Department of Land Surveying & Geo-Informatics

The Hong Kong Polytechnic University

LSGI3220 Building Information Modelling & 3D GIS

Lab 3 Working with LiDAR
This exercise describes how to work with light detection and ranging (LiDAR) data in ArcGIS
Pro. LiDAR data are commonly made available as point clouds in LAS format. You will learn to
process these LiDAR data to create LAS Datasets, digital elevation models (DEMs), digital
surface models (DSMs), normalized digital surface models (nDSMs), first return intensity
images, and hillshades. You will also explore the statistics of the point cloud.
The LiDAR data used in this lab are large files, so it may take your computer some time to
complete each task. These data will support the creation of 1 m spatial resolution raster grids;
however, you will produce 5 m resolution products in this lab to reduce processing time and
file size. (Cited from and thanks to Prof. Maxwell, Geography in the Department of Geology
and Geography at West Virginia University)

Topics covered in this exercise include:

1. Create LAS Datasets
2. Filter point clouds
3. Rasterize LiDAR point cloud data as DEMs, DSMs, nDSMs, and intensity
images.
4. Analyze and interpret LiDAR derivatives
5. Extract raster values at points

Step 1. Create and Prepare a New Project

You will begin the analysis by creating a new project to work
within.

 Open ArcGIS Pro. This can be done by navigating to

All Apps followed by the ArcGIS Folder. Within the
ArcGIS Folder, select ArcGIS Pro. Note that you can
also use a Task Bar or Desktop shortcut if they are
available on your machine.
 Once ArcGIS Pro launches, select Map.aptx under
Create a new project on the right side of the page.

1
  In the Create a New Project Dialog Box, name your new project Lab 3 and save it to
your personal folder. You can leave the “Create a new folder for this project” option
selected.

You have now created a new project. Since you used the Map.aptx template, a map has
already been added to the project. You will now need to add data to the map. You will need to
download the required data from Blackboard. You will need to uncompress the folder.
 Click on the Add Data Button. Navigate to your copy of the lab data. Navigate to the
downloaded Data folder. Add the parsons.img and tree_points.shp files.
Note: The map will take on the datum and projection of the added layers: NAD 1983 UTM
Zone 17 North.

2
Step 2. Create a LAS Dataset
In order to work with LiDAR data in ArcGIS Pro, you must create a LAS Dataset from the
LAS files. This can be accomplished using the Create LAS Dataset Tool.
Note: In this lab, we will access tools from ArcToolbox. However, there are many ways to
access tools in ArcGIS Pro. For example, some of the more common tools are provided in the
Tools list in the Analysis Tab.

Once you open the Geoprocessing Pane, you can access favorite tools or search for Tools.

We have decided to demonstrate ArcToolbox here so that you get a sense of where the tools are
located in the Toolbox directory.
 In the Analysis Tab, select Tools from the
Geoprocessing Area. This should open the
Geoprocessing Pane.
 In the Geoprocessing Pane,
navigate to the Toolboxes.
Note: We will not provide these directions for accessing other tools. We will just tell you
where to find them within ArcToolbox.
 Navigate to Data Management Tools followed by the LAS Dataset subtoolbox.
Click on the Create LAS Dataset Tool.

 Add the four LAS files from the Data folder to the Input Files list (C08582.las,
3
 C08772.las, C08773.las, and C08581.las).
 Name the Output LAS Dataset lidar_data.lasd and save it to a location of your
choosing.
 Make sure the Coordinate System is set to NAD_1983_Zone_17N/VCS: NAVD88.
 You do not need to change any of the other settings.
 Click Run to execute the tool.

 In order to view the LiDAR points, you will need to move the
lidar_data.lasd layer to the top of the list in the Contents Pane.
When you are zoomed out, the extent of the LiDAR tiles are drawn as red boxes. If you
want to see individual points, you will need to zoom in.
Question 1. What are the dimensions of each of the four LiDAR tiles in kilometers?

4
Step 3. Create a DEM
You will now convert the point cloud data to a digital elevation model or DEM. To create this
surface, you only want to use ground returns. So, you will first need to adjust the filter.
 Right-click on the lidar_data.lasd layer in the Contents Pane then select LAS Filter
followed by Ground.

Now, only ground returns should be displayed or used in an analysis. You are now ready to
create the DEM.

5
  Navigate to Conversion Tools followed by the LAS subtoolbox. Click on the LAS
Dataset to Raster Tool.
 Set the Input LAS Dataset to the lidar_data.lasd layer.
 Name the Output Raster dem and save it to a location of your choosing.
 Make sure the Value Field is set to Elevation.
 Make sure the Interpolation Type is set to Binning, the Cell Assignment is Average, and
the Void Fill Method is Linear.
 Make sure the Output Data Type is set to Floating Point.
 Make sure the Sampling Type is set to Cell Size.
 Set the Sampling Value to 5.
 Click Run to execute the tool.

Step 4. Create a Hillshade

To visualize the terrain, you will now create a hillshade from the DEM.
 Navigate to Spatial Analyst Tools followed by the Surface subtoolbox. Click on the
Hillshade Tool.
 Set the Input Raster to the dem layer.
 Name the Output Raster hs and save it to a location of your choosing.
 You do not need to change any of the other settings.
 Click Run to execute the tool.

