GIS/MEA582: Geospatial Modeling and Analysis

Geomorphometry I: Terrain Modeling

Resources: ESRI virtual campus Create the folder 'Modeling'.
Download the Raleigh 2013 lidar data as LAS file LAS tile 0793_016 in NC State Plane Meters and the orthophoto to the Modeling folder. Unpack the package containing orthophoto.

Bring lidar data into ArcGIS

There are multiple ways of handling lidar data in ArcGIS. We will use LAS datasets and multipoint files. For other workflows refer to the FYI section at the end of the assignment

Analyze multiple return lidar data

Create a LAS dataset from .las point cloud

LAS datasets provide a way to quickly view the properties of a point cloud. You can also use binning or TINs to create a DSM or DEM. This is especially useful if you want to process several .las files at once.

Open ArcToolbox
Select 'Data Management->LAS Dataset->Create LAS Dataset'
Set 'Input Files' to 'tile_0793_016_spm.las'
Set 'Output LAS Dataset' to '.\Modeling\lidar_all'
Set 'Coordinate System' to 'NAD_1983_HARN_StatePlane_North_Carolina_FIPS_3200' (you can type 3358 in search window)
Check 'Compute Statistics'
Click 'OK'

View LAS dataset properties and filter points

Open ArcCatalog
Right click on LAS dataset and select 'Properties...'
Select the 'Files' tab and take note of average point spacing (will use in later step).
Select the 'Statistics' tab to view point statistics by returns and classes.
Click 'Ok'

What percentage of the points are classified as ground?

Open ArcMap and add the LAS dataset to the layer manager ('Add data' button or drag and drop from ArcCatalog)
Right click on LAS dataset in Layers and click 'Properties...'
Select 'Filter' tab.
Try applying several filter settings using classes and returns and view the results.
Check box next to '2 Ground' under 'Classification Codes'

What is the difference between 'Last Return' and 'Last of Many'?

With the LAS dataset filtered for ground points only, now we can analyze the point density of ground points and create a DEM with binning.

Compute a raster map representing number of points per cell (at two different resolutions)

Calculate point density at 2m resolution:

Open ArcToolbox
Select 'Data Management->LAS Dataset->LAS Point Statistics as Raster'
Set 'Input Features' to 'lidar_all'
Set 'Output Raster' to '.\Modeling\ground_count2m'
Set 'Method' to 'POINT_COUNT'
Set 'Sampling Type' to 'CELLSIZE'
Set 'Sampling Value' to '2'                            
Click 'OK'
Right click on resulting raster in Layers and select 'Properties'
Click 'Symbology' tab and change color ramp to something appropriate.
Set 'Display NoData as' to red to highlight any areas with zero points.

Calculate point density at 6m resolution:

Open ArcToolbox
Select 'Data Management->LAS Dataset->LAS Point Statistics as Raster'
Set 'Input Features' to 'lidar_all'
Set 'Output Raster' to '.\Modeling\ground_count2m'
Set 'Method' to 'POINT_COUNT'
Set 'Sampling Type' to 'CELLSIZE'
Set 'Sampling Value' to '6'                            
Click 'OK'
Right click on resulting raster in Layers and select 'Properties'
Click 'Symbology' tab and change color ramp to something appropriate.
Set 'Display NoData as' to red to highlight any areas with zero points.

Compute a raster map representing range of elevation values per cell at 6m resolution

Open ArcToolbox
Select 'Data Management->LAS Dataset->LAS Point Statistics as Raster'
Set 'Input Features' to 'lidar_all'
Set 'Output Raster' to '.\Modeling\ground_range6m'
Set 'Method' to 'Z_RANGE'
Set 'Sampling Type' to 'CELLSIZE'
Set 'Sampling Value' to '6'                            
Click 'OK'

Use binning to compute a raster map representing average elevation per cell for ground points

Based on the point density rasters created in the previous steps, what would be an appropriate resolution for binning the ground points? What does the range raster tell you about binning at this resolution?

