Waikato LiDAR 1m DSM (2021)

Data Acquisition:

Airborne Laser Scanner (ALS) data was acquired from a fixed wing aircraft from 5th January 2021 to 25 March 2021, using a Lecia TerrainMapper-LN LiDAR system.

Survey Specification:

- Scanner: Lecia TerrainMapper-LN
- Scan angle: 20 degrees
- Pulse rate: 1,025,000 Hz
- Swath overlap: 20%
- Swath points per M^2: 4

Data Processing:

The LAS files created during the project during the Geopositioning phase and subsequent data
cleaning and deliverable products are in ASPRS LAS 1.4 format, with PDRF 6.

To verify that the project deliverables are in the correct LAS file format/version with PDRF, LP360 File Analyst was used. File Analyst performs exhaustive testing on the LAS file header and data records, and outputs the analysis results to an excel spreadsheet

LINZ PGF specification 6.2 requires that Each Global Navigation Satellite System (GNSS) aircraft positional measurement must be time stamped using Adjusted Global Positioning System (GPS) Time, at a precision enough to allow unique timestamps for each LiDAR pulse. For this project, LAS 1.4 with PDRF 6 has been delivered including, with Global Encoding bit set to 1.

To verify GPS time is set to adjusted GPS time, LP360 File Analyst was used. File Analyst performs exhaustive testing on the LAS file header and data records, and outputs the analysis results to an excel spreadsheet. The Lidar technical manager reviewed this analysis to ensure that all deliverable LAS files are compliant.

LINZ PGF specification 6.3 requires that the required datum for latitude, longitude, and ellipsoid heights is the New Zealand Geodetic Datum 2000. The required vertical datum for normal-orthometric heights is NZVD2016 (Reference 9). Projected data products are to be delivered in NZTM2000 projection (Reference 10) with NZVD2016 normal-orthometric heights. The Survey Datums, Ground Control, Check Points and Lidar Geopositioning sections of the Project Methodology Statement describes in detail how the data is transformed and connected to the required project datum.

To verify that the correct datum information is recorded in the LAS 1.4 header, LP360 File Analyst was used. File Analyst performs exhaustive testing on the LAS file header and data records, and outputs the analysis results to an excel spreadsheet. The Lidar Phase Manager reviewed this analysis to ensure that all deliverable LAS files have the correct CRS applied in the header.

Laser point position was calculated by associating the SBET position to each laser point return
time, scan angle, intensity, etc. Raw laser point cloud data was created for the whole project
area in LAS format. Line-to-line calibrations were then performed for system attitude
parameters (pitch, roll, heading), mirror flex (scale) and GPS/IMU drift.

Laser point position was calculated by associating the SBET position to each laser point return time, scan angle, intensity, etc. Raw laser point cloud data was created for the whole project area in LAS format. Line-to-line calibrations were then performed for system attitude parameters (pitch, roll, heading), mirror flex (scale) and GPS/IMU drift.

Statistical reports were generated for comparison and used to make the necessary adjustments to remove any residual systematic error. These calibration adjustments were performed using a combination of automated and manual corrections to the data to develop the geometrically calibrated data set to be utilized for all downstream processes.

Validation of the point cloud positional accuracies is the primary outcome of the Lidar Geopositioning workflow phase, using surveyed ground control & check points

Points classified as low noise (coverage class 7) and high noise (coverage class 18) will have the LAS withheld flag set.

Classification of the overlap points was done using TerraScan in Microstation. A macro step was created which allowed for points to be classified with the overlap bit by cutting the scan angle at zero degrees. The step used for the project dataset is illustrated in the figure below. This allowed for the overlap points to be withheld from DEM, DSM and intensity imagery generation.

Verification of appropriate overlap flag application is performed using LP360 File Analyst. File Analyst performs exhaustive testing on the LAS file header and data records, and outputs the analysis results to an excel spreadsheet.

Point cloud classification is performed by automated classification algorithms developed by Woolpert’s senior Lidar analysts and reviewed by the Lidar technical manager. A first run automatic classification was carried out on the raw LiDAR points using TerraSolid’s TerraScan software to classify the LiDAR points into ground hits and non-ground hits.

As documented by TerraSolid, the ground routine classifies ground points by creating a triangulated surface model iteratively. The routine is best suited for classifying ground in airborne laser data sets and in data sets where there is mainly natural terrain.

