GNSS, Remote Sensing, Drones and Geospatial Systems

Introduction

Surveying has developed from observations made with optical instruments to integrated geospatial measurement using satellites, aircraft, unmanned aerial vehicles (UAVs), laser scanners and geographic information systems (GIS). These technologies do not eliminate the need for fundamental surveying principles. Instead, they extend the surveyor’s ability to establish positions, model terrain, collect imagery and manage spatial information.

This chapter explains the practical foundations of Global Navigation Satellite Systems (GNSS), static and real-time kinematic (RTK) surveying, satellite limitations and positioning errors. It also introduces UAV photogrammetry, LiDAR, GIS and digital terrain models (DTMs). Particular attention is given to accuracy, quality control, legal permissions, safety, data security and the appropriate selection of technology for a surveying task.

Learning objectives

By the end of this chapter, a student should be able to:

• distinguish GNSS, GPS and related satellite constellations;
• explain how satellite positioning determines coordinates;
• describe static, rapid-static, network-RTK and conventional RTK workflows;
• identify major sources of GNSS error and methods of mitigation;
• explain the principles and workflow of UAV photogrammetry;
• compare photogrammetry and LiDAR for terrain and infrastructure surveys;
• distinguish a DTM from a digital surface model (DSM);
• describe how GIS stores, analyses and presents spatial information;
• select an appropriate technology for a specified surveying purpose;
• apply accuracy, safety, permission and quality-control requirements professionally.

1. Geospatial positioning and reference systems

A geospatial measurement has two essential components: a position and a reference system. Position may be expressed as geographic latitude, longitude and ellipsoidal height, or as projected eastings, northings and a practical elevation datum. Coordinates without a stated datum, projection, epoch and units can be misleading or unusable.

GNSS is the collective term for satellite navigation systems. GPS is operated by the United States; other systems include Galileo (European Union), GLONASS (Russia) and BeiDou (China). A receiver may use several constellations simultaneously. Multi-constellation and multi-frequency observations generally improve availability and geometry, but they do not automatically guarantee accuracy.

The Earth is represented approximately by an ellipsoid for horizontal and ellipsoidal-height computations. Orthometric height, commonly called elevation above mean sea level, is related to ellipsoidal height by the geoid separation:

Here, H is orthometric height, h is ellipsoidal height, and N is geoid height; all are measured in metres and must refer to compatible datums. The relationship assumes the adopted geoid model and ellipsoid are consistent. A GNSS receiver directly determines ellipsoidal height more naturally than orthometric height, so a geoid model or local levelling connection is needed for engineering elevations.

A projected coordinate system converts curved-earth coordinates to a plane. The choice of projection, datum and zone affects coordinate values and distances. A draftsman must record these settings in drawings, survey reports and GIS metadata.

2. How GNSS positioning works

GNSS satellites transmit coded signals containing a precise time reference, satellite orbit information and system status. The receiver compares the transmission time with the reception time. Multiplying the apparent travel time by the speed of light gives a pseudorange. It is called a pseudorange because it includes receiver-clock error and propagation effects.

For a simplified receiver–satellite observation:

ρ_i is the measured pseudorange to satellite i in metres; X, Y and Z are unknown receiver coordinates; X_i, Y_i and Z_i are satellite coordinates; c is the speed of light in metres per second; δt is receiver-clock bias in seconds; and ε_i represents combined measurement and modelling errors. The satellite clock correction and other corrections are normally applied before or during computation. At least four suitably placed satellites are required to solve the three coordinates and receiver-clock bias, although more observations are normally used for redundancy and quality assessment.

The receiver solves a nonlinear set of equations by linearisation and least-squares estimation. Satellite geometry is important. If satellites are spread widely across the sky, the solution is generally stronger; if they are clustered, small measurement errors produce larger position errors. Dilution of precision (DOP) describes the effect of geometry. Common forms include GDOP for position and time, PDOP for three-dimensional position, HDOP for horizontal position and VDOP for vertical position. DOP is a geometry indicator, not an accuracy value by itself.

