Geoid Undulation from Gravity Anomaly
Problem
A region in Luzon exhibits a regional gravity anomaly of Δg = −25 mGal due to a deep sedimentary basin. Using the approximate relationship ΔN ≈ −Δg/(9.81), estimate the change in geoid undulation associated with this anomaly.
Solution
Using the approximate relationship: ΔN ≈ −Δg / ḡ ΔN ≈ −(−25 mGal) / (9.81 m/s² converted to mGal) ΔN ≈ +25 mGal / 9810 mGal per m/s² ΔN ≈ +25 / 9810 m ΔN ≈ +0.00255 m ≈ +2.5 mm (rough estimate using linear approximation) Alternatively, using the more accurate relationship from geoid science (Stokes integral): ΔN ≈ −(0.01 m) × Δg(mGal) over a typical anomaly region ΔN ≈ −(0.01) × (−25) ≈ +0.25 m ≈ +25 cm Interpretation: A −25 mGal gravity deficit in a sedimentary basin causes the geoid to rise (N becomes more positive or less negative) by approximately 20–30 cm relative to the surrounding region. This demonstrates why geoid models must be based on dense gravity observations: local geology directly affects local geoid shape.
Example 2C: Latitude Effect on Gravity in Philippine Context
Problem
Two survey stations in the Philippines are at the same elevation (100 m) but at different latitudes: Station A in Luzon (latitude 15°N) and Station B in Mindanao (latitude 8°N). Calculate the approximate difference in normal gravity between them using the simplified international gravity formula.
Solution
The International Gravity Formula (IGF80) is: g = 9.780318 × (1 + 0.0053024 sin²φ − 0.0000058 sin²(2φ)) m/s² For Station A (φ = 15°): sin(15°) = 0.2588; sin²(15°) = 0.0670 sin(30°) = 0.5; sin²(30°) = 0.25 g_A = 9.780318 × (1 + 0.0053024 × 0.0670 − 0.0000058 × 0.25) g_A = 9.780318 × (1 + 0.000355 − 0.0000015) g_A ≈ 9.780318 × 1.0003535 g_A ≈ 9.78130 m/s² For Station B (φ = 8°): sin(8°) = 0.1392; sin²(8°) = 0.01938 sin(16°) = 0.2756; sin²(16°) = 0.0759 g_B = 9.780318 × (1 + 0.0053024 × 0.01938 − 0.0000058 × 0.0759) g_B = 9.780318 × (1 + 0.0001028 − 0.00000044) g_B ≈ 9.780318 × 1.0001028 g_B ≈ 9.78096 m/s² Difference: Δg = g_A − g_B = 9.78130 − 9.78096 = 0.00034 m/s² = 3.4 mGal Interpretation: Station A at 15°N (Luzon) has gravity ~3.4 mGal higher than Station B at 8°N (Mindanao), despite being at the same elevation. This latitude-dependent variation must be accounted for in precision gravity work and geoid computations. The variation is systematic and predictable using international formulas, but it significantly affects the geoid undulation across the Philippine archipelago.
Key Points
- Gravity g is the resultant of gravitational and centrifugal accelerations; it varies with latitude, elevation, and local mass.
- At Earth's equator: g ≈ 9.780 m/s²; at poles: g ≈ 9.832 m/s²; variation ~52 mGal globally.
- Free-air gravity gradient: −3.086 mGal per 100 m elevation; topographic anomalies: ±20 mGal locally.
- The geoid is an equipotential surface: where g increases (positive anomaly), geoid rises (N > 0); where g decreases (negative anomaly), geoid sinks (N < 0).
- Geopotential numbers C integrate levelling with gravity to give rigorous orthometric (dynamic) heights.
- Philippine levelling network and height system per PD 1529 and RA 8560 use gravity-referenced heights for design standards.
Three distinct height systems are used in geodetic and engineering practice, each with its own definition, measurement method, and application. **Ellipsoidal Height (h):** Ellipsoidal height is the geometric height of a point above the ellipsoid, measured perpendicular to the ellipsoid surface. - **Definition:** Distance from the ellipsoid surface to the point, along the ellipsoidal normal (perpendicular to the ellipsoid at the point). - **Measured by:** GNSS receivers directly, after processing to WGS84 or PRS92 coordinates. - **Range in Philippines:** Approximately −30 m (offshore) to +2800 m (mountaintops like Mount Apo). - **Why it exists:** The ellipsoid is a mathematical convenience; it provides a global, well-defined surface for coordinate systems (latitude, longitude, ellipsoidal height). - **Drawback for engineering:** Ellipsoidal height does NOT correspond to natural gravity equipotentials, and therefore has no direct physical meaning for water flow, drainage, or potential energy calculations. **Orthometric Height (H):** Orthometric height is the height above the geoid, measured along the direction of gravity (the plumb line). - **Definition:** The height of a point above the geoid surface, measured perpendicular to the geoid along the local plumb line (direction of local gravity). - **Physical meaning:** Orthometric height is directly related to gravitational potential and is the true "elevation above mean sea level" used in engineering. - **Measured by:** Spirit levelling (using optical levels, total stations, or digital levels to read vertical distances along the plumb line), combined with gravity observations to compute geopotential numbers. - **Symbol and range:** H (typically 0 to 3000+ m in the Philippines, with 0 corresponding to mean sea level). - **Applications in Philippine context:** - Water infrastructure: Irrigation channel design and water supply systems must account for orthometric heights to ensure gravity-driven flow. - Building codes (National Building Code of the Philippines): Foundation and structural design elevations are specified in orthometric heights. - Flood risk mapping: Flood elevation models use orthometric heights relative to MSL. - Roads and railways: Grades, slopes, and drainage are designed based on orthometric heights. - Property registration: Under RA 4374 (Cadastral Law) and PD 1529, property boundary descriptions may reference orthometric heights for