Test application of a neural network model for classification of coastal wetlands vegetation structure using moderate resolution imaging spectro-radiometer (MODIS) data
Abstract
This paper demonstrates the use of a neural network model in classifying Louisiana‟s coastal marshes. Four image maps were produced for each year between 2001 and 2004. Accuracy assessment of the classification has indicated that the neural network techniques using MODIS data achieved an overall accuracy of over 80% (at 0.95 confidence level). Using the classified images change detection was performed to determine the loss and gain of four marsh types; saline marsh, brackish marsh, intermediate marsh, and, fresh water marsh found in the south eastern coastal areas of Louisiana. The greatest gain was in the intermediate marsh, 4.4% of the study area, and the greatest loss was in the saline marsh, 2.7% percent of the study area. The assessment of changes of water salinity levels in various parts of the Barataria bay indicate the introduction of fresh water into the marsh from the Mississippi River has the greatest influence into marsh dynamics in south eastern Louisiana.
Key words: Coastal marshes, classification
Downloads
Published
How to Cite
Issue
Section
License
Authors must agree to abide by the terms of the Copyright Statement as per the licensing information of this journal.
Licensing Information for this journal:
Attribution-Non Commercial (CC BY-NC). This journal License allows others to download works, remix, adapt and build up works non commercially, and new works must also acknowledge other authors and be non commercial.