Carrier-phase observations measure the phase of the carrier wave. Their wavelength is much shorter than the code measurement, allowing millimetre- to centimetre-level relative precision after integer ambiguities are resolved. The unknown whole number of cycles between satellite and receiver is the integer ambiguity. A fixed solution has successfully resolved these ambiguities; a float solution has not fully resolved them and should not be treated as equivalent to a fixed solution.

Important GNSS terminology

3. Static and rapid-static GNSS surveying

In static surveying, two or more receivers observe simultaneously while remaining stationary over survey marks. One receiver may occupy a known control point and another an unknown point. The observations are processed as baselines, using differences between receivers and satellites to reduce common errors. Longer occupation times generally support longer baselines and more reliable ambiguity resolution, although the required time depends on equipment, satellite geometry, baseline length, ionospheric conditions and the required accuracy.

A typical static workflow is:

1. inspect the site and select stable, unobstructed stations;
2. verify marks, antenna type, antenna height and instrument settings;
3. configure the same datum, coordinate system, observation interval and satellite systems;
4. observe simultaneously for the planned period;
5. record station descriptions, antenna heights, weather and obstructions;
6. download raw data and precise orbit products where required;
7. process baselines and inspect residuals and ambiguity status;
8. adjust the network and compare independent or check baselines;
9. transform coordinates only after confirming the required datum and epoch;
10. archive raw data, processing reports and final coordinates.

Rapid-static surveying uses shorter occupations, often on relatively short baselines under favourable conditions. It is useful for control densification but is more sensitive to poor geometry, multipath and cycle slips. It must not be selected merely because it is faster; the occupation duration must be justified by the required accuracy and field conditions.

A closed or redundant network is preferable to isolated observations. Independent occupations at different times can reveal blunders and environmental effects. A computed coordinate with many displayed decimal places is not necessarily accurate.

4. RTK and network RTK

Real-time kinematic surveying uses carrier-phase observations and corrections transmitted from a base station or reference-station network to a rover. The rover computes a relative position while operating in the field. Conventional RTK uses a local base. Network RTK uses corrections generated from several permanent reference stations and delivered through a communication service, often via cellular internet.

A practical RTK workflow includes:

• confirming the project coordinate system, geoid model and datum transformation;
• checking radio, cellular or internet connectivity;
• setting the base over a known point or initializing a properly surveyed base;
• entering the antenna height using the correct measurement method;
• allowing the rover to track adequate satellites and monitoring solution status;
• waiting for a fixed solution rather than accepting an unverified float solution;
• occupying each point for an appropriate time and recording quality indicators;
• measuring independent check points and repeating critical points;
• avoiding obstructions, reflective surfaces and radio interference;
• storing point codes, descriptions, photographs and field notes.

RTK is efficient for topographic detail, construction set-out and control densification. It may fail or degrade beneath trees, beside buildings, near vehicles or fences, in deep cuttings and during ionospheric disturbance. A fixed status can be lost after a cycle slip. The operator should reinitialise and recheck the point rather than continuing under an assumed solution.

A base coordinate error propagates into rover coordinates. If a local base is assigned an incorrect coordinate, the survey can be internally consistent but displaced from the true datum. Known control, independent checks and appropriate transformations are therefore essential.

Figure: Simplified RTK observation, correction and quality-control workflow.

5. GNSS errors and limitations

GNSS accuracy is controlled by the measurement environment, satellite geometry, atmospheric conditions, equipment, processing method and reference data. Major error sources include:

• Satellite orbit and clock errors: imperfect satellite coordinates or time corrections;
• Ionospheric delay: frequency-dependent delay caused by the ionosphere;
• Tropospheric delay: delay caused by the neutral atmosphere, especially pressure, temperature and water vapour;
• Multipath: reflected signals arriving with direct signals from buildings, water, vehicles or rock faces;
• Receiver noise: electronic and signal-tracking limitations;
• Cycle slips: interruptions in carrier tracking;
• Poor geometry: satellites concentrated in one part of the sky;
• Obstructions and interference: trees, structures, radio sources and deliberate or accidental jamming;
• Datum and transformation errors: incorrect reference frame, epoch, geoid or projection;
• Operator errors: wrong antenna height, point identification, units or coordinate settings.

