Show simple item record

dc.contributor.advisorMylroie, John
dc.contributor.authorHo, Hung Chak
dc.date2012
dc.date.accessioned2020-06-30T17:04:17Z
dc.date.available2020-06-30T17:04:17Z
dc.identifier.urihttps://hdl.handle.net/11668/17930
dc.description.abstractThis project developed a series of spatial models to classify the island karst landforms and predict the island karst feature distribution. Spatial models with unsupervised classified images, and fuzzy-based spatial models were used in this study. Forecasting verification and spatial regressions were used to validate the models. The case study was conducted on San Salvador Island, the Bahamas, a recognized carbonate island with island karst features. Fieldwork data on banana holes on the island were used for model validation. The results showed that most models had accuracy higher than 90%, and were statistically proved that they could be used as predictors of island karst features. Further study may be conducted to solve the Modified Areal Unit Problem (MAUP) in the future.
dc.publisherMississippi State University
dc.subject.lcshKarst--Geographic information systems--Bahamas--San Salvador Island.
dc.subject.lcshGeographic information systems--Bahamas--San Salvador Island.
dc.subject.lcshGeospatial data--Bahamas--San Salvador Island.
dc.subject.otherRemote Sensing
dc.subject.otherSpatial Modeling
dc.subject.otherKarst
dc.subject.otherGIS
dc.titleIsland Karst Classification: Spatial Modeling-Oriented Approach with Multispectral Satellite Imageries
dc.typeThesis
dc.publisher.departmentDepartment of Geosciences
dc.publisher.collegeCollege of Arts & Sciences
dc.date.authorbirth1986
dc.subject.degreeMaster of Science
dc.contributor.committeeMishra, Deepak
dc.contributor.committeeRodgers, John


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record