land classification and risk zones. **Geoidal Height (N):** Geoidal height, or geoid undulation, is the height of the geoid above (or below) the ellipsoid, measured along the ellipsoidal normal. - **Definition:** The perpendicular distance from the ellipsoid surface to the geoid surface at a given point, positive when geoid is above the ellipsoid, negative when below. - **Symbol:** N (note: N can be positive or negative). - **Range in Philippines:** Approximately −12 m to +5 m. - Northern regions (Luzon north of ~13°N): N ≈ −8 to −5 m (geoid below ellipsoid) - Central Philippines: N ≈ −10 to −8 m - Southern regions (Mindanao): N ≈ −12 to −10 m - Some northern island groups: N ≈ 0 to +5 m - **Derived from:** Geoid models based on satellite gravity data, surface gravity measurements, and levelling networks. Current models for the Philippines include the Philippine Geoid Model (PGM, developed by NGA and NAMRIA) and international models like GEOID12B, GEOID18, and EGM2020. - **Critical role:** Without N, you cannot convert GNSS ellipsoidal heights (h) to orthometric heights (H). **Relationship Between the Three Heights:** $$h = H + N \quad \Rightarrow \quad H = h - N \quad \Rightarrow \quad N = h - H$$ All three relationships are equivalent and critical for geodetic work. **Orthometric vs. Dynamic Heights (Advanced Topic):** For extremely precise work (centimeter or millimeter level), a distinction exists between orthometric heights (H) and dynamic heights (H_dyn): $$H = H_{dyn} + \frac{C}{g_0(\phi)}$$ where C is the geopotential number and g₀(φ) is the normal gravity at the station's latitude. However, for typical engineering surveys in the Philippines, this distinction is negligible (< 1 cm for most applications), and orthometric and dynamic heights are used interchangeably. **Practical Workflow for Philippine Engineers (GNSS to Orthometric):** 1. Perform GNSS survey; obtain coordinates and ellipsoidal heights (h) in WGS84/PRS92. 2. Access the Philippine Geoid Model or an appropriate geoid model for the project location. 3. Look up or interpolate the geoid undulation (N) at the survey point. 4. Calculate orthometric height: H = h − N. 5. Use H for all engineering design, including drainage slopes, building foundations, and flood elevations. 6. Document the geoid model used, as future resurveys should use the same model for consistency within a project.
Heading
3. The Height Systems: Ellipsoidal, Orthometric, and Geoidal
Examples
Example 3A: Multi-Point GNSS to Orthometric Height Conversion (Network Survey)
Problem
A cadastral survey in Cavite, Philippines surveys four corners of a land parcel. GNSS processing yields ellipsoidal heights (h) in WGS84. The Philippine Geoid Model (PGM) provides the undulation (N) at each point. Calculate the orthometric heights (H) and verify the vertical geometry of the parcel. Data: - Point 1 (SW corner): h₁ = 45.23 m, N₁ = −28.56 m - Point 2 (SE corner): h₂ = 46.18 m, N₂ = −28.54 m - Point 3 (NE corner): h₃ = 48.67 m, N₃ = −28.48 m - Point 4 (NW corner): h₄ = 47.89 m, N₄ = −28.51 m
Solution
Apply H = h − N at each point: Point 1: H₁ = 45.23 − (−28.56) = 45.23 + 28.56 = 73.79 m Point 2: H₂ = 46.18 − (−28.54) = 46.18 + 28.54 = 74.72 m Point 3: H₃ = 48.67 − (−28.48) = 48.67 + 28.48 = 77.15 m Point 4: H₄ = 47.89 − (−28.51) = 47.89 + 28.51 = 76.40 m Orthometric Heights: - Point 1 (SW): H₁ = 73.79 m - Point 2 (SE): H₂ = 74.72 m - Point 3 (NE): H₃ = 77.15 m - Point 4 (NW): H₄ = 76.40 m Vertical geometry analysis: - South edge slope (P1 to P2): ΔH = 74.72 − 73.79 = 0.93 m over estimated horizontal distance (from xy coords) = slope percentage - East edge slope (P2 to P3): ΔH = 77.15 − 74.72 = 2.43 m (steeper) - North edge slope (P3 to P4): ΔH = 77.15 − 76.40 = 0.75 m (slight decline) - West edge slope (P4 to P1): ΔH = 76.40 − 73.79 = 2.61 m (steeper) Interpretation: The parcel has a moderate slope, with the NE corner (Point 3) at the highest elevation. For property registration per RA 4374, these orthometric heights define the vertical extent of the property. For drainage or building design, the slope data (orthometric heights) guides foundation levels and stormwater management.
Example 3B: Correcting a GNSS-Only Elevation Mistake
Problem
A municipal infrastructure project in Nueva Ecija references a critical water intake structure. A GNSS survey reports the intake location as h = 38.50 m elevation. A structural engineer designs a pump station intake at h = 38.50 m, expecting it to be near MSL. However, the project fails because the intake is actually 26 m BELOW the design elevation. The project geoid model shows N = −65.00 m at the location. Explain the error and calculate the true orthometric height.
Solution
The error: The engineer confused GNSS ellipsoidal height (h) with orthometric height (H). GNSS gives h = 38.50 m, but this is NOT the elevation above MSL. True orthometric height: H = h − N H = 38.50 − (−65.00) H = 38.50 + 65.00 H = 103.50 m above MSL Instead of h = 38.50 m, the TRUE elevation is H = 103.50 m. Error magnitude: Design error = Designed h − Correct H = 38.50 − 103.50 = −65.00 m The intake was designed 65 m too LOW because the geoid undulation was not applied. Lesson: The large negative N = −65.00 m indicates this region has a significant gravity deficit (perhaps a large sedimentary basin or crustal thinning). In such regions, GNSS heights can be dangerously misleading without a geoid model. This example illustrates why RA 4374 and engineering standards in the Philippines mandate the use of proper height systems (H) for infrastructure design, and why geodetic professionals must always apply a geoid model.