Differential methods reduce errors common to receivers observing the same satellites at similar times. They do not remove local multipath, incorrect antenna heights or a wrong base coordinate. Dual-frequency observations assist ionospheric modelling and ambiguity resolution, but difficult sites can still produce unreliable results.

Quality indicators should be interpreted together: solution type, number of satellites, PDOP, estimated precision, residuals, baseline length, age of corrections and independent checks. A low numerical estimated precision is not proof of correctness if the model is wrong or the point was incorrectly occupied.

6. UAV photogrammetry

UAV photogrammetry derives measurements and three-dimensional products from overlapping photographs taken by a drone. The process depends on image geometry, camera calibration, image overlap, identifiable ground control and suitable processing software.

A normal workflow is:

1. define the survey boundary, accuracy requirement and deliverables;
2. obtain aviation, landowner and other required permissions;
3. plan flight height, image overlap, speed, lighting and safe take-off and landing areas;
4. establish and survey ground control points (GCPs) and independent checkpoints;
5. fly the mission while monitoring weather, battery, aircraft and airspace;
6. inspect images for blur, exposure problems, gaps and inadequate overlap;
7. align images using common features and estimate camera positions;
8. introduce GCPs and process the point cloud, mesh, orthomosaic and elevation products;
9. use checkpoints to estimate accuracy independently of the adjustment;
10. classify, edit, export and archive the data with metadata.

Image overlap allows the same ground features to be observed from several positions. Forward overlap is overlap between successive images along a flight line; side overlap is overlap between adjacent flight lines. High-relief terrain, vegetation, wind and repetitive surfaces may require more overlap than a simple flat site.

GCPs constrain the model to a known coordinate system. Checkpoints are not used to control the solution; they test it. A survey that reports only software internal accuracy, without independent checkpoints, provides weaker evidence of real-world accuracy.

Photogrammetry performs best when the ground has texture and lighting is adequate. Uniform surfaces, deep shadows, moving vegetation, water, glass and highly reflective materials can produce gaps or mismatches. The camera should be calibrated or its self-calibration assessed. The final ground sample distance (GSD) is not the same as coordinate accuracy; a small pixel size does not guarantee an accurate model.

7. LiDAR and sensor selection

LiDAR (Light Detection and Ranging) measures distance by transmitting laser pulses and recording return time. A simplified range relationship is:

R is the sensor-to-target range in metres, c is the speed of light in metres per second, and ∆t is the round-trip travel time in seconds. Division by two is required because the pulse travels to the target and back. Actual systems account for scanning geometry, sensor position and orientation, timing, atmospheric effects and multiple returns.

Airborne or UAV LiDAR can penetrate some vegetation gaps and may provide useful ground points beneath a canopy. It is often more reliable than image matching on low-texture or shaded surfaces, but equipment, processing and calibration are more demanding. Water, highly reflective or absorptive materials and dense vegetation can still cause missing or unreliable returns.

The correct selection depends on required accuracy, terrain, vegetation, deliverables, budget, permissions, weather and operator competence.

8. Digital terrain models and elevation products

A digital elevation model (DEM) is a general term for digital representation of elevation. Terminology varies among organisations, but a common distinction is:

• DSM: elevation of the visible surface, including buildings, trees and other objects;
• DTM: terrain elevation after removal or classification of above-ground objects, often including breaklines;
• Digital surface or terrain grid: a raster or triangulated representation of elevation.

A point cloud may be converted to a raster grid or triangulated irregular network (TIN). A TIN represents terrain using connected triangular facets and can preserve breaklines efficiently. A raster uses regular cells and is convenient for GIS analysis, but its resolution and interpolation method affect results.

A useful raster cell-size relation is:

A is the represented area in square metres, n is the number of cells, and s is cell size in metres. This relation describes coverage, not accuracy. Interpolation across a cliff, drainage line or sparse point area may be inappropriate unless breaklines and survey observations are included.

