NOTE: This item is not available outside the Texas A&M University network. Texas A&M affiliated users who are off campus can access the item through NetID and password authentication or by using TAMU VPN. Non-affiliated individuals should request a copy through their local library's interlibrary loan service.
The use of TM imagery to monitor the general health of coastal marshes
Abstract
General marsh-health monitoring without field observations was demonstrated to be feasible using TM (Thematic Mapper) Imagery. As monitoring is very expensive and time-consuming because of field work and personnel required, the procedure proposed should cut costs. Areas of declining health can be detected, at little expense; therefore limited budgets can be used more efficiently, by channeling resources directly to locations where they are needed. Preprocessed imagery (prepared by using training sites from the National Wetlands Inventory Map for the area) of a 1983 Summer scene and a 1990 Winter scene were overlaid for change-detection to be interpreted into general marsh-health classes. A new procedure was devised and compared to the traditional 2-layer overlay procedure for change detection. It involved using a supervised and an unsupervised classification of each image to be studied. Therefore, a 4-layer overlay is prepared. The 4-layer overlay has superior structure for change detection, in that two varied sources of information, offer more information to be processed in a GIS. Errors are modelled out of the change-detection process; thus more accurate and reliable assessments concerning marsh health are possible. The traditional 2-layer overlay procedure pern-tits no truly convincing means to correct errors. When classifying an image to be used as a GIS layer in the 2-layer over-Jay, each pixel is given a classification that may be correct or not. There is information to distinguish between true-change and no-change situations, from change and no-change errors. As was the case in this study, the processed change-detection map difficult conbecomes speckled with error. The 2-layer overlay could not cope with the conditions of the test. These were: (1) seasonal differences, 2) no ground-truthed image, 3) no field work to guide the image processing; 4) no information used concerning alterations in the study area, except that which could be extracted from the images and the NWI Map, and 5) excessive care by the image-processing operator to capture any degradation and loss, causing unintentionally-introduced error, by unrealistically shifting the water regime to a much drier one. The 4-layer overlay procedure fared well under these difficult conditions. The classes were much cleaner (less speckled with error). Interpretations concerning marsh health were possible. The interpretations given in this study are visual. Area quantification of marsh-health classes do not seem likely under the conditions stipulated, but less harsh conditions may give promising results. The field-corrected, 4-layer overlay appears to be of such clarity to warrant quantitative investigations to evaluate its full potential (not the subject of this study). Practically all areas of digital map making can benefit from the concepts involved in the procedure proposed. Multi-layer processing of TM imagery permits classifications to the most logical mapping unit, and eliminates error. The results are much superior to those of 2-layer processing. Applying the DeSousa procedure, or modifications thereof to suit specific requirements, can greatly enhance the quality of most digitally produced maps.
Description
Due to the character of the original source materials and the nature of batch digitization, quality control issues may be present in this document. Please report any quality issues you encounter to digital@library.tamu.edu, referencing the URI of the item.Includes bibliographical references.
Collections
Citation
De Sousa, David Alexandre (1994). The use of TM imagery to monitor the general health of coastal marshes. Master's thesis, Texas A&M University. Available electronically from https : / /hdl .handle .net /1969 .1 /ETD -TAMU -1994 -THESIS -D467.
Request Open Access
This item and its contents are restricted. If this is your thesis or dissertation, you can make it open-access. This will allow all visitors to view the contents of the thesis.