Spatial Referencing and Positioning
Open in the Padhaiko app3.1 Shape of the earth: geoid, ellipsoid and datums
Accurate spatial referencing begins with modeling the true physical shape of the Earth. Because the Earth is neither a rigid sphere nor a smooth mathematical surface, geodesists define three nested surfaces: the topographic surface (actual terrain), the geoid, and the reference ellipsoid.
Geoid, Ellipsoid, and Vertical Height Relation
- Geoid: The equipotential surface of the Earth's gravity field that best fits global mean sea level (extended continuously beneath the continents). Because mass density varies inside the Earth, the geoid is gently undulating and serves as the physical reference for vertical datums (orthometric heights , i.e., elevation above Mean Sea Level).
- Reference Ellipsoid (Spheroid): A smooth, mathematically tractable oblate spheroid formed by rotating an ellipse around its shorter polar minor axis. It is parameterized by its semi-major equatorial radius and semi-minor polar radius , with flattening and first eccentricity squared :
- Geoid Undulation (): The vertical separation between the ellipsoid and the geoid at any location, linking ellipsoidal height (measured by GNSS) and orthometric height (surveyed above Mean Sea Level):
Horizontal and Vertical Datums
A datum ties a mathematical reference ellipsoid (or equipotential surface) to the physical Earth:
- Horizontal Datum: Defines the size, shape, origin, and orientation of the reference ellipsoid for horizontal coordinates .
- Local (Regional) Datum: Aligned to fit the geoid closely over one country or continent (e.g., Everest 1830 for South Asia and Nepal; , ), leaving larger offsets elsewhere.
- Global (Geocentric) Datum: Origin coincides with the Earth's center of mass, giving a global best fit (e.g., WGS 84 used by GPS; , ).
- Vertical Datum: Defines the zero-elevation surface () from which heights or depths are measured—typically local Mean Sea Level (MSL) realized through tidal benchmarks and geoid models (such as EGM2008).
| Aspect | WGS 84 Datum | Everest 1830 Datum (Nepal) |
|---|---|---|
| Datum Type | Global geocentric datum | Regional / local datum |
| Semi-Major Axis () | ||
| Inverse Flattening () | ||
| Primary Usage | Global GNSS positioning, Web GIS, international datasets | Traditional Survey Department mapping and MUTM of Nepal |
3.2 Coordinate systems and coordinate reference systems
A Coordinate System (CS) defines the axes, units, and angle/length conventions used to locate points, whereas a Coordinate Reference System (CRS) binds a coordinate system to the Earth through a specific datum (plus a map projection when planar).
- Geographic Coordinate System (GCS): Locates points directly on the 3D ellipsoidal surface using angular units (degrees or radians):
- Geodetic Latitude (): Angle between the equatorial plane and the ellipsoid normal at the point, ranging from (South Pole) to (North Pole).
- Geodetic Longitude (): Angle east or west of the Prime Meridian (Greenwich), ranging from to .
- DMS to Decimal Degrees: (with sign for S/W).
- Example:
EPSG:4326represents WGS 84 geographic coordinates.
- Projected Coordinate System (PCS): Transforms the curved 3D ellipsoid onto a flat 2D Cartesian plane using Easting ( or ) and Northing ( or ) in linear units (meters or feet). Essential for accurate planar distance, area, and engineering CAD/GIS calculations (e.g.,
EPSG:32645for WGS 84 / UTM Zone 45N). - Local Coordinate System: Uses an arbitrary local origin and local axis orientation in meters or feet without binding directly to a global Earth datum—common in site surveys, indoor floor plans, and CAD drafting before georeferencing.
3.3 Map projections, standard systems and MUTM of Nepal
A map projection is a systematic mathematical transformation from the curved ellipsoid onto a flat map plane. By Gauss's Theorema Egregium, no flat map can simultaneously preserve all geometric properties of a curved surface without distortion.
Classification and Distortion Properties
Map projections are classified by two criteria:
- By Developable Surface & Aspect:
- Planar (Azimuthal): Projects onto a flat tangent or secant plane (good for polar regions).
