What is LiDAR? A Complete Guide for 2026

LiDAR (Light Detection and Ranging) is a remote sensing technology that uses rapid laser pulses to measure the distance between a sensor and the surfaces around it. Each pulse reflects off the ground, buildings, vegetation, or other objects, and the sensor records the time it takes for the light to return. By firing millions of these pulses per second, a LiDAR system builds a dense three-dimensional representation of the environment called a point cloud. It is the foundational technology behind modern topographic surveys, autonomous vehicles, forestry inventories, and countless other applications where accurate 3D measurement matters.

I’ve been working with LiDAR data for years — from airborne corridor surveys to UAV-mounted systems on construction sites. This guide covers everything you need to know about the technology in 2026: how it works, the different types, where it’s used, the file formats involved, and how it compares to photogrammetry.

How LiDAR works

The basic principle: time of flight

LiDAR operates on a simple concept. A laser emits a short pulse of light (typically near-infrared at 905 nm or 1550 nm wavelength). That pulse travels at the speed of light, hits a surface, and bounces back to the sensor’s detector. The sensor measures the round-trip time with extreme precision.

The formula is straightforward:

Distance = (Speed of Light x Time) / 2

Because light travels at approximately 299,792,458 metres per second, even nanosecond timing differences translate into centimetre-level distance measurements.

Scanning mechanisms

A single laser firing straight down would only measure one point. To build a full 3D picture, LiDAR systems use scanning mechanisms to sweep the laser across the landscape:

• Oscillating mirrors — The most common approach in airborne systems. A mirror rocks back and forth, directing the laser in a zigzag pattern across the ground as the aircraft moves forward.
• Rotating prisms or polygons — Used in some terrestrial and mobile systems. A spinning prism redirects the laser in a continuous circular sweep.
• MEMS mirrors — Micro-electromechanical mirrors used in compact, solid-state sensors. Common in automotive LiDAR and newer UAV sensors.
• Flash LiDAR — Instead of scanning, a single wide pulse illuminates the entire scene at once, and a detector array captures the returns simultaneously. Lower range but no moving parts.

Multiple returns

One of LiDAR’s most powerful features is its ability to record multiple returns from a single laser pulse. When a pulse hits the edge of a tree canopy, part of the energy reflects back immediately (first return) while the rest continues through gaps in the foliage, hitting branches (intermediate returns) and eventually the ground (last return).

This is why LiDAR is so valuable for forestry and terrain modelling — it can “see through” vegetation to measure the ground surface beneath. Photogrammetry, which relies on visible light, cannot do this.

A typical airborne LiDAR system records up to five returns per pulse, while full-waveform systems digitise the entire return signal for even more detailed analysis.

Point cloud generation

The raw output of a LiDAR survey is a point cloud: millions or billions of individual XYZ coordinate measurements, each with additional attributes like intensity (how strongly the surface reflected the laser), return number, GPS time, and scan angle.

After collection, the point cloud goes through processing:

1. GPS/INS integration — Combining the laser measurements with the sensor’s position (GPS) and orientation (inertial navigation system) to compute accurate ground coordinates.
2. Strip adjustment — Aligning overlapping flight lines or scan passes to remove systematic offsets.
3. Classification — Labelling each point as ground, vegetation, building, water, noise, etc. This is where you separate the bare earth from everything above it.
4. Quality control — Checking accuracy against ground control points, removing outliers, and verifying point density meets specification.

The resulting classified point cloud is typically stored in LAS, LAZ, or E57 format.

Types of LiDAR

Airborne LiDAR (ALS)

Airborne laser scanning mounts the sensor on a fixed-wing aircraft or helicopter, typically flying at 500-3000 metres above ground. It’s the workhorse of large-area topographic mapping.

Common applications: National elevation datasets, flood plain mapping, transmission line corridors, large infrastructure projects, coastal erosion monitoring.

Terrestrial LiDAR (TLS)

Terrestrial laser scanning uses a tripod-mounted sensor that rotates 360 degrees, capturing everything visible from a single position. Multiple scan positions are registered together to build a complete model.

Common applications: Building facades, heritage documentation, industrial plant as-builts, forensic scene capture, deformation monitoring.

Terrestrial scanners from manufacturers like Leica (RTC360, P-series), FARO (Focus), Trimble (X7, X9), and Riegl (VZ series) typically export data in E57 format, which supports multiple scan positions and embedded panoramic imagery.

Mobile LiDAR (MLS)

Mobile laser scanning mounts the sensor on a vehicle — car, boat, rail vehicle, or backpack. It captures data continuously while moving, combining LiDAR with GPS and inertial navigation.

Common applications: Road corridor mapping, rail surveys, utility pole inventories, streetscape modelling, indoor mapping (SLAM-based).

UAV LiDAR

UAV-mounted LiDAR has transformed the survey industry over the past five years. Compact sensors from DJI (Zenmuse L2), YellowScan, Riegl (miniVUX), and Phoenix are now routinely deployed on commercial drones.

Common applications: Topographic surveys, construction site monitoring, mine stockpile volumes, forestry canopy analysis, archaeological prospection, power line inspection.

UAV LiDAR sits in the sweet spot between terrestrial (very high detail, small area) and airborne (lower detail, very large area). For most survey companies, it has become the primary data collection method for sites under 200 hectares.

