You are here
Brett Lord-Castillo's M.S. Degree, 2007
Arc Marine as a Spatial Data Infrastructure: A Marine Data Model Case Study in Whale Tracking by Satellite Telemetry
Master of Science, Geography, Oregon State University, Fall 2007
Emphasis in Geographic Information Science, Minor in Marine Resource Management
Graduate committee: D. Wright, B. Mate, A.J. Kimerling, R. Reitsma
Dept of Geosciences, Oregon State Univ
Corvallis, OR 97331-5506
The Arc Marine data model is a generalized template to guide the implementation of geographic information systems (GIS) projects in the marine environment. Arc Marine developed out of a collaborative process involving research and industry shareholders in coastal and marine research. This template models and attempts to standardize common marine data types to facilitate data sharing and analytical tool development. The next step in the development of Arc Marine is adaptation to the problems of specific research communities, and specific programs, under the broad umbrella of coastal and marine research by community specific customization of Arc Marine. In this study, Arc Marine was customized from its core model to fit the research goals of the whale satellite telemetry tagging program of the Oregon State University Marine Mammal Institute (MMI). This customization serves as a case study of the ability of Arc Marine to achieve its six primary objectives in the context of the marine animal tracking community. These objectives are: 1) to create a common model for assembling, managing, and publishing tracking data sets; 2) to produce, share, and exchange these tracking data in a similar format and standard structure; 3) to provide a unified approach for community software developers extending the capabilities of ArcGIS; 4) to extend the power of marine geospatial analysis through a framework for incorporating object-oriented behaviors and for dealing with scale dependencies; 5) to provide a mechanism for the implementation of data content standards; and 6) to aid researchers in a fuller understanding of object-oriented GISs and the power of advanced spatial data structures. The primary question examined in this thesis is:
How can the Arc Marine data model be customized to best meet the research objectives of the OSU MMI and the marine mammal tracking community, in order to explore the relationship of the distribution and movement of endangered marine mammal species to underlying physical and biological oceanographic processes?
The MMI customization of Arc Marine is focused on the use of Argos satellite telemetry tagging. The customized database schema was described in Universal Markup Language by modification of the core Arc Marine data model in Microsoft Visio 2003 and implemented as an ArcGIS 9.2 geodatabase (personal, file, and ArcSDE). Tool development and scripting were carried out predominantly in Python 2.4.
The two major schema modifications of the MMI customization were the implementation of the Animal and AnimalEvent object classes. The Animal class is a subclass of Vehicle and models the tagged animal as a tracked instrument platform carrying an array of sensors to measure its environment. The AnimalEvent class represents interactions in time between the Animal and an open-ended range of event types including field observations, tagging, sensor measurements, and satellite geolocating. A programming interface is described for AnimalEvent (AnimalEventUI) and the InstantaneousPoint feature class (InstantaneousPointUI) that represents observed animal locations. Further customization came through the development of a comprehensive development framework for animal tracking in Arc Marine. This framework implements front-end analysis tools through Python scripting, ArcGIS extensions, or standalone applications developed in VB.NET. Back-end database loading is implemented in Python through the ArcGIS geoprocessing object and the DB-API 2.0 database abstraction layer.
Through a description of the multidimensional data cube model of Arc Marine, Arc Marine and the MMI customization are demonstrated to be foundation schemas for a relational database management system (RDBMS), object relational database management system (ORDBMS), or enterprise spatial data warehouse. This modeling method shows that Arc Marine is built upon atomic measures (scalar quantities, vector quantities, points, lines, and polygons) that are described by related dimensional tables (such as time, data parameters, tagged animal, or species) and concept hierarchies of different levels generalization (for example, tag Download Thesis (10.6 Mb PDF file)
Also available in the ScholarsArchive@OSU permanent collection
Publication of M.S. thesis in Transactions in GIS: Lord-Castillo, B., Wright, D.J., Mate, B., and Follett, T., A customization of the Arc Marine data model to support whale tracking via satellite telemetry, Transactions in GIS, 13(s1): 63-83.
This is the authors' version of the work. It is posted here by permission of Transactions in GIS and Wiley-Blackwell Publishers for academic use, not for mass redistribution or for commercial purposes. The definitive version is available at www.blackwell-synergy.com.