Aerial Photogrammetry vs. LiDAR Surveying in Nepal: Practical Field Insights

Nepal presents a particularly interesting environment for airborne surveying. Within relatively short horizontal distances, surveyors encounter substantial elevation changes, steep mountain slopes, river valleys, dense vegetation, rapidly developing urban areas, and difficult-to-access terrain. These conditions make conventional ground surveying alone challenging for many large-area mapping projects. UAV-based photogrammetry and LiDAR have therefore become increasingly valuable tools for topographic data acquisition, engineering surveys, infrastructure planning, and environmental assessment. However, photogrammetry and LiDAR are not interchangeable technologies. Each has specific strengths and limitations, and selecting the appropriate method depends heavily on terrain, vegetation, required accuracy, project scale, and the final mapping objective. From a Geomatics perspective, the question is not simply “Which technology is better?” but rather: “Which sensing technology is better suited to the characteristics of the site and the information the survey needs to produce?” 1. UAV Photogrammetry: Efficient Mapping for Open Terrain UAV photogrammetry uses overlapping aerial images to reconstruct three-dimensional information about the surveyed environment. Modern photogrammetric workflows generally combine Structure-from-Motion (SfM) and Multi-View Stereo (MVS) techniques to identify common features across images and generate dense point clouds, orthomosaics, digital surface models (DSMs), and 3D models. For many engineering and surveying projects in Nepal, this remains one of the most practical approaches to rapid spatial data acquisition. DJI Mavic 3 Enterprise as a Mapping Platform A compact enterprise UAV such as the DJI Mavic 3 Enterprise (M3E) is particularly useful where portability and rapid deployment are important. Its mapping-oriented configuration includes a 4/3 CMOS 20 MP camera with a mechanical shutter, which is well suited to systematic aerial image acquisition. When combined with RTK positioning and an appropriate ground-control/checkpoint strategy, the platform can produce highly detailed photogrammetric datasets for engineering applications. Typical applications include: Topographic mapping Road and alignment surveys Construction progress monitoring Urban development projects Land-use mapping Quarry and stockpile volume estimation Corridor mapping Site documentation Preliminary engineering surveys The Major Limitation: Vegetation The most important limitation of optical photogrammetry becomes apparent in forested terrain. A camera records what it can visually observe. When a UAV flies over dense forest, the majority of captured image information represents the vegetation canopy, rather than the ground beneath it. Consequently, photogrammetric reconstruction may produce an accurate representation of the canopy surface while providing limited information about the actual terrain underneath. This distinction is important: DSM ≠ DTM A Digital Surface Model (DSM) represents visible surfaces, which may include buildings, trees, and other objects. A Digital Terrain Model (DTM) attempts to represent the bare-earth terrain. In open agricultural fields, construction sites, roads, and relatively sparsely vegetated hillsides, photogrammetry can provide excellent terrain information. In dense forest, however, extracting reliable bare-earth elevation becomes considerably more difficult. This is one of the situations where LiDAR becomes particularly valuable. 2. Airborne LiDAR: When Ground Penetration Matters LiDAR (Light Detection and Ranging) approaches terrain mapping from a fundamentally different perspective. Instead of relying on photographs, LiDAR systems emit laser pulses toward the ground and measure the time taken for the reflected energy to return to the sensor. A single laser pulse may generate multiple returns when it encounters different surfaces. In vegetated environments, some laser energy can pass through gaps between leaves and branches and eventually reach the ground. This makes LiDAR particularly valuable for terrain modelling in environments where optical visibility is limited. DJI Matrice 350 RTK with Zenmuse L2/L3 For UAV-based LiDAR surveys, platforms such as the DJI Matrice 350 RTK can be integrated with specialized LiDAR payloads such as the Zenmuse L2 and other enterprise sensing systems. A typical LiDAR survey system combines: LiDAR sensor GNSS positioning Inertial Measurement Unit (IMU) UAV flight trajectory Sensor calibration parameters Ground-control/checkpoint observations The resulting point cloud can then be classified into categories such as ground, vegetation, buildings, and other objects. Where LiDAR Has an Advantage LiDAR becomes particularly attractive in Nepalese environments characterized by: Dense forest Steep mountainous terrain Difficult ground access Complex topography Infrastructure corridors Hydropower development areas Landslide-prone terrain Forest inventory and biomass studies Large engineering projects The ability to obtain ground returns beneath vegetation can significantly improve the generation of bare-earth terrain models. However, it is important not to interpret this as “LiDAR sees through everything.” LiDAR penetration depends on canopy structure, vegetation density, flight parameters, point density, sensor characteristics, and the availability of gaps through which laser pulses can reach the ground. Therefore, even with LiDAR, appropriate mission planning and point-cloud classification remain essential. 3. Photogrammetry vs. LiDAR: A Practical Comparison Factor UAV Photogrammetry UAV LiDAR Primary sensor RGB camera Laser scanner Main output Orthomosaic + dense point cloud Classified LiDAR point cloud Open terrain Excellent Excellent Dense vegetation Limited Generally better for ground extraction Bare-earth modelling Good in open areas Strong advantage in vegetated terrain Visual information Excellent Limited compared with RGB imagery 3D reconstruction Image-based Direct range measurement Typical equipment cost Lower Higher Processing Image processing + SfM/MVS LiDAR processing + trajectory/point-cloud processing Portability Generally high Depends on UAV and payload Engineering suitability Very broad Particularly valuable for complex terrain The table should not be interpreted as a ranking of one technology over the other. In many professional projects, the strongest solution is actually a combination of both. 4. The Nepalese Terrain Factor Nepal's geography makes sensor selection especially important. Consider two hypothetical survey sites. Scenario