Return to search

Exploration of Very High Spatial Resolution Data for Vegetation Mapping using Cartographic Ontologies: Identifying Life Forms to Mapping Formations

Vegetation mapping is often considered the process of identifying landscape patterns of individuals or clusters of species or life forms (LF). At the landscape scale, the larger pattern represented by individuals or clusters represents the conceptualization of "vegetation mapping" and can be used as a building block to describe an ecosystem. To represent these building blocks or LF a "common entity (CE)" concept is introduced to represent the components of Formations as described by the National Vegetation Classification (NVC) system. The NVC has established protocols to consistently represent plant communities and promote coordinated management, particularly across jurisdictional boundaries. However, it is not a universal standard and the methods of producing detailed maps of vegetation CE from very high spatial resolution (VHR) remote sensing data are important research questions.This research addressed how best to understand and represent plant cover in arid regions, the most effective methods of mapping vegetation cover using high spatial resolution data, how to assess the accuracy of these maps, and their value in establishing more standardized mapping protocols across ecosystems. Utilizing VHR products from the IKONOS and QuickBird sensors the study focused on the Coronado National Memorial and Chiricahua National Monument in Arizona and Los Ajos and Pinacate - Grand Desierto Biosphere Reserves in México. Individual CE were semi-automatically mapped incorporating spectral, textural and geostatistical variables. The results were evaluated across sensors, study sites, and input variables. In addition, multiple methods of acquiring field data for accuracy assessment were evaluated and then an evaluation was made of a semi-automatic determination of Formation based on CE.The results of the study suggest consistency across study sites using the IKONOSdata. A comparison between VHR products from the same place is feasible but sensor spectral differences may affect which derived bands would improve classification. CE classification procedures were not significantly different across sensors. The overall accuracy obtained for each Park was 59.5% for Chiricahua using QuickBird and 51.9% using IKONOS; at Pinacate 70.0% using IKONOS, and 55.9% for Ajos. Incorporating the geostatistical semi-variogram variables improved CE accuracy for some CE but not all.

Identiferoai:union.ndltd.org:arizona.edu/oai:arizona.openrepository.com:10150/194484
Date January 2009
CreatorsRodriguez-Gallegos, Hugo Benigno
ContributorsMarsh, Stuart E., Marsh, Stuart E., Comrie, Andrew C., Hutchinson, Charles F.
PublisherThe University of Arizona.
Source SetsUniversity of Arizona
LanguageEnglish
Detected LanguageEnglish
Typetext, Electronic Dissertation
RightsCopyright © is held by the author. Digital access to this material is made possible by the University Libraries, University of Arizona. Further transmission, reproduction or presentation (such as public display or performance) of protected items is prohibited except with permission of the author.

Page generated in 0.0017 seconds