AmericaView's national applied research goals are broad and generally accomplished through the Research Committee. The Research Committee is one of several special interest groups of the AmericaView national consortium. The committee was established to effectively organize and support StateView consortia members who are particularly interested in applied remote sensing research. Because applied research can take many forms, the committee collaborates on various projects driven by state and regional applied research needs.
The Research Committee exists to support research efforts in the areas of objective-based analysis, ecosystem science, classification meta-analysis, and cyber infrastructure. Specifically, the Research Committee provides a forum for:
The AmericaView Classification Methods Accuracy Comparison (ACMAC) project.
ACMAC is an AmericaView project lead by Dr. Rick Lawrence at Montana State University as part of MontanaView. The purpose of ACMAC is to provide the tools and infrastructure to conduct rigorous comparisons of classification algorithms. ACMAC is implemented in the R statistical programming language. Two tools are provided by ACMAC to support classification algorithm comparisons: (1) a collection of 30 datasets, each including separate training and validation data, in a series of comma-delimited (csv) files and (2) sample R code for automatically analyzing the 30 dataset collection (or other user provided datasets) using Random Forest (the Random Forest algorithm can readily be replaced with most other classification algorithms available in R). To download a zip file containing these ACMAC tools, click here.
StateViews cooperate in a range of applied research projects depending on common needs and funding from AmericaView and other organizations and agencies. Organized around the need to better understand the structure and function of Earth's ecosystems and the sustainability of natural resources, StateViews identify, develop, and seek funding for cooperative projects that benefit society. Data used in these projects, such as Landsat and MODIS, is designed to provide wide coverage at moderate resolutions, and is often best suited to landscape-scale and regional applications.
At the 2010 Fall Technical Meeting, the Research Committee created four working groups to advance the research interests of AmericaView. The four groups and their lead contact person are:
The CI research group is interested in pursuing improved cyberinfrastructure elements at and between state view sites to improve collaboration potential, the ability to share large geospatial datasets, and leverage high performance computer clusters when available. To accomplish this, state view investigators are asked to self-identify their interest in participating in an AmericaView CI network and assist in writing a grant proposal(s) to deploy this technology.
Contact: Keith Weber, Idaho State University firstname.lastname@example.org or (208) 282-2757
Contact: Rick Lawrence, Montana Statue University email@example.com or (406) 994-5409
The quality assessment group focuses on an issue for decision-makers in which a land cover dataset may be very accurate by peer-reviewed literature standards (>85%) yet be nearly completely incomprehensible to the end user. The goal of the group is to develop a method to assess the quality of the data produced. The current approach is to do manual corrections. For example, in a recent tree canopy mapping project in Philadelphia, the University of Vermont’s QA/QC team made over 30,000 corrections. The resulting land cover layer looked much better for the decision-makers, however, neither the estimate of tree canopy nor the accuracy were statistically improved.
Contact: Seeking a lead for this group
The Object-Based Image Analysis (OBIA) group is dedicated to advancing the use of object-based/object-oriented techniques for extracting information from remotely sensed data within the AmericaView community.
Contact: Jarlath O’Neil-Dunne, University of Vermont firstname.lastname@example.org or (802) 656-3324
Working through our StateView lead institutions, AmericaView supports a wide range of undergraduate research activities in cooperation with university scientists, government agencies, non-profit organizations, and the private sector. StateViews partners, such as local planning agencies, state departments of natural resources, and federal partners including the National Park Service, U.S. Forest Service, and Bureau of Land Management provide a variety of exciting opportunities for undergraduate students to engage in applied geospatial projects. StateViews also support undergraduate students through scholarships, research assistantships, internships, and travel grants for field work and conference presentations. Examples of student research projects include mapping forest ice storm damage, analyzing changes in floodplain land use, assessing wildfire damage, quantifying agricultural productivity, detecting and mapping invasive species, and monitoring rangeland conditions using satellite imagery and aerial photographs. These educational and professional experiences increase students’ geospatial knowledge and skills and significantly enhance their graduate school and employment potential.
Graduate research occurs in cooperation with StateView academic scientists and their partners who direct remote sensing research laboratories at their respective colleges and universities. Most of our StateView Principle Investigators are tenured academic faculty members who work on a wide variety of theoretical and applied research sponsored by the federal and state government, foundations, and other organizations. Graduate students are critical to the research missions and gain valuable experience as a central part of their academic training.