Google Cloud: Processing large volumes of seismic data in the search of highly correlated waveforms.

Abstract:

Currently, a large amount of data is being generated as a result of an increasing number of seismic instruments both inland and offshore. Thus, detection threshold along the areas where large and destructive earthquake nucleate is drastically increasing. Then, the need for the implementation of an affordable and fast method to process seismic data becomes an urgent necessity. Google cloud is a platform that allows the rapid implementation of multiprocessor software without the need of purchasing, updating and maintaining a dedicated cluster of computers. This drastically reduces the cost of processing data by just purchasing the computing time and processing capacity needed to perform a specific task. Here, we compare the results obtained to identify highly correlated waveforms using a multiprocessor computer and the results obtained using the Google Cloud platform. The proper and rapid identification of waveforms will allow examining large sets of data as they become available within a reasonable amount of time and mainly without the need of actually owing a large capacity computer.