DTMs support contour generation, slope and aspect analysis, drainage studies, cut-and-fill estimation, road design and flood modelling. Contours are derived representations; they should be checked against spot heights, breaklines and known control. Vegetation classification errors can make a DSM appear to be a DTM.

9. GIS and geospatial data management

A GIS links spatial features with descriptive attributes. Vector data represent points, lines and polygons; raster data represent values in cells. A road may be a line with attributes such as name, width and surface type. A parcel may be a polygon linked to ownership or land-use data. A satellite image or DTM is normally raster data.

GIS operations include coordinate transformation, buffering, overlay, spatial query, terrain analysis, network analysis and map composition. Overlay results are meaningful only when layers use compatible coordinate reference systems, scales, dates and accuracy standards.

A professional geospatial dataset should include metadata stating:

• source, survey date and responsible organisation;
• coordinate reference system, datum, projection and epoch;
• vertical datum and geoid model where elevations are used;
• instrument, sensor, flight or observation method;
• control points, checkpoints and accuracy assessment;
• processing software, significant settings and classification rules;
• limitations, completeness, resolution and intended use.

GIS should not be treated as a substitute for field verification. A visually attractive map can contain incorrect coordinates, outdated imagery, topology errors or false precision.

10. Accuracy, specifications and quality assurance

The surveyor must begin with the required purpose and accuracy, not with a preferred instrument. A reconnaissance should identify control availability, obstructions, terrain, access, safety hazards, airspace restrictions, privacy concerns and expected deliverables.

Accuracy specifications should distinguish absolute accuracy from relative precision. Absolute accuracy describes agreement with an external datum or control network. Relative precision describes consistency between nearby points. A model may be internally smooth yet shifted from the correct datum.

A practical quality-assurance plan includes:

• independent check measurements;
• redundant observations and repeat occupations;
• calibration and equipment checks;
• closure or residual analysis where applicable;
• documented rejection and editing criteria;
• comparison with known benchmarks or control;
• clear reporting of uncertainty and limitations;
• preservation of raw observations and processing records.

For a set of residuals e_i, a commonly used root-mean-square error is:

Here, e_i is the error or residual for checkpoint i and n is the number of checkpoints. RMSE has the same units as e_i. It summarises the magnitude of errors but does not by itself reveal systematic bias, outliers, spatial patterns or whether the checkpoints represent the full survey area. It should therefore be accompanied by maximum error, sample distribution and the stated accuracy standard.

11. Permissions, safety, ethics and governance

UAV operations must comply with the aviation authority’s rules, airspace restrictions, pilot competency requirements, aircraft registration and operational limitations applicable in the country of work. Permission from landowners or site managers may also be required. Flights near airports, people, roads, critical infrastructure or restricted facilities require particular caution and may be prohibited without authorization.

A safe operation includes a documented risk assessment, weather check, battery and aircraft inspection, emergency procedure, exclusion zone, visual observation where required, and protection against loss of control. GNSS equipment also requires care around traffic, excavations, water, high-voltage installations and unstable ground.

Imagery may reveal homes, people, vehicles, industrial processes or sensitive infrastructure. Data collection should follow privacy, data-protection, confidentiality and cybersecurity requirements. Access controls, encryption, secure transfer and appropriate retention periods are good practice. A surveyor must not remove inconvenient observations merely to meet a target accuracy or alter a point cloud without recording the edit.

AI-assisted classification and automated feature extraction can accelerate point-cloud and image processing. However, automated outputs may misclassify vegetation, vehicles, shadows or structures. Human review, traceable processing and independent validation remain necessary, especially for legal boundaries, safety-critical infrastructure and construction quantities.

12. Illustrative case: UAV and RTK survey for a road-widening project

Background

A municipal authority requires a topographic base map for widening a two-kilometre rural road. The deliverables include an orthomosaic, terrain model, drainage features, utility indications, cross-sections and a design control network.

Problem

Dense roadside vegetation obstructs some GNSS observations. The project corridor includes private land, overhead power lines and a nearby controlled airspace zone. The design team requires a consistent local coordinate system and reliable elevations.

