Generating space-time composite images for GIS data

• Main Author: Holness, Carolyn
• Format: Dissertation
• Language: English
• Published: University of Cape Town 2014
• Subjects: Information Technology
• Online Access: http://hdl.handle.net/11427/10608

Includes bibliographical references. === For many decades scientists have been collecting marine data from survey boats that spend weeks at sea covering large areas and collecting data at regular points along the journey. With the introduction of remotely sensed satellite images, large amounts of environmental data (e.g. sea surface temperature) is now available to be used in studies on the effect of the ocean on phenomena observed on such journeys. Up until recently, mapping this environmental data from the satellite images has been done by averaging daily images for the duration of the survey to obtain a single image approximating the entire time span and the whole area covered in the survey. The problem with this however, is that in creating these temporal composite images, values are averaged out so that small variations that occur over short time scales are lost. In an effort to overcome this problem, the idea of a spatio-temporal composite image was proposed. This entails using pieces of daily images, where each piece represents the area covered by the survey on the same date, and merging these pieces to create a single image. This allows the original values from the satellite images to be retained which provides a much more realistice mapping of the environment. Satellite images and survey data can be viewed and analysed in a Geographical Information Systems (GIS), however creating these spatio-temporal composite images manually using the software is a time consuming and difficult task. This project investigates whether it is possible to develop an application to automatically create these images. On completion of the project it was found that not only was this possible, but that it also provided the means to create these images more accurately and far quicker than could be done manually and it was also easy to use even without prior GIS experience.