It was the focus of this manuscript to use passive optical remote sensing to classify the benthic facies of a large study area such as the southeastern Arabian Gulf. Landsat TM and Quickbird sensors were also evaluated for the determination of benthic facies. Spatial distributions were further examined from the classified image to study facies patterns. It was found that Landsat TM sensors could be used to accurately classify benthos of a large study area such as the southeastern Arabian Gulf if sufficient ground control data was available. This was determined by using both the unsupervised and supervised classification techniques in the ENVI 4.1 program. When discussing the issue of scale in relevance to classification for small areas considered in isolation (i.e. Butina Island), the Landsat TM sensor returned classification results comparable to those obtained with a higher spatial resolution (Quickbird sensors). By using the classification results from the southeastern Arabian Gulf, the patch frequency of the facies concluded that patch frequency and area were inversely related, with smaller areas being more common and larger areas rare. The data showed a linear relationship on log-log plots and therefore could be termed a power function. Due to the linear relationship, perhaps patch frequency and area follow a power function.
| Date of Award | Jan 14 2007 |
|---|
| Original language | English |
|---|
| Supervisor | Sam Purkis (Supervisor) & Bernhard Riegl (Advisor) |
|---|