The wonders of LiDAR never cease to amaze. With what originates as a laser scan of the Earth, one can create digital elevation models (DEMs) of the Earth's surface using geospatial software. DEMs are commonly used as the raster data type to model a continuous elevation surface from discrete LiDAR points. The high density laser scan points allow for accurate mapping of Earth's features; in this case we are focusing on streams. While this is based on an actual case, the situation is confidential. The data illustrated was not part of any private work; it was generated from public data.

Special techniques and procedures that examine characteristics of the DEM allow for some pretty cool findings otherwise not apparent in an aerial photograph, or even the DEM.

Worth noting, the stream being illustrated to the left appears historically ditched.

Problem

Regulations required all streams, as other surface waters, to be delineated for jurisdiction. Requirements were to use global positioning systems (GPS) to delineate the stream banks. There were about 13 miles of stream banks to delineate, and time was not a friend. How could we use GIS and supporting data to update this process with modern technology?

Field Mapping of Stream Feature

The tradiditional way to map a stream bank or other surface water, is to use GPS hardware to collect points along the stream bank where there were significant bends and angles. The points were later connected in GIS to build stream bank limits as jurisdictional areas for other surface water.

Feature Extraction

One of the remarkable aspects of LiDAR is being able to take a laser scan of the Earth's surface with relatively high density (~1m post spacing) and subsequently extract features of the earth. Feature extraction comes from understanding how the data originates, how it is converted into a DEM, and then how functions applied to the DEM can derive more geographic data. A few examples of feature extraction are for: ditches, streams, roads, wetlands as topographic depressions, drainage networks, drainage areas and seepage slopes.

LiDAR-Based Methodology

Using data at ~1m post spacing was dense enough to map stream features and their banks. Point data was extracted from the LiDAR point cloud and used to create a DEM at one meter resolution. An astounding 90% of the effort was able to be done using the DEM and special GIS analysis and techniques to create vector lines modeling stream bank limits. This method was credible and proved accurate enough such that the Agency dropped their requirements for GPS points to define stream banks. The LiDAR based approach proved equal or better to previous technology and obliterates any comparison to doing the work manually with GPS point collection.

Shortcomings of Methodology

While maybe not evident here, tree and shrub canopy can affect the ability of LiDAR scanners to get high density ground points. Although newer LiDAR technology is incredibly better, but In these dense vegetation areas with obscured points, GPS teams can supplement the LiDAR data with GPS points from the field. Obvious caution applied to collecting GPS data under canopy as it can disrupt the signal. When this dense vegetation blocks LiDAR points from hitting the ground, fewer points are available to interpolate a DEM with. Fewer points mean reduced ability to capture accurate sharp changes in the elevation, thus creating a wider delineation output than should be. This is no problem so long as some field verification is done, especially in dense vegetation areas.

It is worth noting in the example above, that under heavy canopy (northern segment) the stream bank appears to deteriorate and widen. Some may recommend using breaklines that can be provided with data, but that is another topic to discuss. Banks under canopy areas appear wider than the open field areas of the stream bank model. The remarkable aspect is that the feature is still exposed under canopy, dispelling the myth that LiDAR cannot penetrate canopy cover.