Lab IX: Using LIDAR Data With FUSION
Introduction: Airborne LIDAR data has just become widely available over the past few years. There are several commercial software packages available for working with LIDAR data but all are quite expensive. Fortunately, there is also a free software package that has been developed under funding provided by the US Forest Service. Since the development of this software has been supported by federal research dollars, the software is free. Unlike many free software packages, this software is very powerful, there is great technical support and it seems to be very user-friendly. The software is called FUSION and it available for download from this site:
http://www.fs.fed.us/eng/rsac/lidar_training/
From this site, select the “Lidar Fusion Tutorial.” There are a number of other useful links at this site as well.
http://www.fs.fed.us/eng/rsac/fusion/index.html
This site includes instructions for downloading and installing the software (Exercise One) and a very good tutorial that introduces the use of the software. The software is currently installed on the machines in our lab. Simply go to Start-Programs-Misc_Programs-FUSION to launch FUSION.
For today’s lab, I would like you to work through Exercises Two and Three in the tutorial. However, instead of using their sample data (which is located in J:\saldata\esci442\FUSION\Example) I would like you to use some data that cover the WWU campus and vicinity. These data are located in:
J:\saldata\Esci442\lidar
You should get a copy of the entire “lidar” folder and paste it into your workspace under C:\Temp. The \lidar folder includes a subfolder /allreturnsdata, that contains the all-returns lidar data (in four “tiles”, PS1842.las, PS1843.las, PS1914.las, and PS1915.las). You will use these .las files in place of the “lda_4800_data.lda” file that is mentioned in the tutorial. The /wwu_image subfolder includes a version of our 2004 color-IR image that we used for the image segmentation lab. This will be used as your reference image in place of the “orthophoto_4800k.jpg” file that is mentioned in the tutorial. Use /lidar/wwu_surface.dtm in place of the 4800K_ground_surface.dtm that is mentioned in the tutorial.
There is no write-up for this lab. I would just like you to gain some exposure to the software and some appreciation for what can be done with LIDAR data.
Other Stuff to try:
As discussed in class, we will begin by heading out to Sehome Hill to collect some field data using the
Point-centered Quarter method. More
information that you’d ever want to know about this method is available here: PCQM.pdf. The field
data collection process is described in the Preface:
Collecting Ground Truth Data Using the Point-Center-Quarter Method.
Use the Catalog command to obtain information about your raw data such as coverage and point density (Described in Exercise 6.4 of the tutorial)
Create your own ground surface layer from the all-returns data (Described in Exercise 7 of the tutorial)
Create a canopy model and a canopy height model (Described in Exercise 8 of the tutorial)
Use the Gridmetrics command to generate a bozillion different layers that describe the vertical distribution of biomass.
Also, if you are interested in finding LIDAR data for some other part of Whatcom or Skagit counties, we have it on our system. Go to this locations and follow the links:
J:\GEO\GEO_data\NW_LIDAR\NorthPugetSound_USGS\DOCS\HTML\_Quad_index.html
General information about this data set can be found here:
\\ptarmigan\DATA1\GEO\GEO_data\NW_LIDAR\NorthPugetSound_USGS\DOCS\Report_of_Survey.pdf
This site includes the all-returns data as .las files and there are also bare earth layers (as zipped .e00 files) that were created by the vendor.
Updates for 3/11:
Unfortunately, there are problems with the ground truth
data that we have. For starters, we only
have 14 points. Furthermore, at least
some of the locations for these points seem to be off to some degree. Nevertheless, I’d like to walk you through
several processing steps with these data to illustrate the approach that could
be used to model various aspects of forest structure using the all-returns
LIDAR data.
The data:
The data that we have is located in J:\saldata\Esci442\lidar\point-quarter-data.xls. Note that the coordinates in this file are in
UTM NAD83. I imported these points into
ArcMap and reprojected them to State Plane NAD83(feet) and created a shapefile
J:\saldata\Esci442\lidar\pts_sp83.shp.
If you plot these points over our image, you will see that one of the
points (#12) is in the grass to the west of the CF building but the data in the
excel file clearly indicate that this point is in the forest. Not sure what happened here.
Optional: Try running these .bat files:
Catalog.bat
Gridmetrics.bat
Dtm2ascii.bat
I’ll then show you how to convert all of these to grids
and extract data for all of our points.
We will then look at correlations between our lidar
metrics and our ground metrics.
Return to ESCI 442/542 Lab Page
Return to ESCI 442/542 Syllabus
Return to David Wallin's Home Page