Example 3C: Checking Geoid Model Consistency Across a Survey Network
Problem
A survey of a provincial highway route includes three levelling benchmarks with known orthometric heights from the Philippine Levelling Network (under PD 1529). A new GNSS survey of the same benchmarks is conducted using WGS84. Verify that the geoid model used is internally consistent. Data (Legacy levelled vs. New GNSS): - Benchmark A: H_lev = 125.43 m, h_GNSS = 101.67 m → N = h − H = 101.67 − 125.43 = −23.76 m - Benchmark B: H_lev = 128.55 m, h_GNSS = 104.12 m → N = h − H = 104.12 − 128.55 = −24.43 m - Benchmark C: H_lev = 132.08 m, h_GNSS = 106.89 m → N = h − H = 106.89 − 132.08 = −25.19 m
Solution
Geoid undulation values derived from GNSS and levelling: - At Benchmark A: N_A = −23.76 m - At Benchmark B: N_B = −24.43 m (difference: ΔN_AB = −0.67 m over horizontal distance ~15 km) - At Benchmark C: N_C = −25.19 m (difference: ΔN_BC = −0.76 m over horizontal distance ~12 km) Geoid slope analysis: Slope A→B: −0.67 m / 15 km ≈ −0.45 mm/m (or −0.045%) Slope B→C: −0.76 m / 12 km ≈ −0.63 mm/m (or −0.063%) Consistency check: A typical geoid model variation of ~0.5 to 1.0 mm per km is expected due to regional gravity anomalies. The observed slopes (0.45–0.63 mm/m) are realistic. If the same geoid model (e.g., PGM or GEOID18) is used across all three points, the N values should follow the model's spatial variation smoothly. If N values from the model are: - N_A(model) ≈ −23.74 m (close to −23.76) ✓ - N_B(model) ≈ −24.41 m (close to −24.43) ✓ - N_C(model) ≈ −25.18 m (close to −25.19) ✓ Then the model is consistent, and GNSS heights can be reliably converted to orthometric heights using the model across the entire network. Any significant discrepancies would indicate either GNSS/levelling errors or an outdated/inconsistent geoid model, requiring investigation.
Key Points
- Ellipsoidal height (h) is geometric, measured perpendicular to the ellipsoid; obtained directly from GNSS.
- Orthometric height (H) is physical, measured along gravity direction; the true engineering elevation above MSL.
- Geoidal height (N) is the geoid's position relative to the ellipsoid; essential for h-to-H conversion.
- The fundamental relationship h = H + N must be memorized and applied correctly in all survey work.
- In the Philippines, N ranges approximately −12 m to +5 m; negative values dominate due to regional gravity deficit.
- Orthometric heights are used for engineering design (water flow, building codes, flood mapping); ellipsoidal heights are not.
- Geopotential numbers and dynamic heights are more rigorous for first-order networks but typically negligible for practical engineering.
- Sign convention: N < 0 means geoid is below ellipsoid (Δ = h − H is negative); N > 0 means geoid is above.
Modern geoid models are essential tools for converting GNSS heights to usable elevations. Understanding the basis and limitations of these models is crucial for Filipino geodetic professionals. **What is a Geoid Model?** A geoid model is a mathematical representation of the geoid surface, typically expressed as a grid of geoid undulation (N) values covering a geographic region. Modern geoid models are derived from: 1. **Satellite gravity data:** Missions like GRACE, GOCE, and CHAMP measure Earth's gravitational field at satellite altitude. These measurements are globally uniform and provide the long-wavelength (regional) geoid features. 2. **Surface gravity measurements:** Ground-based and ship-based gravity surveys provide high-frequency (local) details. In the Philippines, gravity data comes from: - National gravity campaigns by NAMRIA (National Mapping and Resource Information Authority) - Academic institutions (e.g., University of the Philippines, Manila Observatory) - International surveys (e.g., GEBCO bathymetric gravity, EMAG2 regional models) 3. **Levelling networks:** Classical spirit levelling over benchmarks provides an independent way to estimate N at specific points (N = h − H), which validates and improves the model. 4. **Computational methods:** Stokes integral, spherical harmonic models, and local fitting techniques combine these data sources into a continuous geoid surface. **Current Geoid Models Used in the Philippines:** 1. **Philippine Geoid Model (PGM):** - Developed jointly by the National Geospatial-Intelligence Agency (NGA) and NAMRIA. - Latest version: PGM18 (based on EGM2020) or similar; previous versions include PGM08. - Coverage: Entire Philippine archipelago. - Resolution: Typically 1×1 arc-minute grid (approximately 1.8 km × 1.2 km spacing). - Accuracy: ±0.5 to ±1.0 m depending on local gravity data coverage. - Access: Available from NAMRIA via web GIS portals or downloadable grids. 2. **EGM2020 (Earth Gravitational Model 2020):** - Global model maintained by the National Centers for Environmental Information (NOAA), US. - Based on satellite data (GRACE-FO, Sentinel-1), surface gravity, and altimetry. - Spatial resolution: Up to 2.5×2.5 arc-minutes globally. - Accuracy: ±0.2 to ±0.5 m in well-surveyed regions; poorer in sparsely surveyed areas. - Freely available; can be queried via online calculators. - Recommended for international projects or when local models are unavailable. 3. **GEOID12B and GEOID18 (US-specific, but sometimes referenced):** - These are primarily for US territory but are occasionally referenced as examples of high-quality models. - For Philippine use, PGM or EGM2020 is preferred. **Understanding Geoid Model Accuracy and Limitations:** - **Systematic error (bias):** All geoid models have a mean error relative to local levelling networks. For the Philippine Geoid, the systematic error is typically ±0.3 to ±0.7 m. This means all N values in a region might be off by a similar amount. - **Random error (scatter):** Local geoid height values can vary by ±0.5 to ±1.5 m from reality due to incomplete gravity data, unmodelled local anomalies, or interpolation artifacts. - **Data gaps:** Remote areas with sparse gravity measurements (e.g., some mountain ranges, deep offshore regions) have larger uncertainties. The Philippines, with its complex geology and limited offshore gravity coverage, has regional model uncertainties. - **Temporal changes:** The geoid is essentially static on human timescales, but the ellipsoid (WGS84) is updated periodically. Using an outdated geoid model with a current ellipsoid (or vice versa) can introduce systematic errors. **Improving Local Geoid Estimates:** For high-precision projects (±0.1 m or better), engineers can improve the geoid model by: 1. **Gravity surveys:** Conducting local gravity measurements at critical points and fitting a local geoid model. 