Bathymetric LiDAR

A specialised variant that uses green wavelength lasers (532 nm) that penetrate water. The sensor records returns from both the water surface and the seabed or riverbed below.

Common applications: Coastal zone mapping, river channel surveys, reef monitoring, harbour depth measurement.

LiDAR file formats

The data from all these systems needs to be stored, processed, and delivered. Three formats dominate:

LAS is the uncompressed standard maintained by ASPRS. Every point cloud tool supports it, but file sizes are large — a typical drone survey might produce 1-15 GB of LAS data.

LAZ is the losslessly compressed version of LAS, typically 5-15x smaller with zero data loss. For delivery, LAZ should be your default. I covered this in detail in LAS vs LAZ vs E57: Which Point Cloud Format Should You Deliver?

E57 is the ASTM standard format designed for terrestrial scanners. It supports multiple scan positions, scanner metadata, and embedded panoramic images — features that LAS/LAZ lack. Read more in What is E57 Format?

Other formats you may encounter include PLY (common in photogrammetry and research), PTS/PTX (Leica legacy), XYZ/CSV (plain text), and RCP/RCS (Autodesk proprietary). For client delivery, stick to LAZ or E57.

LiDAR accuracy and point density

Two metrics define LiDAR data quality: accuracy and point density.

Accuracy

These figures assume proper ground control, calibrated sensors, and correct processing. Without ground control points (GCPs), accuracy degrades significantly — PPK GNSS on a UAV might get you 3-5 cm vertical without GCPs, but a poorly configured system could be off by 30+ cm.

Point density

Point density (points per square metre) determines the level of detail in your data. Higher density means smaller features can be resolved:

Most survey specifications call for 10-50 points per square metre from UAV LiDAR, which modern sensors achieve comfortably at normal flying heights.

Applications of LiDAR

Surveying and construction

LiDAR is now the primary data source for topographic surveys, earthworks measurement, and construction monitoring. UAV LiDAR surveys a 50-hectare site in a single flight, delivering a classified point cloud that can generate contour maps, DTMs, cross-sections, and volume calculations.

For construction monitoring, repeat surveys at regular intervals create a time series showing progress against design. The point cloud becomes the single source of truth — any measurement can be extracted after the fact.

Mining

Stockpile volume measurement, pit progression tracking, and highwall stability monitoring. LiDAR’s ability to measure vertical surfaces accurately makes it particularly valuable for open-pit mining.

Forestry

Canopy height models, stem density estimates, biomass calculation, and forest inventory. LiDAR’s multiple returns allow measurement of both the canopy surface and the ground beneath it — a capability no other remote sensing technology matches.

Archaeology

LiDAR has revealed thousands of previously unknown archaeological sites hidden beneath forest canopy. The most famous example is the rediscovery of Maya cities in Central America, where airborne LiDAR penetrated dense jungle to reveal pyramids, roads, and urban grids invisible from the ground or in satellite imagery.

Autonomous vehicles

Self-driving cars use LiDAR to build real-time 3D maps of their surroundings. Automotive LiDAR sensors are solid-state, compact, and designed for 100+ metre range at high update rates. This is a different world from survey-grade LiDAR — lower accuracy but real-time operation.

Power and utilities

Transmission line inspection, vegetation encroachment analysis, and corridor mapping. LiDAR can measure the sag in power lines, the clearance to nearby trees, and the condition of towers — all from a helicopter or drone flying along the corridor.

LiDAR vs photogrammetry

This is one of the most common questions in the survey industry. Both technologies create point clouds and 3D models, but they work differently and have different strengths.

For a deeper dive, see LiDAR vs Photogrammetry: Which Should You Use?

Many survey companies use both: LiDAR for the terrain and accurate measurements, photogrammetry for visual outputs like orthomosaics and textured 3D models. The two datasets complement each other well.

How to share LiDAR data with clients

LiDAR point clouds are large, specialised files. Most clients don’t have CloudCompare or Cyclone installed. This creates a delivery problem — you’ve captured beautiful, accurate data, but the client can’t open it.

The traditional approach is FTP or WeTransfer, forcing clients to download multi-gigabyte files and install desktop software. That’s a poor experience.

Browser-based point cloud viewers solve this. Platforms like Swyvl let you upload LAZ or E57 files and share them via a link — no software installation required. The client opens the link, and the point cloud renders in their browser using Potree streaming.

For more on this topic, see How to Share LAS Files with Clients and How to Send Large LiDAR Files.

Summary

LiDAR has gone from an expensive, specialised technology to an everyday tool for surveyors, engineers, and environmental scientists. UAV-mounted sensors have made it accessible to small survey companies, and browser-based viewers have made the data accessible to clients who don’t have specialised software.

The key things to remember:

• LiDAR measures distance using laser pulses — millions per second, generating dense 3D point clouds.
• Multiple return capability lets it map terrain beneath vegetation — something photogrammetry cannot do.
• Four main platforms — airborne, terrestrial, mobile, and UAV — each suited to different scales and applications.
• LAZ is the delivery format for most LiDAR data. E57 for terrestrial scans.
• Accuracy ranges from millimetres (terrestrial) to centimetres (UAV/airborne) depending on the platform and processing.

Whether you’re commissioning a LiDAR survey or processing one yourself, understanding how the technology works gives you the context to specify, evaluate, and deliver the data properly.