Friday, June 1, 2012

Satellite images monitor vegetation response in Wyoming rangelands

Article originally published in Reflections a publication of the UW College of Agriculture and Natural Resources (Publication date: June 2012; pages 26-29)

Access the issue online at: http://www.uwyo.edu/uwexpstn/publications/reflections/2012/reflections-2012-web.pdf (5.5 MB)


* * * * * * * * * * *

Satellite images monitor vegetation response in Wyoming rangelands

Students capitalize on no-cost Landsat images to map rangeland vegetation


Rangeland vegetation responds to both environmental variations (e.g., drought, precipitation) and human interventions (e.g., grazing management). These responses can be rapid (e.g., wildfire), seasonal (e.g., grazing) or slow (e.g., invasion of non-native species).

Ranchers, land managers, and planners need periodic information about where and how changes are occurring. Collecting change information through field surveys is expensive and time-consuming and often does not provide a complete picture because only a portion of the area is sampled.

Monitor and Map Vegetation Responses

Remotely sensed images can be used for monitoring and mapping vegetation response to environmental and human influences. Since information derived from remotely sensed data is not always similar to data collected in the field, researchers have developed indices and metrics that can provide valuable insights for monitoring and mapping natural resources. With the availability of more types of remotely sensed data, new indices and metrics are also being developed for monitoring vegetation in rangelands and elsewhere.

University of Wyoming students enrolled in the Remote Sensing for Agricultural Management course are trained in the use of remotely sensed data for monitoring and mapping vegetation in rangelands and forests. In addition to learning remote sensing concepts, students are required to complete an inquiry-based research project of their choice involving use of remotely sensed data for answering questions pertaining to natural resource management.

Images Cover 100 x 100 miles

Most students use Landsat data provided by the US Geological Survey (USGS) through the GloVis website for monitoring or mapping vegetation response in rangelands, forests, and agricultural fields. Each image (referred to as a scene) covers roughly 100 miles x 100 miles on the ground and contains information in the visible and infrared regions of the electromagnetic spectrum. Landsat images can be used for mapping present and past (since 1972) conditions of rangelands, forests, crop fields, lakes, and other earth surface features (Figure 1).

Figure 1: Landsat images acquired in 1999, 2002, 2005, and 2011 for an area adjacent to Keyhole Reservoir near Moorcroft demonstrate their utility for monitoring changes in natural resources such as vegetation rangeland and riparian zones and water. Intensity of green color can be associated with high (bright) and low (dark) vigor. Vegetation growth is low (few areas in green) during drought years (2002 and 2005) and relatively higher in normal and wet years (1999 and 2011). The surface area of the reservoir also fluctuates between these years as a result of natural and human influences.

Although Landsat satellites have collected data since 1972, their widespread use was somewhat limited because of the high cost associated with acquiring those images. However, since December 2008, USGS is providing all Landsat images at no-cost to users, which has generated an unprecedented opportunity for students enrolled in this class.

Students interested in rangeland management have obtained numerous Landsat images for monitoring vegetation growth and mapping changes in vegetation conditions over time. Almost all students selected ranches either their parents owned or where the students worked in the summer.

Monitors Rangeland Allotments

Selecting familiar study sites provides a unique opportunity for evaluating the utility of Landsat image information. Clint Beiermann, an agroecology major, analyzed vegetation responses in three rangeland allotments where he worked one summer. Analyses of Landsat images from May 17, June 2, June 18, and July 4, 2007, (Figure 2) showed different vegetation growth rates between allotments. Grasses and forbs grew early in the season in one type of allotment while growth was delayed in another. During the end of the growing season, differences between allotments were minimal. He concluded Landsat images can be used for routine monitoring of rangeland vegetation.

Figure 2: Landsat images acquired from May through July 2007 were used for monitoring vegetation (green) growth in allotments under different management regimes. Water in the ponds appears black, and allotment boundaries (white) were superimposed on the image. Clouds and their shadows (bottom left of July 7 image below) could limit the use of Landsat data.

Studies Grazing Effects

Having image data for the entire study area enabled identification of patterns and anomalies (areas of high, medium, and low growth). Brandon Greet, also an agroecology major, monitored the grazing impact on a ranch in the Big Horn Mountains using Landsat images from June 29, August 25, and September 17, 2008. He categorized vegetation in this rangeland into high, medium, and low vigor  and found the area of the low-vigor vegetation class increased during the growing season. Further, he noticed vegetation re-growth in September in some of the areas classified as bare ground in the August image. He generated maps that showed where these changes have occurred, which could help ranchers improve grazing by focusing cattle on under-utilized areas.

Most students (and owners of the ranches where they worked) had a general sense of the changes in vegetation composition and patterns. Visualizing those changes in Landsat images acquired several years apart sheds new light on the types and magnitude.

Researches Forage Availability

Matthew Allshouse (a rangeland and ecology management major) monitored forage availability and riparian vegetation at a ranch near Laramie (Figure 3). Over the last several years, this ranch shifted from intensive grazing to multiple-objective management. Ranch owners have implemented a rest rotation grazing regime, fenced off riparian areas, and also built a reservoir to improve riparian habitat. Allshouse obtained a Landsat image from 2001 to establish baseline vegetation conditions and used the 2006 and 2011 images for monitoring and mapping how vegetation responded to the changes in management practices.

Students gained experience processing remotely sensed data and extracting information pertinent to their research questions by working independently on
their projects. They evaluated several vegetation indices derived from Landsat
data and determined the suitability for monitoring and mapping rangeland vegetation. Over the past few years, students enrolled in this course have used Landsat and other remotely sensed data for monitoring natural resources in Wyoming and other states. Several students have also presented their research work at the Wyoming Undergraduate Research Days.

As more and more students take advantage of the no-cost Landsat data, findings from their research projects may benefit the community by providing insights about the utility and limitations of Landsat data for natural resource monitoring in Wyoming and elsewhere.