Submitted to: Rangelands
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 6/6/2011
Publication Date: 8/22/2011
Publication URL: http://handle.nal.usda.gov/10113/57952
Citation: Schrader, T.S., Duniway, M.C. 2011. Image interpreter tool: An ArcGIS tool for estimating vegetation cover from high-resolution imagery. Rangelands. 33(4):35-40. Interpretive Summary: This article describes an automated, digital image interpretation system that includes calibration and quality control features and can be implemented by anyone with a basic knowledge of ArcGIS. This system is designed to allow land managers to quickly (1) calibrate observers, (2) and collect, organize and archive information on vegetation cover and composition, and ground cover in the plant interspaces. This knowledge transfer system allows remote sensing data on key rangeland indicators to be collected by users at remote locations and potentially can reduce the workload of local land management and monitoring agencies.
Technical Abstract: Land managers need increased temporal and spatial resolution of rangeland assessment and monitoring data. However, with flat or declining land management and monitoring agency budgets, such increases in sampling intensity are unlikely unless new methods can be developed that capture data of key rangeland indicators at a lower cost. Remote sensing techniques have shown promise for collecting plant community composition and ground cover data efficiently. However, many image analysis techniques require software and expertise not always available to field offices. This article describes Image Interpreter Tool (IIT): a series of ArcGIS 9.3 tools and workflow procedures that have been developed to meet this need. The tool and procedures were designed to streamline: 1) the calibration of image interpretation users, and 2) the collection of vegetation and ground-cover types for a study site or project. IIT is distributed as a customized ArcMap document or template with nothing to install, and is compact enough to be used on portable storage devices such as USB thumb drives. IIT can be used by people with little or no GIS experience and reduces recording errors by providing an automated system for attributing data files. IIT mimics point-intercept field-sampling methods using remotely sensed data, “virtual” points along transects, and a simple and intuitive interface-to-estimate cover. Three main cover categories are used: noncanopy (rock, litter, soil, and lichens), herbaceous cover (grass and forbs), and woody canopy (subshrub, shrub, tree, and succulent). Additionally, users can toggle between true color and color-infrared versions of the imagery (assuming four-band imagery is used as a source) with a simple click of a button on the interface. IIT is easy to learn and is designed to facilitate multiple users producing consistent results. IIT is divided into two modules: 1) an observer training and calibration module that includes quality assurance and quality control procedures, and 2) a data collection module.