The main purpose of this lab was to gain basic knowledge and understanding of LiDAR data structure and data processing. To do this, I was tasked with assuming the role of a GIS manager working in Eau Claire, Wisconsin, and was tasked with acquiring Eau Claire LiDAR point cloud data in LAS format. I was to use this data to process and retrieve surface and terrain models of Eau Claire and generate an intensity image and other products using the terrain models.
Methods
First, an LAS dataset was created in ArcCatalog, and the necessary LAS files for the exercise were added to the dataset. The statistics of the dataset were then calculated and analysed for reference. The necessary XY Coordinate System and the the Z Coordinate System were applied to the dataset, the "NAD 1983 HARN Wisconsin CRS Eau Claire (US Feet)" and "NAVD 1988 US Feet", respectively. These were chosen by analyzing a set of the metadata provided for the cloud point data. The LAS dataset was then displayed within ArcMap. The LAS dataset was then overlayed on a shapefile of the larger Eau Claire area to verify it had been formatted correctly.
Using the LAS dataset toolbar, the cloud point data was rendered as points color coded by elevation. From here, various functions of the rendering tool were explored. This includes rendering the point clouds elevation as polygons, rendering the aspect of the point clouds, rending the slope of the surface features, and displaying the contour lines of the features. Afterwards, the various preset classification and return filters were explored. This includes the Ground, Non Ground, and First Return preset settings for classification and return. Afterwards, the feature points were viewed in both 2D and 3D interactive viewers.
Using the Conversion Tool "LAS Dataset to Raster", digital models of both the the surface first return and the terrain were created to a cell size of two meters. Using these digital models, hillshades were constructed of both digital models using the 3D Analyst Tool "Hillshade". The output hillshades were then compared and analysed in order to interpret the surface features.
Using the point clouds set to "First Return", an intensity image was generated using the "LAS Dataset to Raster" Conversion Tool to the same cell size as the previous digital models. The output intensity image was then saved as a TIFF file and displayed in ERDAS, as the image appeared darkened in ArcMap. ERDAS automatically enhanced the image, allowing for analysis.
Results
Using the hillshades of both the DSM and DTM digital models (Figure 1), terrain and surface features can be identified for the Eau Claire area. The hillshade of the DSM shows that the Eau Claire shows that the area is densely packed with both vegetation and buildings of various size and height. However, this obscures much of the terrain. The DTM hillshade shows that the area is largely flat except for where it approaches the river, sloping greatly to meet the surface of the river. Using the DTM hillshade also allows for the analysis of the various road networks throughout the area. This is because roads and streets are a ground level feature picked up along with the terrain.
Using LiDAR data, an accurate intensity image of the area can be generated similarly to optical remote sensing images produced through Landsat and other systems. However, unlike Landsat, the visual detail of this image is far greater. LiDAR outputs in the .7-1.5 μm NIR Band. A Landsat 5 image produced in this spectral range outputs to a resolution of 30x30 meters. However, using the "LAS Dataset to Raster" Conversion Tool an image was generated using the LiDAR data to a spatial resolution of 2x2 meters. This far exceeds the capabilities of Landsat imaging, which means that LiDAR can be used for fine details and measurement analysis.
Sources
Eau Claire County. www.co.eau-claire.wi.us/. Accessed 2013.
Price, Margaret H. Mastering ArcGIS. 6th ed., 2014.
Wilson, C. (2016). Lab 5: LiDAR Remote Sensing. Eau Claire, Wisconsin.
No comments:
Post a Comment