Generating landscape elevation rasters from raw LiDAR data

Table of contents

1. Summary of repository
2. How to download raw LiDAR data
3. Description of scripts
4. Point classification table

Summary of repository

The tools within this repository allow users to go from a group of raw LiDAR files (las files) to an elevation raster for select point classifications. For example, we utilize LiDAR data from the Chicago Height Modernization project to generate canopy height and building rasters for all of Cook County, IL. These output rasters can be utilized in a slew of spatial analyses.

Back to top ⤒

How to download raw LiDAR data

Raw LiDAR data is stored as las files. The las files used for this project can be found on the Chicago Height Modernization Webiste. For this analysis all files relevant to Cook County for the year 2022 were used.

Raw las files can be downloaded under the data tab. You can also download DSM and DTM rasters in this tab if you wish to forego the LiDAR processing stage. If you are unsure of what files you need but know the area you want to analyze, use the viewer tab to select only the neccesary files.

BEWARE! LiDAR data is infamous for its aggregious filesizes. Make sure you have proper storage and processing power before considering LiDAR analysis. We specifically recommend storing raw .las files and rasterized output on an external drive if converting across broad extents (e.g. all of Cook County).

Back to top ⤒

Description of scripts

setup.R

This script is called at the top of subsequent scripts. It loads/installs necessary packages.

lidar_processing_folderloop.R

This script reads in .las files, converts them into rasters, and save them as .tif files. You will need to edit folder paths to where you're storing .las files and the folder where you'd like to store output. Additionally, we stored .las files in folders of ~200 files, as we found that stitching rasters together larger than that tended to crash our R environment.

stitch_rast.R

This script can be used to stitch together multiple rasters. Again, we found that attempting to rasterize across a large aggregation of .las files crashed our R environment, so we instead created rasterized output in ~200-file 'chunks.' We used this script to stitch them together.

Back to top ⤒

Point classification table

These classifcations are relevant to the Chicago Height Modernization dataset. Do not assume that other datasets are using these exact classifications.

Class	Value
Never Classified	0
Unassigned	1
Ground	2
Low Vegetation	3
Medium Vegetation	4
High Vegetation	5
Building	6
Noise	7
Model Key/Reserved	8
Water	9
Rail	10
Road Surface	11
Overlap/Reserved	12
Wire - Guard	13
Wire - Conductor	14
Transmission Tower	15
Wire - Connector	16
Bridge Deck	17
High Noise	18
Overhead Structure	19
Ignored Ground	20
Snow	21
Temporal Exclusion	22

Back to top ⤒