2. **Levelling and GNSS collocation:** Precisely levelled benchmarks with coincident GNSS observations directly give local N values (N = h − H), which can be gridded to create a local refinement. 3. **Remove-restore technique:** Subtract the regional model from observations, compute local refinement, and then add the model back. This often improves accuracy to ±0.1 to ±0.2 m. **Selecting a Geoid Model for a Project:** For Filipino engineers, the recommended workflow is: 1. **Check project specifications:** Does the client or standard require a specific model (e.g., PGM for government projects)? 2. **Evaluate available models:** - For national/cadastral work: Use PGM (latest version). - For international or remote areas: Use EGM2020. - For very high-precision work: Combine multiple models or conduct local surveys. 3. **Document the choice:** Always record which geoid model was used. Future surveys must use the same model for consistency within the project area. Changing models mid-project introduces artificial discontinuities. 4. **Assess the impact:** For a ±1 m geoid model uncertainty, the resulting orthometric height uncertainty is also ±1 m. If the project requires better accuracy, either improve the model or acknowledge the limitation in the final report. **Case Study: Philippine Geoid Variation and Its Engineering Impact** Across the Philippines, the geoid undulation ranges from −12 m (southern Mindanao) to +5 m (northern regions). This 17 m variation has real consequences: - **Southern Mindanao (N ≈ −12 m):** GNSS heights must be reduced by ~12 m to get true elevation. An engineer who forgets this oversight will design a structure 12 m too low. - **Northern Luzon (N ≈ −5 to +3 m):** The variation is smaller but still significant for precision work. - **Open ocean regions:** Geoid models are less reliable offshore due to sparse gravity data. Using WGS84 ellipsoidal heights for hydrographic surveys requires extra caution. This spatial variation is why geoid-model selection is project-specific and location-dependent.
Heading
4. Geoid Modelling and Philippine Geoid Data
Examples
Example 4A: Geoid Model Selection for a Major Infrastructure Project
Problem
A large water resources project spanning three provinces (Nueva Ecija, Laguna, Batangas) requires GNSS positioning and elevation surveys to ±0.5 m accuracy. The client requires compliance with RA 8560 (National Geodetic Control Modernization Program) standards. Which geoid model should be used, and what are the implications for the survey?
Solution
Recommended approach: 1. **Model selection:** Use the latest Philippine Geoid Model (PGM, e.g., PGM18 based on EGM2020). - Rationale: PGM is the official national model for government infrastructure under RA 8560 and NAMRIA standards. - Accuracy: PGM typically provides ±0.5 to ±0.8 m in well-surveyed regions like Central Luzon. - Coverage: Seamless across the three provinces. 2. **Uncertainty budget:** - GNSS positional error: ±0.3 m (good RTK/PPP solution) - Geoid model error: ±0.5 to ±0.8 m (PGM typical) - Total orthometric height uncertainty: √(0.3² + 0.6²) ≈ ±0.67 m (root-sum-square) - This exceeds the ±0.5 m target. 3. **Refinement strategy:** - Conduct gravity measurements at 20–30 locations across the project area. - Obtain precise levelled heights (or collocate GNSS with existing benchmarks) at the same locations. - Compute local N values and compare to PGM predictions. - If discrepancies exceed ±0.3 m, develop a local geoid model (remove-restore technique). - This can reduce model uncertainty to ±0.2–0.3 m, meeting the ±0.5 m project requirement. 4. **Documentation:** - Document that PGM (version X) was used as the base model. - Report any local refinement applied. - Include a geoid model uncertainty statement in the survey report. - Provide GIS shapefiles or digital data showing the N field across the project area. 5. **Compliance:** This approach aligns with RA 8560 mandates for rigorous height determinations in national projects.
Example 4B: Detecting a Geoid Model Discontinuity at a Provincial Boundary
Problem
A cadastral survey in Quezon Province reveals an apparent jump in geoid undulation at the provincial boundary with Laguna Province. GNSS and levelling collocation shows: - Quezon Province (site A): h = 85.43 m, H_levelled = 112.50 m → N_A = h − H = −27.07 m - Laguna Province (site B, ~5 km away): h = 84.98 m, H_levelled = 110.38 m → N_B = h − H = −25.40 m - Apparent geoid jump: ΔN = N_A − N_B = −1.67 m over 5 km The PGM model at the same locations shows: - N_A(model) = −27.08 m ✓ - N_B(model) = −25.41 m ✓ Is this a real geoid feature or a data error?
Solution
Analysis: 1. **Model vs. observations:** The observed N values (−27.07 and −25.40) closely match PGM predictions (−27.08 and −25.41). Root-mean-square difference ≤ 0.01 m. This is excellent agreement, suggesting the geoid model is correct. 2. **Is the geoid jump real?** Yes. A 1.67 m geoid undulation change over 5 km (slope of ~0.33 mm/m) is realistic and typical of regional gravity anomalies. This region likely sits at the boundary between two geological provinces: - Positive gravity anomaly in Laguna (possibly denser basement rock, uplifted crustal block) → N is less negative (−25.40 m) - Negative gravity anomaly in Quezon (possibly sedimentary basin) → N is more negative (−27.07 m) 3. **Practical implication:** The cadastral survey should accept this geoid variation and use PGM (which captures it correctly). If the survey had used an older, coarser geoid model that smoothed over this feature, it would have introduced systematic errors (~0.8–1.0 m) on either side of the boundary. 4. **No action needed:** The PGM model is fit-for-purpose. The apparent discontinuity is a real geophysical feature, not a model error. Surveyors should document this in their report as a regional geoid gradient. Conclusion: This example shows that modern geoid models capture real geological/gravity features accurately, and apparent discontinuities often reflect true Earth structure rather than model failures.
