Conservation, restoration, and mitigation efforts around the world increasingly rely on large volumes of data. Nowadays, the problem is less about having lots of data, and more about how to acquire it cost-effectively and how to make efficient use of it. Yet public agencies and nonprofits typically don’t include data management in their lists of core strengths; as a result, they often don’t readily leverage related paradigm shifts. We hope this article exposes you to some new tech that can help you accomplish your mission better, faster, cheaper.
It’s a bird. It’s a plane. It’s a toy. It’s a super data capture device! Unmanned Aerial Vehicles (UAV), or drones, are creating more than just a buzz—they’re forcing a paradigm shift in the way data are collected. Once a technology that was largely confined to military uses, UAVs have flown the coop and are featured in promising civilian applications that serve people in the fields of agriculture, forestry, ecology, geology, energy, and engineering.
The cost of off-the-shelf UAVs are declining, and the capabilities of the sensors available that can attach to them are increasing. Today, you can purchase a very capable UAV platform that includes a high-resolution digital video-enabled camera for under $1,500. As with other disruptive technologies, lower cost and higher quality drives adoption and investments in innovative services and applications that would have previously been cost-prohibitive.
Being perpetually curious and eager to learn how these benefits can be applied to environmental sciences, Sitka Technology Group conducted an experiment this summer on the Yankee Fork River in Idaho. Our objective was to use an inexpensive, commercially-available UAV platform to capture high-resolution images and construct 2D imagery and 3D models that might be used to assess the current condition of habitat characteristics at the site. Along with our FAA-licensed and section 333-authorized operator, we planned our mission and were excited to get into the field.
We learned a lot during this trip. The biggest lesson: site layout and flight mission planning is absolutely essential. We started by creating GIS files that delineated the extent of the area we planned to survey and the ideal locations of ground control points (clearly visible markers placed on the ground at precise GPS locations). We used a surveying technology called Real Time Kinematic (RTK) to pinpoint the ground control points to within 2 centimeters of accuracy.
When the imagery from the UAV is harvested, the ground control points are visible in the imagery, enabling sophisticated image processing software to align the imagery with real-earth coordinates so that the resulting 2D and 3D products are uniform in scale (geekily speaking, orthographically-rectified) and can be used in popular tools like ArcGIS or Google Earth. We are currently engaged in techniques to segment and classify objects visible in the 2D imagery we gathered and are integrating UAV-captured data into our flagship monitoring program platform, GeoOptix™.
Counting things that matter. Since the 2D imagery is mapped to real world coordinates, it is possible to compute the areas of features on the surface (e.g., surface area of water, vegetation, etc.) as well as to count objects of interest that appear in the image. These objects can be anything from the number of cobbles in a stretch of stream, or the number of shrubs in an area. It’s one thing to count dollars spent, acres treated, or citizens contacted (and believe me, we help folks do a lot of that). It’s a whole other thing to be able to accurately and inexpensively count cobbles and shrubs—the things that spawning salmon and nesting sage grouse depend on to survive.
With ground sampling distances of 2 cm, the technology now exists to be able to rapidly count and classify relatively small features on landscapes over large areas—a capability not possible just a few short years ago.
Detecting changes in critical ecosystems. The 3D data products are used to produce elevation models, which feed into a vast array of geomorphic and hydraulic modeling efforts. Considered over time and combined with sound science, these products offer the ability to detect relatively small changes in critical ecosystems, leading to a better understanding of how our world is changing and what we can do to mitigate for harmful anthropogenic and natural effects.
We will be going into greater depth on the 3D aspects of this experiment at the Western Forestry and Conservation Association’s Field Technology Conference in Portland OR, November 18-19. We will have some conference materials posted on our website after that, if you’re interested in learning more and checking out some of the results.
Our experimental flight certainly wasn’t a model of aviation perfection, but we learned a lot and generated new information at a mind-blowing resolution of one point per 2 cm of surface area. While the datasets generated by this technology are very large and consume extensive amounts of CPU time to distill down into meaningful, actionable information, the advent of inexpensive cloud computing engines and vast storage capabilities make it possible for most organizations to reap the benefits. It is always exciting to see benefits formerly enjoyed only by the largest of institutions become available to environmentally-based nonprofits who must make every dollar count.