Jens Ammon
DEM Source Comparison
This exercise demonstrates the advantages and disadvantages of Digital Elevation Models created from various source elevation data and resolutions: 0.5 m LiDAR data, 5 m Auto-Corrected data, and 10 m USGS NED data.
​
The accompanying DEMs depict the peak of Monroe Mountain, UT, an area of interest due to the high-elevation biodiversity it supports in an otherwise low-elevation region. In this scenario, we are comparing the different DEMs in their effectiveness in identifying specific elevation zones where such biodiversity might thrive. The identified regions could then be protected from trail and recreational development.
Map Creation Process:
- Once the elevation data was imported into GIS Pro, topographic contour lines were created with the Contour tool. The contour interval was set to 10 m and the base contour was set to a rounded minimum elevation of the data tile.
​
- Before making a hillshade, the 10 m NED data had to be changed from a national geographic coordinate system to a local projected coordinate system. This was done with the Project Raster tool using the Bilinear Interpolation resampling technique.
​
- The Hillshade tool was then used to make hillshades for each elevation raster.
Hillshade Comparison
0.5 m LiDAR (top)
The top hillshade was created with the the smallest resolution raster. As such, this hillshade has the most detail, with rock outcrops, gullies, channels and other slope features easily distinguishable. However, this comes at the cost of small tile size. Each tile only covers a 2000 m by 2000 m area, meaning numerous tiles will need to be analyzed for projects of larger scope.
5 m Auto-Correlated (middle)
The middle hillshade has an intermediate resolution, but the data is created with images from the National Agriculture Imagery Program. Issues arise with this data source because photographs are taken during peak crop season, leading to erroneous elevation undulations on the ground surface anywhere vegetation is present. However, the area covered by a single tile is much larger than that of the LiDAR data, and elevation abnormalities are not as noticeable at larger scales.
10 m National Elevation Data (bottom)
The bottom hillshade was made from the coarsest resolution raster and therefore loses much of the finer detail of the slope features. However, unlike the 5 m auto-corrected data, this data source preserves the larger outcrops, channels and general profile of the slopes. Also, the area covered by each data tile is the largest of the three, meaning file management and analysis procedures are the easiest with this data type.
Contour Comparison
In this comparison, the contours of the 10 m NED data and the 5 m auto-correlated data are compared to the contours of the 0.5 m LiDAR data. In both maps, LiDAR contours are red and the contours of the data being compared are black.
​
The top map shows the 10 m NED contours (black) overlaid onto the 0.5 m LiDAR contours (red). The lack of visible red contours in this comparison means the contours of the two datasets align fairly well. While the 10 m resolution has much less detail, it creates the same accurate slope patterns as the LiDAR data.
The bottom map shows the 5 m auto-correlated contours (black) overlaid onto the 0.5 LiDAR contours (red). The abundance of visible red LiDAR contours indicates how poorly the contours from the two datasets align. Unlike the LiDAR data, there is poor data adjustment for vegetation in the auto-correlated dataset, causing erroneous elevation values.
In the scenario of this exercise, the LiDAR dataset is the most effective. LiDAR data at a 0.5 m resolution has ample detail to identify overused roads/trails, and locate small regions of suitable habitat. While the data tiles are small and file size can increase quickly with LiDAR data, the region of interest is small enough for this project that LiDAR would be manageable.