- Cylindrical: Projects onto a cylinder wrapped around the Earth (normal/equatorial, transverse, or oblique).
- Conical: Projects onto a cone coaxial with or inclined to the polar axis (good for mid-latitude east–west belts).
- Tangent vs. Secant: A tangent surface touches the globe along one point or line (zero distortion there); a secant surface cuts through the globe along two standard lines, distributing scale error more evenly.
- By Preserved Geometric Property:
- Conformal (Orthomorphic): Preserves local angles and shapes of small features ( at every point; Tissot's indicatrix remains a circle of varying size). Essential for navigation and topographic mapping.
- Equal-Area (Equivalent): Preserves true area ratios everywhere (), distorting shapes/angles away from standard lines. Essential for thematic density and land-cover area statistics.
- Equidistant: Preserves true scale along specific lines (e.g., along all meridians or radially from a center point).
- Azimuthal (True-Direction): Preserves true bearings from a chosen central point.
Standard Map Projections
| Projection | Class & Preserved Property | Key Parameters | Typical Use & Limitations |
|---|---|---|---|
| UTM (Universal Transverse Mercator) | Transverse cylindrical, secant, conformal | zones of longitude (–); ; False Easting ; False Northing (N) / (S) | Global topographic & engineering GIS (–); scale error within a zone. Nepal spans Zones 44N () and 45N () |
| Lambert Conformal Conic (LCC) | Conical, secant (2 standard parallels), conformal | Two standard parallels , central meridian | Mid-latitude countries or states with wide east–west extent; aeronautical charts |
Web Mercator (EPSG:3857) | Normal cylindrical (spherical radius ), pseudo-conformal | Bounded at (produces a square tile world) | Slippy web maps (OSM, Google Maps); massive polar area inflation (Greenland appears as large as Africa) |
| MUTM of Nepal (Modified UTM) | Transverse cylindrical, secant, conformal | Everest 1830 ellipsoid; -wide zones with Central Meridians ; ; False Easting ; False Northing | Official national cadastral & large-scale Survey Department mapping of Nepal; narrower zone keeps scale distortion |
1
A GNSS benchmark in Kathmandu is measured on WGS 84 at , , with ellipsoidal height . The local geoid undulation is .
- Convert to decimal degrees and determine the standard UTM zone and its central meridian .
- Compute the orthometric elevation above Mean Sea Level.
- Compute the semi-minor axis of the WGS 84 ellipsoid (, ).
Convert longitude to decimal degrees and find the UTM zone:
Compute orthometric height : Using :
(Notice that because the geoid lies below the WGS 84 ellipsoid in Kathmandu, .)
Compute the WGS 84 semi-minor axis : Since :
- , which lies in UTM Zone 45N with central meridian .
- Orthometric height .
- WGS 84 semi-minor axis .
3.4 Coordinate transformations and map scale
Combining GIS datasets from different surveys or scanned paper maps requires mathematical coordinate transformations:
- Datum Transformation: Converts coordinates between two 3D datums (e.g., Everest 1830 WGS 84) because their ellipsoid centers, orientations, and sizes differ:
- 3-Parameter Geocentric Shift: Translates ECEF origins by only (accuracy –).
- 7-Parameter Helmert Transformation: Applies translations , small axis rotations , and scale factor (sub-meter accuracy).
- Grid-Based Shift (NTv2): Interpolates empirical datum shifts over a precomputed latitude/longitude grid.
- 2D Affine Transformation (6 Parameters): Maps planar or pixel coordinates to target projected coordinates using translation, rotation, differential scaling, and non-orthogonal shear/skew while keeping parallel lines parallel:
Each Ground Control Point (GCP) provides equations, so at least non-collinear GCPs ( equations) are required to solve the unknowns (usually GCPs with least squares so Root Mean Square Error, , can be evaluated).
