Large scale processing of landsat data on various software platforms
George Hamer, Jason Werpy · 2014
This paper will propose, plan, model, and investigate the feasibility and cost of utilizing cloud computing (aka Software Platforms) to perform processes on very large sets of binary data. In this case the binary data is the Landsat TM and ETM+ sensor archive. Due to the costs of running trials against processing large datasets, a model will be created that can be utilized for predicting cloud performance and cloud costs. This model will be evaluated against real world examples and will be utilized to measure performance of an existing private cloud implementation of a distributed processing system. The determination of the feasibility of the public cloud for this kind of computing will influence future activities undertaken by the owners of large binary datasets for their processing.