Analysis

RTK alone can efficiently collect open-ground points but cannot provide complete visual coverage beneath vegetation. UAV photogrammetry can map visible surfaces and road details, while selected ground observations and breaklines are needed to construct a suitable terrain model. A local datum transformation and geoid model must be agreed with the authority before fieldwork.

Approach and solution

The survey team first obtains land access and aviation permissions and prepares a risk assessment. Static GNSS observations establish several control stations tied to the authority’s known control. RTK is then used to survey road edges, drainage inlets, culvert locations, utility markers, breaklines and GCPs. Independent checkpoints are distributed along the corridor.

The UAV flight is planned with adequate overlap and launched only after checking airspace, weather and power-line hazards. Images are processed into an orthomosaic and dense point cloud. Vegetation and non-ground points are classified and removed cautiously. The DTM incorporates retained ground points and surveyed breaklines. Checkpoint residuals are analysed using horizontal and vertical components rather than relying only on software-reported internal precision.

Outcome

The authority receives a GIS-ready dataset, CAD linework, a documented DTM, orthomosaic, control report and metadata. Areas where vegetation prevents confident terrain extraction are marked for supplementary ground survey rather than being presented as equally reliable.

Learning points

The strongest result comes from combining methods rather than expecting one sensor to solve every problem. Control, checkpoints, permission, metadata and transparent limitations are as important as image resolution or receiver specifications.

Contemporary developments and professional practice

Modern receivers increasingly combine multiple constellations, frequencies, inertial sensors and cloud correction services. UAV platforms may carry multispectral, thermal or LiDAR sensors. Cloud processing improves collaboration but introduces dependence on connectivity, licensing and data governance. Digital twins combine surveyed geometry with operational information for infrastructure management.

These developments increase productivity but also increase the consequences of poor specification and undocumented processing. A diploma-level professional should be able to ask: What is being measured? In which datum? With what expected accuracy? How will it be checked? Is the operation legal and safe? Can another professional reproduce the result?

Chapter Summary

GNSS determines position from satellite signals, with receiver coordinates and clock bias estimated from pseudorange and carrier-phase observations. Static surveying provides strong relative control through simultaneous observations and baseline processing, while RTK provides efficient real-time positioning using corrections from a base or reference network. Satellite geometry, atmospheric delay, multipath, cycle slips, datum errors and operator mistakes can all limit results.

UAV photogrammetry creates maps and three-dimensional products from overlapping images, but it requires careful flight planning, control points, independent checkpoints, camera and image-quality assessment, and legal authorization. LiDAR measures range using laser pulses and is valuable where vegetation or low image texture limits photogrammetry. DSMs represent visible surfaces, whereas DTMs aim to represent bare terrain.

GIS integrates spatial features, attributes, coordinate systems and analysis. Professional outputs require metadata, appropriate projections, quality control, accuracy reporting and preservation of raw data. Technology selection must be based on project purpose, terrain, required accuracy, safety, permissions, cost and deliverables. Automated and AI-assisted processing can improve efficiency but must be reviewed and validated by a competent surveyor.

Key Terms

GNSS: Collective term for satellite navigation systems.

GPS: The United States satellite navigation system.

Pseudorange: Apparent satellite-to-receiver range derived from signal travel time and affected by clock and propagation errors.

Carrier phase: Measurement of the phase of a GNSS carrier signal used for high-precision positioning.

Integer ambiguity: Unknown whole number of carrier cycles between satellite and receiver.

Static surveying: GNSS surveying in which receivers remain stationary while observing simultaneously.

RTK: Real-time kinematic positioning using carrier-phase corrections.

CORS: Continuously Operating Reference Station.

Multipath: Error caused by reflected signals arriving with direct signals.

DOP: Indicator of the effect of satellite geometry on positioning strength.

Photogrammetry: Measurement and three-dimensional reconstruction from photographs.

UAV: Unmanned aerial vehicle used for remotely operated data collection.

GCP: Ground control point used to constrain a spatial model.

Checkpoint: Independent point used to assess model accuracy.

LiDAR: Light Detection and Ranging using laser pulses.

Point cloud: Collection of three-dimensional points representing measured surfaces.