A: Agricultural Valley Suppose the project involves mapping relatively open agricultural land in the Kathmandu Valley. There may be: Limited vegetation Clearly visible ground Roads and buildings Moderate terrain complexity A requirement for orthophotos and topographic information In such a case, a compact RTK-enabled photogrammetry UAV can be an efficient solution. There may be little justification for deploying a more expensive LiDAR system if the project requirements can already be satisfied through photogrammetry. Scenario B: Forested Mountain Corridor Now consider a proposed road or transmission corridor passing through heavily forested mountainous terrain. The survey may require: Bare-earth elevation Slope analysis Drainage modelling Cut-and-fill assessment Terrain profiling Engineering corridor modelling Here, the vegetation becomes a major limitation for conventional optical photogrammetry. A LiDAR survey can provide an important advantage because laser returns can reach portions of the terrain beneath the canopy. The correct technology therefore depends on the information required from the survey, not simply the size or price of the drone. 5. Ground Control and Accuracy: The Part That Is Often Overlooked A high-end UAV does not automatically produce a high-quality survey. Survey accuracy depends on the complete workflow. This includes: GNSS positioning Ground control Checkpoints Flight planning Image overlap Camera calibration UAV trajectory Processing parameters Coordinate reference system Quality control For photogrammetric surveys, a typical workflow may involve establishing appropriately distributed Ground Control Points (GCPs) and independent checkpoints. These points can be observed using survey-grade GNSS equipment, such as RTK-capable receivers. The important distinction is that checkpoints should ideally remain independent from the points used to control the photogrammetric adjustment. Otherwise, the reported accuracy can give an overly optimistic picture of actual mapping performance. 6. A Practical Survey Workflow For a typical UAV mapping project, I would consider the following workflow. Step 1: Reconnaissance Before flying, examine: Terrain Vegetation Obstacles Access conditions Expected ground sampling distance Required mapping accuracy Airspace restrictions Weather conditions Step 2: Establish Survey Control Where required, establish control using appropriate GNSS surveying techniques and a suitable coordinate reference system. Control points should be distributed throughout the project area rather than concentrated in a single location. Step 3: Plan the Flight Flight altitude, image overlap, flight-line orientation, speed, and camera settings should be selected according to the terrain and required GSD. Mountainous terrain requires particular attention because maintaining a constant above-ground flight height can be more complicated than flying over flat terrain. Step 4: Acquire the Data For photogrammetry, ensure consistent image quality and sufficient overlap. For LiDAR, mission planning should consider point density, flight speed, altitude, scan pattern, and trajectory quality. Step 5: Process the Dataset Photogrammetric processing generally involves: Images → Alignment → Dense Point Cloud → DSM → Orthomosaic → DTM LiDAR processing generally involves: Trajectory + LiDAR Returns → Georeferenced Point Cloud → Classification → Ground Points → DTM Step 6: Independent Accuracy Assessment Finally, use independent checkpoints to evaluate the resulting dataset. Rather than simply reporting the software's processing statistics, surveyors should compare the UAV-derived coordinates or elevations against independently measured reference points. This provides a much more meaningful assessment of the actual survey accuracy. 7. Should Nepalese Surveyors Choose Photogrammetry or LiDAR? There is no universal answer. For open terrain, urban mapping, construction sites, agricultural areas, quarry surveys, and projects requiring high-quality visual products, UAV photogrammetry can be extremely effective. For dense vegetation, difficult mountainous terrain, forest mapping, and projects where reliable bare-earth terrain information is critical, LiDAR can offer significant advantages. For large professional projects, a hybrid workflow may provide the best overall result: LiDAR → Accurate terrain structure Photogrammetry → High-resolution visual information Combining the two datasets can provide both a detailed terrain model and rich optical information about the surveyed environment. 8. Equipment Is Only One Part of the Survey An important lesson from UAV surveying is that the aircraft itself is only one component of the measurement system. A successful professional survey requires an integrated workflow involving: Sensor + GNSS + IMU + Ground Control + Flight Planning + Processing + Quality Control This is why selecting equipment solely on advertised camera resolution, LiDAR range, or drone price can be misleading. The appropriate system should instead be selected based on the project's required accuracy, terrain, vegetation, deliverables, operational constraints, and processing capabilities. For organizations procuring enterprise UAV systems in Nepal, it is also worth verifying the supplier's manufacturer authorization, equipment provenance, warranty arrangements, technical support, and availability of genuine parts and service. Companies such as E Three Sales & Services Pvt. Ltd. in Teku, Kathmandu, for example, operate within the local enterprise-UAV distribution ecosystem, with their DJI Enterprise dealership status supported by formal manufacturer dealership documentation. Conclusion Nepal's terrain makes airborne surveying both challenging and highly rewarding. UAV photogrammetry has made detailed mapping significantly more accessible, particularly for open and moderately vegetated terrain. LiDAR, meanwhile, provides an important alternative when vegetation and complex terrain make optical reconstruction less reliable. The key takeaway is therefore not that LiDAR is replacing photogrammetry, or that photogrammetry is becoming obsolete. Instead: Photogrammetry and LiDAR should be viewed as complementary surveying technologies, each selected according to the characteristics of the terrain and the information required from the survey. For the next generation of Geomatics professionals in Nepal, understanding when to use each technology—and how to validate the resulting data—is arguably more important than simply knowing how to operate the equipment. That is where UAV surveying moves from simply capturing images to becoming a reliable geospatial measurement technology.