Example 4C: Impact of Using the Wrong Geoid Model — A Cautionary Tale
Problem
A survey project in Mindanao in 2010 used an old geoid model (EGM96) which showed N ≈ −23 m in the region. A resurvey in 2024 uses EGM2020 which shows N ≈ −25.5 m at the same location. This 2.5 m difference has caused systematic discrepancies when comparing old and new surveys. Explain the source of error and recommend a remedy.
Solution
Source of error: 1. **Model evolution:** EGM96 (released 1996) was based on limited satellite data and sparse surface gravity in offshore areas. EGM2020 (released 2020) incorporates modern satellite data (GRACE-FO, Sentinel altimetry), expanded surface gravity databases, and improved computational methods. 2. **Mindanao's specific issue:** Mindanao's complex subduction zone and volcanic arc create complex gravity fields. EGM96 likely underestimated (or incorrectly placed) the negative anomaly associated with the Mindanao Trench, resulting in N values that were too high (less negative). - EGM96: N ≈ −23 m (underestimate of the negative anomaly) - EGM2020: N ≈ −25.5 m (more accurate, capturing deeper negative anomaly) - Difference: 2.5 m 3. **Systematic error in old elevations:** Any orthometric height computed using EGM96 is biased low by approximately 2.5 m: - Old H = h − N_EGM96 = h − (−23) = h + 23 (too low by 2.5 m) - New H = h − N_EGM2020 = h − (−25.5) = h + 25.5 (correct) 4. **Remedy:** - Do NOT simply add 2.5 m to old elevations; this assumes the ellipsoidal heights (h) were recorded and are still valid, which may not be true. - Instead, re-process the original GNSS data (if available) using modern datum (WGS84, latest epoch) and modern geoid model (EGM2020 or PGM). - If original GNSS data are not available, obtain new GNSS observations at a subset of old survey points, compare to old levelled elevations, and develop a systematic correction surface across the project area. - Document both models and the 2.5 m discrepancy in the project report. - For new work: Always use the latest approved geoid model (currently PGM or EGM2020). 5. **Prevention:** This example underscores the importance of: - Documenting the geoid model version in all survey reports (not just "used a geoid model"). - Reviewing and updating geoid models every 5–10 years as new satellite data becomes available. - Storing raw GNSS data (not just final coordinates) for future reprocessing with improved models. **Lesson for Filipino engineers under RA 8560 (National Geodetic Control Modernization):** The law mandates periodic updates to the national geodetic network and models. A 2.5 m discrepancy in major infrastructure could have serious consequences (e.g., flood risk miscalculation, pipeline misalignment). Staying current with geoid model updates is not optional—it is a professional and regulatory requirement.
Key Points
- A geoid model is a grid of N (geoid undulation) values derived from satellite gravity, surface gravity, levelling, and computation.
- Philippine Geoid Model (PGM), EGM2020, and other models offer ±0.5 to ±1.0 m accuracy; choice depends on project requirements.
- Geoid models can have systematic bias (±0.3–0.7 m) and random error (±0.5–1.5 m); uncertainty scales with local gravity data quality.
- The remove-restore technique and local gravity surveys can improve geoid estimates for high-precision projects.
- Philippine geoid undulation ranges ~−12 m to +5 m; southern regions have large negative N (geoid below ellipsoid).
- Always document which geoid model is used; changing models mid-project introduces discontinuities.
- Model selection: PGM for national work; EGM2020 for international; local refinement for ±0.1 m projects.
- Geoid model uncertainty directly translates to orthometric height uncertainty (ΔH ≈ ΔN).
Understanding the theory is essential, but applying it correctly in the field requires awareness of common pitfalls and best practices specific to the Philippine context. **Application 1: GNSS-Based Cadastral Surveys (RA 4374, RA 8560)** Under Philippine cadastral law (RA 4374 — Cadastral Act; updated by RA 8560), modern surveys increasingly use GNSS instead of classical traverse and triangulation. However, cadastral surveys must produce orthometric heights (H) for: - Property boundary elevation (for flood or risk mapping) - Foundation design levels - Estate documents and property records **Workflow:** 1. Conduct GNSS survey (static or RTK) in WGS84/PRS92. 2. Obtain ellipsoidal heights (h) from processing. 3. Apply the appropriate geoid model (PGM preferred) to get N. 4. Compute H = h − N for all points. 5. Include both h and H in survey reports, clearly labeled with datum and geoid model version. 6. Verify against existing benchmarks where possible. **Common pitfall:** Reporting only the GNSS ellipsoidal height (h) and labeling it as "elevation" or "MSL height." This is incorrect and can lead to construction errors. **Application 2: Water Resource and Hydraulic Engineering** Water infrastructure (dams, canals, pump stations, gravity water supply systems) is absolutely dependent on correct orthometric heights. Water flows downslope according to gravity, which is defined by the orthometric height system, not the ellipsoidal height. **Example scenario:** A gravity-fed irrigation canal in Nueva Ecija must slope downward from a dam to farmers' fields. If the engineer uses ellipsoidal heights without applying the geoid model, the computed slope could be reversed (uphill instead of downhill), causing total system failure. **Workflow for hydraulic design:** 1. GNSS survey all critical elevations (dam crest, canal bottom at intervals, outlet). 