🎓 Lesson 3
D2
Sensor Physics: How LiDAR Pulse Width Affects Penetration in Vegetated Highwalls
LiDAR pulse width is how long each laser 'flash' lasts—and shorter flashes let the sensor see through gaps in vegetation to hit the rock face behind, while longer ones blur details and miss narrow openings.
🎯 Learning Objectives
- ✓ Explain how pulse width influences multi-return detection capability in dense vegetation
- ✓ Calculate the minimum resolvable distance between two surfaces using pulse width and speed of light
- ✓ Analyze trade-offs between pulse width, signal-to-noise ratio, and highwall point density for a given drone survey mission
- ✓ Design a LiDAR acquisition strategy (pulse width + PRF + scan rate) for reliable highwall toe detection under 70% canopy cover
📖 Why This Matters
In open-pit mines, highwalls obscured by shrubs, saplings, or regrowth pose serious safety and geotechnical risks—if drones can’t reliably map the true rock surface, slope stability analyses fail, and blast design becomes guesswork. Pulse width isn’t just a spec sheet number: it’s the difference between detecting a critical tension crack 15 cm behind a fern curtain—or missing it entirely. This lesson unlocks how to choose and configure LiDAR hardware so your point cloud reveals the *rock*, not just the *roof*.
📘 Core Principles
LiDAR pulse width defines the temporal ‘thickness’ of each emitted laser burst. When this pulse strikes a complex target—like layered vegetation over rock—it generates multiple returns as photons reflect off leaves, branches, and finally the highwall. A narrow pulse (e.g., 2 ns) lasts only ~60 cm in air (c × τ), allowing distinct time-of-flight measurements for returns separated by >30 cm. A wide pulse (e.g., 20 ns) spans ~6 m—blurring returns from foliage and rock into one ambiguous echo. Crucially, pulse width also affects peak power (for fixed pulse energy): narrower pulses demand higher peak power to maintain detectability, increasing eye-safety constraints and detector saturation risk. In practice, optimal pulse width balances penetration (via multi-return separation) against signal integrity and regulatory compliance (e.g., IEC 60825-1 Class 1 limits).
📐 Minimum Resolvable Distance
The shortest distance between two reflective surfaces that a LiDAR system can distinguish is governed by pulse width and the speed of light. This sets the fundamental limit on vertical layer separation—critical for distinguishing vegetation canopy from highwall.
Axial Resolution Limit
δz = (c × τ) / 2Theoretical minimum vertical distance between two surfaces that produce distinguishable returns.
Variables:
| Symbol | Name | Unit | Description |
|---|---|---|---|
| δz | Axial resolution | m | Minimum resolvable distance along line-of-sight |
| c | Speed of light in air | m/s | Approximately 299,792,458 m/s (use 3.00×10⁸ m/s for estimation) |
| τ | Pulse width (FWHM) | s | Full width at half maximum duration of laser pulse |
Typical Ranges:
Survey-grade UAV LiDAR: 2–6 ns
Long-range terrestrial LiDAR: 5–20 ns
💡 Worked Example
Problem: A drone-mounted topographic LiDAR emits pulses with τ = 4.5 ns. What is the theoretical minimum resolvable distance between two surfaces (e.g., leaf layer and rock face)? Assume speed of light c = 299,792,458 m/s.
1.
Step 1: Recall axial resolution δz ≈ (c × τ) / 2 — division by 2 accounts for round-trip travel.
2.
Step 2: Plug in values: δz = (299,792,458 m/s × 4.5 × 10⁻⁹ s) / 2
3.
Step 3: Compute: δz ≈ (1.349) / 2 ≈ 0.6745 m → rounded to 0.67 m.
Answer:
The minimum resolvable distance is 0.67 m, meaning surfaces closer than ~67 cm vertically (e.g., thin brush <30 cm tall over rock) will merge into a single return—requiring multi-return processing or narrower pulses for resolution.
🏗️ Real-World Application
At Newmont’s Boddington Mine (Western Australia), drone LiDAR surveys of vegetated waste dump highwalls initially failed to detect incipient sloughing zones. Post-mission analysis revealed the OEM sensor used τ = 12 ns pulses—producing δz ≈ 1.8 m—so low-density understory (20–50 cm tall) and the underlying rock returned as a single, smeared elevation. Switching to a custom-configured RIEGL VUX-1HA with selectable τ = 2.5 ns (δz ≈ 0.37 m) and optimized waveform digitization enabled clear separation of 3–5 returns per shot. This revealed a previously hidden 0.5 m-deep tension crack zone along the toe—verified by ground truthing—and triggered redesign of the next blast pattern to relieve stress concentrations.