DSM: Digital surface model including visible objects such as buildings and vegetation.

DTM: Digital terrain model representing classified or modelled bare ground.

GIS: Geographic information system for storing, analysing and presenting spatial data.

Geoid: Gravity-related reference surface used to relate ellipsoidal and orthometric heights.

Review Questions

Short Answer Questions

1. What is the difference between GNSS and GPS?
2. Why is a GNSS carrier-phase solution described as “fixed” or “float”?
3. State two important sources of GNSS error.
4. What is the purpose of an independent checkpoint in UAV photogrammetry?
5. Distinguish a DSM from a DTM.

Long Answer Questions

1. Explain the operating principle and field workflow of static GNSS surveying.
2. Describe an RTK survey from base setup through quality control and coordinate delivery.
3. Explain how satellite geometry, atmospheric effects and multipath influence GNSS accuracy.
4. Describe the complete workflow for producing an orthomosaic and DTM from UAV imagery.
5. Compare UAV photogrammetry and LiDAR for a vegetated engineering corridor.

Analytical / Application Questions

1. A rover repeatedly loses its fixed solution beside a high concrete building. Diagnose the likely causes and propose a field procedure for obtaining reliable points.
2. A UAV project has excellent software-reported internal precision but large checkpoint errors in one corner of the site. Analyse possible causes and recommend corrective action.
3. Select a combination of GNSS, UAV, LiDAR and GIS methods for a road-widening project in dense vegetation. Justify the selection in terms of accuracy, safety, permissions, cost and deliverables.

Multiple Choice Questions with Answers

Question 1

Which observation is most directly associated with centimetre-level relative GNSS positioning after ambiguity resolution?

A. Carrier-phase observation

B. Uncorrected pseudorange only

C. Photogrammetric pixel value

D. Barometric pressure observation

Correct Answer: A. Carrier-phase observation

Question 2

In the relationship H = h − N, what does H represent?

A. Satellite clock bias

B. Orthometric height

C. Horizontal dilution of precision

D. Antenna phase-centre offset

Correct Answer: B. Orthometric height

Question 3

What is the principal purpose of a GNSS base station in conventional RTK?

A. To photograph the survey area

B. To provide reference observations and corrections to the rover

C. To classify vegetation in a point cloud

D. To replace all independent checks

Correct Answer: B. To provide reference observations and corrections to the rover

Question 4

Which condition is most likely to cause GNSS multipath?

A. An unobstructed open field

B. A reflective building façade near the antenna

C. A correctly measured antenna height

D. A well-distributed satellite geometry

Correct Answer: B. A reflective building façade near the antenna

Question 5

What is the correct reason for the factor 2 in R = c∆t ÷ 2?

A. The laser pulse travels to the target and back

B. The sensor uses two coordinate systems

C. Two satellites are always required

D. The point cloud has two dimensions

Correct Answer: A. The laser pulse travels to the target and back

Question 6

Which point is deliberately excluded from controlling a photogrammetric adjustment so that it can test accuracy?

A. Ground control point

B. Check point

C. Base station

D. Breakline

Correct Answer: B. Check point

Question 7

Which product normally includes trees and buildings?

A. Bare-earth DTM

B. Orthometric datum

C. DSM

D. Static baseline

Correct Answer: C. DSM

Question 8

Which factor is especially important for successful image matching in UAV photogrammetry?

A. Adequate image overlap and identifiable surface texture

B. Zero metadata

C. Removing all control points

D. Using only one photograph

Correct Answer: A. Adequate image overlap and identifiable surface texture

Question 9

What does a low PDOP primarily indicate?

A. Favourable satellite geometry

B. Correct camera calibration

C. A completed aviation permission

D. A classified bare-earth model

Correct Answer: A. Favourable satellite geometry

Question 10

Why should GIS metadata state the coordinate reference system and datum?

A. To increase image brightness

B. To ensure coordinates can be correctly interpreted and combined

C. To guarantee that every point is error-free

D. To eliminate the need for field checks

Correct Answer: B. To ensure coordinates can be correctly interpreted and combined