2. Apply geoid model to convert to orthometric heights. 3. Compute slopes, design channel dimensions, and verify drainage. 4. Always confirm that the slope is downhill (decreasing H in the flow direction). 5. Double-check by comparing GNSS heights to levelled benchmarks if available. **Application 3: Building Design and Foundation Work** Building codes (National Building Code of the Philippines) specify foundation design based on orthometric height above MSL, particularly for: - Flood risk zones (buildings in areas < 5 m above MSL in flood-prone regions) - Coastal construction (reference mean high water level, which is ~MSL) - Basement and underground utility placement **Workflow:** 1. Obtain site orthometric height (H) from GNSS + geoid model, verified against levelling if available. 2. Compare H to flood elevation maps (which are in orthometric coordinates). 3. Design foundation level accordingly (e.g., in a flood zone, elevation above predicted 100-year flood level). 4. Document the geoid model and survey method used. **Application 4: Flood Risk Mapping and Disaster Management** Flood elevation models for the Philippines (e.g., NOAH (Nationwide Operational Assessment of Hazards) project) use orthometric heights. GNSS-derived elevations for risk mapping must therefore use the same geoid model as the flood model, or results will be inconsistent. **Workflow:** 1. Identify which geoid model was used in the baseline flood map (often EGM96 or an older model). 2. If using new GNSS data, either: a. Convert to the same geoid model for consistency (if possible), or b. Document the difference and apply a systematic correction, or c. Regenerate flood maps using the new geoid model. 3. For critical infrastructure (hospitals, evacuation centers), verify elevations independently using levelling if the model discrepancy exceeds ±0.5 m. **Common Pitfall 1: Confusing h and H** **Scenario:** A GNSS survey reports h = 15 m. An engineer reads this as "the site is 15 m above sea level" and designs drainage accordingly. In reality, if N = −30 m at this location, the true elevation is H = 15 − (−30) = 45 m, a 30 m difference! **Prevention:** Always explicitly label heights as "ellipsoidal (h, WGS84)" or "orthometric (H, MSL)," and include the geoid model used in all reports. **Common Pitfall 2: Sign Errors with N** **Scenario:** An engineer looks up N = −28 m and computes H = h + N = 52 + (−28) = 24 m (incorrect subtraction). Correct: H = h − N = 52 − (−28) = 52 + 28 = 80 m. **Prevention:** Always write out the full formula H = h − N and verify the arithmetic. Negative N values require careful handling; practice the double-negative rule: h − (−N) = h + N. **Common Pitfall 3: Using Outdated or Incompatible Geoid Models** **Scenario:** A 1990s survey used EGM96; a 2020 GNSS survey uses EGM2020. Without realizing the model difference, an engineer assumes the old and new heights should be nearly identical. In reality, regional differences of 1–3 m are common due to model improvements. **Prevention:** Always document the geoid model and version. When comparing old and new surveys, either reprocess old data with the new model or conduct gravity surveys to quantify the model discrepancy. **Common Pitfall 4: Neglecting Geoid Variation Across a Project** **Scenario:** A pipeline survey across 50 km assumes N is constant. In reality, N varies from −26 m at one end to −28 m at the other (a 2 m change). If N is assumed constant at −26 m, then the far end is computed 2 m too low, potentially causing design errors. **Prevention:** For extended surveys (> 10 km), always interpolate N from a grid model at each survey point, rather than assuming a single value. **Best Practices for Philippine Geodetic Engineers:** 1. **Always use a geoid model:** Never report GNSS heights as elevations without applying N. 2. **Document everything:** - Ellipsoid (WGS84/PRS92) - Geoid model name and version (e.g., "Philippine Geoid Model 2018 based on EGM2020") - Survey method (GNSS, RTK, PPP, levelling) - Accuracy/uncertainty statement - Date of survey and model 3. **Verify against existing data:** Whenever possible, compare GNSS-derived heights to levelled benchmarks or previous surveys to detect systematic errors. 4. **Use official models:** For government and cadastral work under RA 8560, use NAMRIA-approved models (PGM) unless the client specifies otherwise. 5. **Account for model uncertainty in the design:** If the geoid model uncertainty is ±1.0 m and project tolerances are ±0.5 m, either improve the model or acknowledge the risk in the project report. 6. **Educate stakeholders:** Non-technical clients (architects, contractors, municipal officials) often misunderstand the difference between h and H. Always provide clear graphics and explanations in project reports. 7. **Stay current:** Geoid models are updated every 5–10 years. Subscribe to NAMRIA updates and periodically review whether your projects should use updated models.
Heading
5. Practical Applications and Common Pitfalls in Philippine Surveys
Examples
Example 5A: A Real-World Irrigation Project Error (Nueva Ecija Case Study)
Problem
A 2008 irrigation project in Nueva Ecija used GNSS to survey a canal alignment. The GNSS survey report showed h = 15 m at the water source (dam) and h = 10 m at the canal outlet (5 km away). The engineer designed the canal with a 1% slope, assuming gravity would drain water from h = 15 m to h = 10 m. However, after construction, water refused to flow; the canal either held water (levelled) or drained backward! What went wrong?