For classifying ground in mobile data sets where the majority of ground is on hard surfaces, such as roads, use the Hard surface routine instead of the ground routine. The routine is sensitive to low error points in the point cloud. Therefore, you should run one or more classification steps using the Low points routine before classifying ground. A more complex classification strategy is required for classifying ground in photogrammetric point clouds.

The ground routine starts by selecting local low points that are confident hits on the ground. The initial point selection is controlled with the Max building size parameter. If the maximum building size is, for example, set to 60.0 m, the routine assumes that any 60 by 60 m area has at least one point on the ground level and that the lowest point is on the ground level. Then, the routine builds a surface model (TIN) from the initial ground points. The triangles in this initial model are mostly below the ground level and only the vertices are touching the ground. In the following iterations, the routine molds the model upwards by adding more and more points. Each added point makes the model following the true ground surface more closely.

The iteration parameters of the routine determine how close a point must be to a triangle plane for being accepted as ground point and added to the model. Iteration angle is the maximum angle between a point, its projection on the triangle plane and the closest triangle vertex. This is the main parameter controlling how many points are classified into the ground class. The smaller the Iteration angle, the less eager the routine is to follow variation in the ground level, such as small undulations in terrain or points on low vegetation. Use a smaller angle value (close to 4.0) in flat terrain and a bigger value (close to 10.0) in mountainous terrain are large. This avoids ground points that are too high, for example within low vegetation or on low buildings.

The iteration angle can be reduced automatically if the triangles become small. This reduces the eagerness to classify more ground points inside small triangles and thus, avoids unnecessary point density of the ground model thus avoiding redundancy of inclusion of unnecessary ground points. Related to this it is common to see default classified points within a classified ground point cloud surface.

The iteration angle inside small triangles approaches zero if the longest triangle edge is shorter than a given Edge length value. Furthermore, the iteration can be stopped completely if triangle edges are shorter than a given limit. After completion of the automated classification, a strenuous manual classification was carried out over the required area to edit the points thus minimizing gross classification errors that may have occurred in the automatic classification process.

Each block’s data was checked in a systematic approach to reduce missing important features. Orthogonal views with background orthoimagery, and profile views are used to review the performance of the automatic classification results. Tools such as ‘Classify using brush’, ’Classify above line’ and ‘Add Point to Ground’ (all found within TerraScan) were used during manual classification, to achieve classification accuracy meeting project specifications. Before handover files were created, checks were done on the project’s points by viewing statistics within TerraScan as illustrated in the figure below.

Please refer to the dataset report for point cloud spatial accuracy check statistics.

All product deliverables supplied in terms of NZTM map projection and NZVD2016 vertical datum.

Classification of the point cloud followed the classifications scheme below:

1- Unclassified
2 - Ground
3 - Low Vegetation
4 - Medium Vegetation
5 - High Vegetation
6 - Buildings
7 - Low Noise
9 - Water
17 - Bridge deck
18 - High Noise

The classification was undertaken in accord with the project specification (Land Information New Zealand, 2020, PGF Version: New Zealand National Aerial LiDAR Base Specification, January 2020).

To generate DEMs, LAS format data files are loaded into TerraScan, only loading in class 2 points and breaklines. Additional data is referenced to avoid tile edge artefacts. Following import, TerraScan-Export Lattice Model command is used to create a grid file with uniform distances between points from one or more selected point classes. For each grid point, the lattice model file stores XY coordinates and elevation. Within the Export Lattice Models command we utilize “Triangulated model Z” to export the DEM tiffs. This triangulated model z derives an elevation value that is calculated from a TIN model of the lidar points using ground and breakline classifications.

To generate DSMs, LAS format data files are loaded into TerraScan, only loading in First, First-of-many and Single returns. Additional data is loaded to avoid tile edge artefacts. Only coverage classed 1 to 6 & 17 points will be used in generation of Digital Surface Models. The elevation grid is generated as a TerraScan Lattice Model using Binning interpolation at 1m cell size.

Hydro-flattening has been performed on the DEM with the incorporation of hydro-flattening features into DEM where islands are 5,000 sq m or larger, ponds and lakes are 10,000 sq m or larger and rivers are ≥30m nominal width. These features are used in the creation of the DEM to reduce the presence of artifacts in the DEM where the point cloud points alone aren’t sufficient to model the landscape. For work where the data is to be used for detailed hydrological modelling hydro-enforcement and enhancement may be required.

The deliverables to LINZ were:

1m gridded bare earth digital elevation model (DEM)
1m gridded digital surface model (DSM)
Classified point cloud

Prior to publication, LINZ made manual fixes to 28 DSM tiles that were missing bridge decks.