- Polynomial Transformations: Used when scanned paper maps have nonlinear shrinkage, folding creases, or lens distortion:
A polynomial of degree has coefficients per axis, requiring at least GCPs for -order and GCPs for -order warping.
- Map Scale:
An area on the map scales by : thus .
2
On a topographic map with Representative Fraction (RF) scale , a straight canal measures in length, and a rectangular reservoir covers on the map sheet.
- Find the true ground length of the canal in kilometers.
- Find the true ground area of the reservoir in square kilometers and hectares ().
Compute the true ground length: Here , so on the map equals on the ground:
Compute the true ground area: Area scales by :
Since and :
The canal has a ground length of (), and the reservoir has a ground area of ().
3.5 Satellite-based positioning (GNSS)
A Global Navigation Satellite System (GNSS) uses a constellation of Medium Earth Orbit (MEO) satellites continuously broadcasting timing signals and orbital ephemerides so any receiver with an unobstructed sky view can determine its 3D position and precise time.
Major GNSS Constellations
| System | Operator | Orbit Altitude & Inclination | Nominal Constellation | Reference Frame |
|---|---|---|---|---|
| GPS | United States | , | satellites ( orbital planes) | WGS 84 |
| GLONASS | Russia | , | satellites ( orbital planes) | PZ-90 |
| Galileo | European Union | , | – satellites ( planes) | GTRF (ITRF-aligned) |
| BeiDou (BDS-3) | China | MEO (+ GEO/IGSO) | satellites | CGCS2000 |
Working Principle: Pseudorange Trilateration
- Each satellite transmits a radio signal tagged with its exact transmission timestamp (from onboard atomic clocks) and its 3D coordinates (navigation ephemeris).
- The receiver measures arrival time using its inexpensive internal quartz clock (which has an unknown clock bias relative to GNSS system time).
- Multiplying travel time by the speed of light yields the pseudorange :
- Because there are four unknowns—receiver position and clock bias —the receiver must track at least satellites simultaneously to solve the system of nonlinear equations.
Sources of Error and Dilution of Precision (DOP)
- Error Sources:
- Ionospheric Delay: Charged particles in the upper atmosphere slow down carrier/code signals (mitigated by dual-frequency receivers).
- Tropospheric Delay: Water vapor and dry gases in the lower atmosphere bend/delay signals.
- Multipath Effect: Signals bouncing off buildings, cliffs, or water surfaces arrive late at the antenna (major problem in urban canyons and steep Himalayan valleys).
- Satellite Clock & Ephemeris Errors: Slight drift in satellite clocks or predicted orbit positions.
- Receiver Noise & Geometry: Internal tracking noise amplified by poor satellite sky distribution.
- Dilution of Precision (DOP): A dimensionless geometric multiplier quantifying how satellite sky geometry amplifies ranging error (User Equivalent Range Error, ):
- GDOP (Geometric): Combined 3D position + time; PDOP (Position 3D); HDOP (Horizontal ); VDOP (Vertical ); TDOP (Time).
- Mathematical identities:
- Well-spread satellites (one near zenith, three spaced apart near the horizon) yield low DOP (–, good); clustered satellites in one sky quadrant yield high DOP (, poor).
3
During a field survey, a single-frequency GNSS receiver reports , , , and a User Equivalent Range Error of .
- Calculate PDOP and GDOP.
- Estimate the expected 3D position error and horizontal error.
Calculate PDOP and GDOP:
Estimate 3D and horizontal positioning errors:
and . The expected 3D position error is and the horizontal error is .
added
Differentiate between the geoid, reference ellipsoid, and datum. How does WGS 84 differ from Everest 1830?
Geoid vs. Ellipsoid vs. Datum:
- The geoid is the physical equipotential gravity surface approximating global Mean Sea Level, used as the zero reference for orthometric heights ().
- The reference ellipsoid is a smooth mathematical oblate spheroid defined by semi-major axis and flattening .
- A datum anchors a reference ellipsoid (or vertical surface) to the Earth by specifying its origin, orientation, and scale.