Solution
Root cause analysis: The engineer used ellipsoidal heights (h) for slope design without applying a geoid model. Let's assume the true geoid undulation at this location was N = −28 m (typical for Nueva Ecija). **True orthometric heights:** - Dam: H_dam = h − N = 15 − (−28) = 15 + 28 = 43 m - Outlet: H_outlet = 10 − (−28) = 10 + 28 = 38 m - True elevation drop: ΔH = 43 − 38 = 5 m over 5 km → downhill slope is correct **Engineer's incorrect assumption (using h only):** - Dam: h = 15 m (misidentified as H) - Outlet: h = 10 m (misidentified as H) - Calculated slope: (15 − 10) / 5 km = 1 m/5 km = 0.2% (matches design) **The real issue:** The engineer did not realize that both the dam and outlet have the SAME negative N value (~−28 m) because they are only 5 km apart in a region with uniform gravity field. So the true slope (5 m over 5 km = 0.2%) was actually correct. But let's consider a different scenario (which probably was the real cause): **Actual scenario (revised):** Suppose the GNSS data were misinterpreted and the h values at the two points were: - Dam: h = 40 m (measured correctly) - Outlet: h = 35 m (5 km away, but GNSS error or processing mistake gave h = 35 instead of true h = 40 − (elevation drop)) If the geoid undulation was N = −28 m at the dam but N = −25 m at the outlet (a 3 m geoid variation over 5 km due to regional gravity gradient), then: - H_dam = 40 − (−28) = 68 m - H_outlet = 35 − (−25) = 60 m (if h = 35 were correct) - True slope: (68 − 60) / 5 km = 1.6% → downhill, correct But if the engineer used the incorrect N = −28 m for both points: - H_dam = 40 − (−28) = 68 m - H_outlet = 35 − (−28) = 63 m (wrong!) - Computed slope: (68 − 63) / 5 km = 1%, which is correct by coincidence. **More likely real cause:** The project did not account for the geoid model AT ALL. GNSS ellipsoidal heights were treated as elevations (h ≈ H, which is false when N ≠ 0). The actual geoid variation across the project may have been incorrectly estimated or ignored, causing the computed slope to be wrong. **Corrective action (post-construction):** 1. Conduct precise levelling survey along the canal to measure true orthometric slope. 2. Measure gravity at multiple points to compute local N values and verify geoid variation. 3. If the true slope was negative (uphill), either: a. Install a pump station to lift water against gravity (expensive retrofit), or b. Redesign the canal with a revised route to follow the true downslope direction. 4. For all future projects, ALWAYS apply a geoid model and cross-check with levelling. **Prevention lessons:** - GNSS h ≠ H unless N is applied. - Gravity-fed systems (water, sewage, drainage) must use orthometric heights from the start of the design phase. - Verify GNSS-derived slopes with independent levelling or gravity data before construction. - Document the geoid model in the design report for future reference and audits.
Example 5B: Disaster Risk Assessment Mismatch (Metro Manila Flood Resilience)
Problem
The city of Manila commissioned a flood risk map in 2010 using EGM96 geoid model, with results showing that areas below 2 m orthometric height would experience > 10-year flood inundation. In 2020, a GNSS survey of the same area using EGM2020 revealed that the true height in some areas was ~0.8 m lower than the 2010 map indicated (due to geoid model differences). This raised concerns that many buildings previously classified as "low risk" were actually "high risk." How should this discrepancy be resolved?
Solution
Analysis of the geoid model discrepancy: 1. **Model evolution:** EGM96 and EGM2020 differ by ~0.8 m in some Metro Manila areas due to: - Improved satellite gravity data (GRACE, GOCE, Sentinel) - Better bathymetric/gravity coverage offshore - Refined computational methods - EGM96 had systematic bias in the Philippines due to sparse offshore data 2. **Impact on flood risk classification:** - 2010 map (EGM96): A site at h = 2.3 m (WGS84) was converted to H = 2.3 − (−20.1) = 22.4 m (orthometric). - 2020 survey (EGM2020): Same site at h = 2.3 m converts to H = 2.3 − (−20.9) = 23.2 m (orthometric). - Difference: ΔH = 0.8 m (site is 0.8 m higher than 2010 map indicated). - For a 2 m flood threshold, this difference is marginal; but for a site near the threshold (e.g., true H = 2.3 m), it could shift the classification from high-risk (H < 2 m) to borderline-safe (H > 2 m). 3. **Resolution strategy:** **Option A: Regenerate the flood map using EGM2020** - Advantages: Internally consistent, uses latest science, improves accuracy. - Disadvantages: Significant cost, all previous analyses become obsolete, public confusion. - Recommendation: Do this for new flood modeling; it is the long-term solution. **Option B: Apply a systematic correction to the 2010 map** - Compute the geoid model difference at grid points across Metro Manila. - Map out the spatial variation of ΔN (EGM2020 − EGM96). - Adjust all orthometric heights on the 2010 map by the appropriate ΔN. - Advantages: Preserves the original flood-inundation analysis; lower cost. - Disadvantages: Two-generation model correction introduces compounded uncertainty; not ideal for high-precision work. - Recommendation: Use for interim updates while planning the Option A regeneration. **Option C: Conduct independent verification at critical sites** - Select 20–30 locations in high-risk flood zones (near the threshold height). - Perform precise levelling (independent of geoid models) to get true orthometric height. - Verify which geoid model (EGM96 or EGM2020) is more accurate. - If EGM2020 is clearly superior, commit to regenerating the entire flood map; if EGM96 was actually more accurate locally, revise the 2020 interpretation. - Advantages: Provides ground truth; settles the debate. - Disadvantages: Time and cost for field levelling; may reveal both models are wrong and a local geoid refinement is needed. - Recommendation: Prioritize this for critical infrastructure locations (hospitals, fire stations, evacuation centers). 4. **Recommended action for Metro Manila:** - **Immediate (0–6 months):** Issue a technical note explaining the 0.8 m geoid model difference. Advise stakeholders that the 2010 map should be considered accurate to ±0.8 to ±1.0 m, and decisions should include a safety margin. - **Short-term (6–18 months):** Conduct Option C verification levelling at 30 critical sites. Use results to assess which model is more accurate locally. - **Medium-term (1–2 years):** Regenerate flood inundation map using EGM2020 and updated hydrodynamic models. Include updated topographic/bathymetric data collected in the interim. - **Documentation:** For all property-level flood risk assessments, require that both the geoid model and survey method be documented, so future updates can assess comparability. 5. **Lesson for Philippine disaster managers:** - Flood risk maps are only as accurate as the elevation data and geoid model they are based on. - Geoid models are updated periodically; risk maps should be reviewed every 5–10 years. - Where ±0.5 m accuracy is critical (flood thresholds, coastal building codes), independent levelling verification is needed, especially at locations near decision thresholds. - Always include uncertainty statements in public risk documents: "flood elevation ±1.0 m" guides design decisions better than a false sense of precision.
