In this tutorial, we will learn how to work with lidar derived raster product that represent topography and snow depth. The objectives include:

1. Understanding the basic concept of lidar remote sensing
2. Deriving snow products from lidar DEMs differencing
3. Analyzing lidar raster data
4. Identifying available lidar data for snow science

The prerequisite for this tutorial is basic experience with python and geospatial data.

Overview¶

We will start by exploring point cloud data and take a quick look at how rasters are created from them. Then we will derive vegetation and snow depth by differencing elevation models followed by analysing these lidar data including DEM. Finally, we will address common issue with working with lidar data from. This is the first of a 3 part series on LiDAR remote sensing tutorial that I presented at the 2022 snowex hackweek. The other two include:

• Elevation Differencing
• Common Pitfall in Elevation Differencing

What is Lidar¶

Lidar stands for light detection and Ranging also called 3D laser scanning. It is an active remote sensing techniques for measuring elevation across wide areas.

• Pulses of laser light are beamed to surface
• Return travel time determines distance to surface

Lidar provides highly accuracte elevation data at high resolution

There are different types depending on: platform used, physical process or scattering process. Based on the platform used, we find 3 types of LiDAR:

1. Ground-based LiDAR
2. Airborne LiDAR
3. Spaceborne LiDAR

For this tutorial, we will focus on Airborne LiDAR. Airborne LiDAR technology is used to measure topography using a laser beam directed towards the ground with GPS and IMU systems providing the location and orientation of the airborne platform.

Lidar records two types of data- Range distance and Intensity. Range is a function of two-way travel time to ground and back. Intensity is the fraction of photons returned controlled by material properties.

Each lidar pulse yields multiple returns. The number and complexity of returns provide information about the surface material.

Depending on the feature the point return is coming from, elevation models can be Digital Terrain Model (DTM) or Digital Surface Model (DSM). DTM is the three-dimensional representation of the terrain without natural or man-made objects. DSM is a three-dimensional representation of the heights of the Earth’s surface, including natural or man-made objects located on it.

Explore Point Cloud¶

In this activity, we will open a .las file, in the plas.io free online lidar data viewer and explore some attributes associated with a lidar data point cloud.

1. Load the data: under file, select Autzen stadium. To navigate around, use left click and drag to tilt up and down, righ click and drag to move around. scroll bar to zoom in and out
2. Adjust Intensity
3. Change the lidar point cloud color options to classification
4. Explore

Point to Pixels¶

Points clouds give us a lot of information (x, y, z and intensity). However, these data might be too large, overwhelming and difficult for scientific application. Lidar data are often provided in raster data format which is easier to work with in many GIS tools. Raster data can be generated from point clouds using PDAL. This is done in two approaches: Filtering and Rasterizing. The filtering pipeline involves various steps such as reprojection, reclassification, noise removal, outlier removal, segmentation and extraction. The filtered point cloud can then be rasterized using preferred algorithm. See this post for detailed process