WGS 84 vs. Everest 1830:
- WGS 84 is a global geocentric datum centered at the Earth's center of mass (, ), used globally by GPS and Web GIS.
- Everest 1830 is a local regional datum fitted specifically to South Asia (, ), historically used for Nepal's national cadastral and MUTM maps.
The geoid is the gravity equipotential surface for heights, the ellipsoid is a smooth mathematical approximation, and a datum ties the ellipsoid to the Earth. WGS 84 is a global geocentric datum (), whereas Everest 1830 is a local South Asian datum () used in Nepal's MUTM mapping.
Compare UTM, Modified UTM (MUTM) of Nepal, Lambert Conformal Conic (LCC), and Web Mercator in terms of zone/projection geometry, distortion properties, and primary use cases.
Analyze each projection:
- UTM: Secant transverse cylindrical conformal projection with zones (–), , false easting , WGS 84; standard for regional/national topographic GIS (Nepal falls in Zones 44N and 45N).
- MUTM (Nepal): Modifies UTM by narrowing zone width to (Central Meridians ), raising on Everest 1830, minimizing scale distortion for Nepal's cadastral surveying.
- LCC: Secant conical conformal projection with two standard parallels; minimizes distortion across mid-latitude regions stretched east–west.
- Web Mercator (
EPSG:3857): Spherical normal cylindrical projection clipped near latitude; ideal for uniform square web tiles, but severely exaggerates area at high latitudes.
UTM ( zones, ) and MUTM ( zones at , , Everest 1830) are conformal transverse cylindrical projections for topographic/cadastral accuracy; LCC is a secant conical conformal projection for east–west mid-latitude regions; Web Mercator is a spherical cylindrical projection optimized for web map tiling despite extreme polar area inflation.
A scanned map in pixel coordinates is georeferenced to a projected CRS in meters using the affine transformation:
- Compute the projected coordinates for a road intersection at pixel .
- How many minimum non-collinear Ground Control Points (GCPs) are needed to solve a -order affine vs. a -order polynomial transformation, and why?
Evaluate at :
Determine minimum GCP counts: Each GCP gives linear equations (one for , one for ):
- 1st-order affine: Has unknown parameters , requiring at least non-collinear GCPs ( equations).
- 2nd-order polynomial: Each axis has terms , totaling parameters, requiring at least well-distributed GCPs.
The intersection maps to . A -order affine transformation requires at least non-collinear GCPs ( parameters), whereas a -order polynomial requires at least GCPs ( parameters).
Why must a GNSS receiver track at least four satellites simultaneously to determine a 3D fix, and how do multipath and Dilution of Precision (DOP) degrade positioning accuracy?
Why four satellites are required: The pseudorange observation equation is:
There are four unknowns—receiver 3D coordinates and receiver clock offset . Thus at least independent satellite pseudorange equations are needed to solve uniquely for all four unknowns.
Impact of multipath and DOP:
- Multipath: GNSS radio signals reflect off nearby buildings, terrain walls, or water before reaching the antenna, lengthening the measured travel path and biasing the pseudorange upward.
- DOP (Dilution of Precision): Measures satellite sky geometry; when satellites are clustered closely together (narrow geometric intersection angles), small ranging errors () are magnified into large position uncertainties ().
At least 4 satellites are required to solve for the unknowns . Multipath delays signals via surface reflections, inflating pseudoranges, while high DOP (poor satellite sky spread) geometrically amplifies ranging errors into large position errors.
At a mountain survey station in Pokhara (), a GNSS receiver measures ellipsoidal height where geoid undulation is , with , , and .
- Determine the UTM zone and the Nepal MUTM Central Meridian for this station.
- Compute the orthometric elevation , PDOP, and expected 3D position error.
Determine UTM zone and MUTM Central Meridian:
For Nepal's MUTM, falls inside the central band (–), whose Central Meridian is .
Compute orthometric height, PDOP, and 3D error:
The station lies in UTM Zone 44N and MUTM CM . Its orthometric elevation is , , and expected 3D error is .