Example 5C: Verification Workflow for a Critical Infrastructure Survey
Problem
A major dam site in Mindanao requires GNSS surveys for the design of spillway elevation, which directly affects flood safety. The project engineer wants to convert GNSS data to orthometric heights with ±0.3 m accuracy, but the regional geoid model (EGM2020) has a stated uncertainty of ±0.8 m. Propose a verification workflow to reduce uncertainty to the required ±0.3 m level.
Solution
Proposed verification and refinement workflow: **Phase 1: Baseline GNSS + Geoid Model Survey (2–3 weeks)** 1. Conduct GNSS observations at 5 key locations around the dam site (dam crest, spillway, abutments, outlet). 2. Process GNSS data to WGS84/PRS92, obtaining ellipsoidal heights (h) with ±0.1 m accuracy. 3. Interpolate EGM2020 geoid undulation (N) at each point from a regional grid. 4. Compute preliminary orthometric heights: H = h − N. 5. Document: GNSS accuracy ±0.1 m, EGM2020 accuracy ±0.8 m → combined uncertainty √(0.1² + 0.8²) ≈ ±0.8 m (not sufficient). **Phase 2: Levelling Survey (3–4 weeks)** 1. Conduct precise spirit levelling (third-order or better) from a stable control point (e.g., nearby benchmark on the Philippine Levelling Network) to each of the 5 GNSS points. 2. Obtain levelled orthometric heights (H_lev) with ±0.1 m accuracy (achievable with proper technique). 3. This provides independent orthometric height, independent of any geoid model. **Phase 3: Geoid Model Validation (1 week)** 1. At each GNSS/levelled point, compute the actual geoid undulation: N_actual = h − H_lev. 2. Compare N_actual to N_EGM2020 (interpolated from the grid): - Point 1: N_actual = −28.2 m, N_EGM2020 = −27.8 m → residual = −0.4 m - Point 2: N_actual = −28.5 m, N_EGM2020 = −28.1 m → residual = −0.4 m - Point 3: N_actual = −29.0 m, N_EGM2020 = −28.4 m → residual = −0.6 m - Point 4: N_actual = −28.7 m, N_EGM2020 = −28.2 m → residual = −0.5 m - Point 5: N_actual = −28.3 m, N_EGM2020 = −27.9 m → residual = −0.4 m 3. Mean residual: −0.48 m (EGM2020 is systematically ~0.5 m too high, i.e., N is less negative than reality). 4. Standard deviation of residuals: ±0.08 m (good local consistency). **Phase 4: Local Geoid Refinement (1–2 weeks)** 1. Develop a local correction surface: N_refined = N_EGM2020 − 0.5 m (or more sophisticated if residuals show spatial gradient). 2. Uncertainty of refined model: ±0.1 m (approximately the standard deviation of residuals). 3. Test on independent points: If possible, check the refined N at any additional control points. **Phase 5: Final Design Heights (1 week)** 1. Use refined geoid model (N_refined) for all design calculations: - Spillway crest elevation: H = h − N_refined (with ±0.2 m uncertainty: √(0.1²+0.1²) from GNSS and refined model) - This meets the ±0.3 m project requirement. 2. Document the entire verification process in the survey report, including: - Levelling network closure and accuracy - Geoid model validation results - Local refinement applied - Final uncertainty statement **Phase 6: Quality Assurance (1 week)** 1. Have an independent surveyor verify: - GNSS processing and accuracy - Levelling notes and calculations - Geoid model selection and interpolation - Final height computations 2. Conduct a sensitivity analysis: If spillway crest H is 45.3 m and the design flood elevation is 45.5 m, then the margin is only 0.2 m. Given a ±0.3 m uncertainty, this is marginal; recommend either: - Increasing the spillway crest by 0.3 m for safety, or - Conducting gravity surveys to further refine the geoid model and reduce uncertainty below ±0.2 m. **Cost-Benefit Summary:** - Levelling survey: ~₱80,000–150,000 (3–4 weeks, team of 3–4 surveyors) - Geoid validation and refinement: ~₱30,000–50,000 (mostly analysis) - Total: ~₱130,000–200,000 - Benefit: Uncertainty reduced from ±0.8 m to ±0.2 m; in a critical dam project (value > ₱500 million), this investment is justified by improved design confidence and reduced risk. **Lessons for Philippine engineers on critical infrastructure:** - Never rely on a geoid model alone when ±0.3 m accuracy is required; always validate with independent levelling. - The remove-restore technique (local refinement) is a practical and cost-effective way to improve geoid estimates for a specific project area. - Document the verification process; it builds credibility for future audits and regulatory reviews. - Uncertainty statements are not signs of weakness; they demonstrate professional rigor and help stakeholders make informed decisions.
Key Points
- GNSS provides ellipsoidal height (h); always apply geoid model to get orthometric height (H) for engineering use.
- Water flows according to orthometric slope (ΔH), not ellipsoidal slope; designs based on wrong heights fail.
- Flood maps use orthometric heights; GNSS surveys for risk mapping must use consistent geoid models.
- Common pitfall: Confusing h and H, mishandling the negative sign in N, using outdated geoid models.
- For extended surveys (> 10 km), interpolate N at each point rather than assuming constant value.
- Document all parameters: ellipsoid, geoid model (name + version), survey date, method, uncertainty.
- Verify GNSS heights against levelled benchmarks when possible to detect systematic geoid model errors.
- Use official geoid models (PGM) for government work under RA 8560; cite model version in all reports.
- Geoid model uncertainty directly impacts design safety; include uncertainty statement in project documents.
- Non-technical stakeholders need clear explanation of h vs. H; use graphics and examples in reports.
Ready to practise for the GELE 2026?
Super Tutor's AI review plan adapts to your weak areas and builds a weekly practice schedule around your target GELE exam date.