6
Question 2. Why was the output type defined as float instead of integer?
Step 5. Create a DSM
You will now need to create a surface that includes not just ground elevations but also
incorporates aboveground features, such as buildings and trees. This is called a digital surface
model or DSM. You will created this surface from the first laser pulse returned, so you will need
to change the filter.
 Right-click on the lidar_data.lasd layer in the Contents Pane then select Properties.
Navigate to LAS Filter.
 Under Classification Codes, make sure only Unassigned and Ground are selected.
Under Return Values make sure only First of Many and Single are selected. Click OK
to change the filter.

7
You are now ready to create the DSM using the LAS Dataset to Raster Tool.

 Navigate to Conversion Tools followed by the To Raster subtoolbox. Click on the

LAS Dataset to Raster Tool.
 Set the Input LAS Dataset to the lidar_data.lasd layer.
 Name the Output Raster dsm and save it to a location of your choosing.
 Make sure the Value Field is set to Elevation.
 Make sure the Interpolation Type is set to Binning, the Cell Assignment is Maximum,
and the Void Fill Method is Linear.
 Make sure the Output Data Type is set to Floating Point.
 Make sure the Sampling Type is set to Cell Size.
 Set the Sampling Value to 5.
 Under Environments, set the Snap Raster to the dem layer. This will make the two
layers align correctly.
 Click Run to execute the tool.

8
Question 3. Why did you use Maximum in this case as the Cell Assignment instead of
Average?
Step 6. Create an nDSM from the DEM and DSM
A normalized digital surface model (nDSM) is a DSM that has had the ground elevations
subtracted out. So, it can be created by subtracting the DEM from the DSM. This surface
represents the height of above ground features, such as trees and buildings. You will now create
this surface using the Raster Calculator Tool.
 Navigate to Spatial Analyst Tools followed by the Map Algebra subtoolbox.
Click on the Raster Calculator Tool.
 Use the following syntax: "dsm" - "dem".
 Name the Output Raster ndsm and save it to a location of your choosing.
 Under Environments, set the Snap Raster to the dem layer. This will make the two
layers align correctly.
 Click Run to execute the tool.

9
Question 4. Using the aerial image and the nDSM, what features on the landscape in this area are
producing above ground heights?

10
Step 7. Determine Heights of Trees using the nDSM
The tree_points layer represents the location of some trees. You will now use the nDSM to
predict the height of these trees. To do this, you will need to extract the raster value at each
point location. This can be accomplished using the Extract Multi Values to Points Tool. This
could also be accomplished using the Extract Values to Points Tool; however, that will not
be demonstrated here.
 Navigate to Spatial Analyst Tools followed by the Extraction subtoolbox. Click
on the Extract Multi Values to Points Tool.
 Set the Input Point Features to the tree_points layer.
 Add the ndsm layer to the Input Rasters list and define the Output Field Name as
“height.” Do not include the quotes.
 Do not select “Bilinear interpolation of values at point locations.”
 Click Run to execute the tool.

Use the attribute table of the tree_points layer to answer the following questions. The
“height” field should have been added.

Question 5. Which tree is the tallest? Use the name provided in the “Tree” field.
Question 6. What is the average height of these trees? Hint: You will need to use
the Summary Statistics Tool.

11
Step 8. First Return Intensity Image
Other than producing DEMs and DSMs, it is also possible to create return intensity images
from the point cloud data. Different surfaces will absorb more or less of the infrared laser
energy. Thus, these images are correlated with land cover.
Note: The DSM was created from first of many returns and single returns. This is the same
filter that is required to create the first return intensity image. So, you will not need to filter
the data again.

 Navigate to Conversion Tools followed by the To Raster subtoolbox. Click on the

LAS Dataset to Raster Tool.
 Set the Input LAS Dataset to the lidar_data.lasd layer.
 Name the Output Raster frst_int and save it to a location of your choosing.
 Make sure the Value Field is set to Intensity.
 Make sure the Interpolation Type is set to Binning, the Cell Assignment is Average, and
change the Void Fill Method to None.
 Make sure the Output Data Type is set to Floating Point.
 Make sure the Sampling Type is set to Cell Size.
 Set the Sampling Value to 5.
 Click Run to execute the tool.

12
We did not have you fill the voids because we assumed that cells that did not contain a return
may have absorbed all of the infrared energy. We would now like you to code these cells as 0.
This can be accomplished using a conditional statement and the Raster Calculator Tool.
 Navigate to Spatial Analyst Tools followed by the Map Algebra subtoolbox.
Click on the Raster Calculator Tool.
 Use the following syntax: Con(IsNull("frst_int"), 0, "frst_int").
 Name the Output Raster frst_int2 and save it to a location of your choosing.
 Click Run to execute the tool.
Note: Here is an interpretation of the syntax. If the cell holds the null assignment, change the
value to 0. If it does not hold the null assignment, or hold any other value, return the original
value. So, this syntax replaces null or NoData values with 0 and does not impact any other
cells.

13
Use the intensity image and aerial image to answer the following questions.
Question 7. Do forests generally show low or high first return intensity? Explain why forests
tend to show different return intensity values than grass or pastureland.
Question 8. Does water generally show high or low return intensity? Does pavement generally
show low or high return intensity?

END.

14