The Land Trust is taking its stewardship monitoring practices to new heights with the help of unmanned aerial vehicles (UAVs).
It’s hard to do anything quickly in the tidal wetlands and Sitka-spruce swamps of Columbia Land Trust’s Kandoll Farm property. Knee-deep in mud on a rainy February day, stewardship team members Ian Sinks, Jeff Malone, and Mitch Attig found that their latest project was no exception. In preparation for a new method of aerial site monitoring, the team set ground control points to establish a series of precise coordinates of latitude, longitude, and elevation throughout the landscape
In 2014, the Land Trust dug three miles of channels, breached a levee, and built mounds of excavated earth to approximate historic conditions and re-establish Sitka-spruce swamp habitat across the site. More than a year later, the team shifted its focus from active restoration to effectiveness monitoring. Was water flowing through the channels restoring wetland conditions as anticipated? Were invasive species such as reed canarygrass outpacing native planting efforts? Had the topography changed after a year of rain and flooding? It’s difficult to answer these questions in a comprehensive and quantitative manner by monitoring 163 acres of challenging terrain on foot, so the Land Trust looked to the sky.
While aerial photography and satellite imagery can both yield valuable results, they have their limitations. Photography captures one angle and one perspective at a time and determining spatial relationships like the height of trees, depth of channels, and species of plants is nearly impossible. In the past, the Land Trust utilized Lidar devices, which mounted from the bottom of a plane, send millions of laser pulses to the earth and measure their rate of return. This data yields precise three-dimensional images of terrain that allows stewardship staff to take measurements such as changes in channel contours or depth with a high degree of accuracy. However, Lidar flights aren’t always cost-effective or easy to orchestrate on smaller sites like Kandoll Farm. With this site in mind, the stewardship team was eager to explore an emerging technology in the field of environmental analytics.
In February the Land Trust partnered with Sitka Technology Group, a Portland-based firm on the leading edge of environmental management software, to monitor a swath of the Kandoll Farm property and the nearby Devil’s Elbow property with an unmanned aerial vehicle (UAV). Automated UAV flights produced a complex digital terrain model, measuring elevation through a dense set of data points while also taking high-resolution photos. The flight over Kandoll Farm covered 163 acres and recorded 1,600 two-dimensional photos that were later stitched together and combined with geospatial data to create a single orthomosaic. In other words, high-resolution aerial photos and geospatial data like distance and surface elevation were integrated to create a hyper-realistic representation of the landscape.
The team at Sitka used the ground control points established by Ian, Jeff, and Mitch to georeference datasets from the UAV flights with the actual project locations. The resulting orthomosaic offers enough detail to show individual thatches of reed canarygrass in locations that are all but inaccessible on foot. This data will serve as a baseline to allow the stewardship team to measure, among other things, changes in the extent of the reed canarygrass over time.
Across two sites, the UAV flights captured nearly 3,000 and 15 gigabytes of compressed imagery across 345 acres. Much of the data in the orthomosaics still need to be analyzed, but at first glance, it’s clear that digital terrain models from UAV flights offer an efficient and accurate way to monitor and measure changes on our lands. We can adapt our management plans in response to our findings, and ultimately work smarter toward our restoration goals. It’s a brave new world for land stewardship professionals, where today’s new technologies are helping restore natural habitats of the pre-settlement past.