Ensure that the LAS dataset is still filtered to include only class 2 (ground) points.
Open ArcToolbox
Select 'Conversion Tools->To Raster->LAS Dataset to Raster'
Set 'Input LAS Dataset' to 'lidar_all'
Set 'Output Raster' to '.\Modeling\ground_mean6m'
Set 'Value field' to 'ELEVATION'
Set 'Interpolation Type' to 'Binning'
Set 'Cell Assignment Type' to 'Average'
Set 'Void Fill Method' to 'LINEAR'
Set 'Sampling Type' to 'CELLSIZE'
Set 'Sampling Value' to '6'                            
Click 'OK'

Use multipoint file to interpolate DSM from multiple return lidar data

LAS to multipoint can be used to create a vector point file from a point cloud. The points can be filtered by class or return and then interpolated to create a raster.

Add some background data to give context:

Select Menu File -> Add Data -> Add Data...
Select ncshape.mdb/streets_wake
Select ncshape.mdb/lakes
Select ncshape.mdb/streams

Display orthophoto:

Select Menu 'File -> Add Data -> Add Data...'
Select ./Modeling/ortho_0793_016_ncspm.tif

Create Multipoint files from .las point cloud and interpolate DSM

Turn on the 3D Analyst Extension
Menu -> Customize ->Extensions...
Open the ArcToolbox
Select '3D Analyst->Conversion->From File->LAS to Multipoint'
Set 'Input' to 'tile_0793_016_spm.las'
Set 'Output Feature Class' to '.\Modeling\lidar_first'
Set 'Average Point Spacing' to value listed in Files tab of LAS dataset properties, noted earlier.
Leave the 'Input Class Codes' empty
For 'Input Return Value', uncheck 'ANY_RETURNS' and check '1'
Set 'Coordinate System' to 'NAD_1983_StatePlane_North_Carolina_FIPS_3200' (you can type 3358 in the search window)
Click 'OK'

Interpolate DSM:

Check out a Geostatistical Analyst Extension license
Turn on the Geostatistical Analyst Extension
Menu -> Customize ->Extensions...
Add Geostatistical Analyst toolbar by right clicking on the empty space next to the Main Toolbar -> check Geostatistical Analyst
Select 'Geostatistical Analyst Toolbar->Geostatistical Wizard'
Under 'Deterministic Methods' select 'Radial Basis Function' 
Set 'Source Dataset' to 'lidar_first'
Set 'Data Field' to 'Z' (or 'Shape.Z')
Click 'Next'
Under 'General Properties->Kernel Functions' select 'Spline with Tension'
Click 'Finish'
Click 'OK' to close report

Find out what areas have multiple returns:

Select '3D Analyst->Conversion->From File->LAS to Multipoint'
Set 'Output Feature Class' to '.\Modeling\lidar_multi'
Fill the rest as before but in 'Input return values', uncheck 'ANY_RETURNS' and check all values greater than 1. (2-8)

Overlay result with orthophoto. Where do we have multiple returns?

Optional - display the multiple return data as points in ArcScene

Start ArcScene, add lidar_multi and manipulate with the 3D scene.

FYI: Handling LAS data in ArcGIS

There are three different formats (datasets) that can be used to manage and work with your lidar data in ArcGIS

1. LAS Datasets

To import LAS files into ArcGIS, run the Create LAS Dataset tool in the Data Management toolbox. This will prompt you to load all of your LAS files and create a dataset from them. You can then use the LAS Dataset to Raster tool in the Conversion toolbox to create a raster DEM. See ArcGIS help.

2. Mosaics

You can also import LAS using the Add Rasters to Mosaic Dataset tool. See ArcGIS help.

3. Terrains

The Terrain Wizard in ArcCatalog. Create a feature dataset inside a geodatabase. Import your LAS files using the LAS to Multipoint tool in the 3D Analyst. Save the multipoint feature class in your new feature dataset. Right click on the feature dataset and select New/Terrain to launch the Terrain Wizard. See ArcGIS help here and here.