|
|
Her initial work will focus on the development of viable methods to classify water types from limited spectral resolution (e.g., SeaWiFS) and hyperspectral (e.g., COIS) remotely sensed ocean colour data. She will explore the utility of both model based and feature based classification approaches, focusing on the development of a hybrid technique which can incorporate both data-based features and model-inverted parameters into the classification. Once basic methods for classifying water type have been established, the focus will be on interpreting the significance of the results for understanding ecological processes. The classification approaches will be refined and adapted specifically for investigating spatial and temporal variability in the Georges Bank / Gulf of Maine area of the northwest Atlantic. This will be accomplished in the context of extensive ecological and hydrographic work in progress as part of the GLOBEC Georges Bank program, as well as this lab’s current research examining in situ optical variability in this region. After developing a viable optical classification scheme for the northwest Atlantic, focus will shift to interpreting the underlying properties which define the classes, and to examining the spatial and temporal variability in water types which can be optically discriminated. Another potential application of these spectral and hyperspectral classification techniques will involve adapting these approaches to the taxonomic classification of phytoplankton blooms occurring in various regions around the globe. This aspect of the work will deal with the subset of water types where the dominant source of variability is a single group of phytoplankton. To achieve this, it will be necessary to compile sea-truth information on bloom occurrences from all available sources. Concurrent ocean colour data will be analysed for the development of comprehensive classification methods to simultaneously delineate several bloom taxa, with the ultimate goal of facilitating the study of these ecologically significant events in the upper ocean.