Using Multispectral Imaging for Ecological Observations
- Saturday, 01 December 2012
With an airborne camera capable of making precise and detailed ecological observations, biologists at Applied Ecological Services (Brodhead, WI) are bringing satellite imagery closer to earth.
After years of using satellite imagery for larger landscape-scale applications, AES has acquired a new high-resolution multispectral camera for imaging and mapping ecological projects. Instead of using a high-flying fast plane with a large format camera, AES and its partner Ayres & Associates (Madison, WI) have opted for a plane that flies low and slow over the ground, even beneath cloud cover, to obtain ecologically relevant imagery.
For AES, the timing of imagery to capture data on dynamic ecological processes is important. So is the flexibility to capture imagery on an as-needed basis — for example, when deciduous trees have lost their leaves but invasive, exotic common buckthorn is still dark green. Flexibility is often difficult for other aerial photography vendors who are more accustomed to imaging for engineering or infrastructure purposes that are not as sensitive to timing issues.
The Leica RCD30 camera acquires imagery as resolved as 2-inch, on-ground pixel size because of its fast shutter and AES’ slow-flying plane (see Figure 1). The digital mapping camera offers four spectral bands (red, green, blue, and infrared) that are capable of achieving engineer- standard mapping accuracy specifications in association with both vertical and horizontal measures.
Using the Infrared Band
The infrared band enables specific applications associated with the study of vegetation, and it offers a unique look at the “greenness” or “productivity” of vegetation. The infrared band is receptive to capturing reflectance associated with the amount and type of chlorophyll A or B pigments present in the tissue or cells of plants. This sensitivity, associated with detecting chlorophyll in vegetation, gives scientists at AES a unique understanding of ecological interactions that might otherwise go undetected (see Figure 2).
Plant productivity is a measure of the condition, vigor, moisture, and health of a plant. Identifying where and how that productive plant tissue is distributed on the landscape can be used to measure ecosystem conditions, including crop productivity, biomass volumes, detection and measurement of pest or disease impacts, and the mapping of vegetation community types. In some cases, it is also used to identify and map specific plant species including various grasses, sedges, forbs, or trees, as well as different aquatic patterns associated with algae growth.
After aerial images are collected, they are typically brought into high-powered software programs designed to seamlessly “mosaick” multiple images into a single image or set of image tiles that cover the entire area of interest. During this process, imagery is ortho-rectified, which is a method that adjusts the imagery to account for the topographic relief of the landscape. In addition, imagery is color balanced and histograms are normalized, which is critical for spectral analysis and remote sensing applications.
Once the imagery has been processed, it becomes an ecological tool for a team of trained remote-sensing and ecological professionals. By combining the understanding of advanced mathematics, statistics, and computer science with ecological expertise and in-depth knowledge in vegetation seasonality, senesence, and phenology, AES discovers spatial patterns that become the blueprints to understanding landscape change. Features are mapped using variables such as texture, size, shape, and reflectance. Analysis of imaged features offers spatial solutions to complex ecological trends, and informs critical decisions associated with landscape restoration.
Mapping an Invasive Shrub
One recent AES project involved the mapping of a woody invasive shrub, bush honeysuckle, which has colonized large forested areas of southern Illinois. The mapping of this 100-square-mile area, which includes Shawnee National Forest, was initiated with the (strategically timed) aerial flight, followed by on-theground work by AES ecologists. Ecologists took measurements of plant conditions, and then used survey-quality GPS-location to find dozens of the individual honeysuckle shrubs.