GPT: Scalable geophysical survey data analysis and visualisation

Summary Geophysical survey datasets used for resource exploration and detection are large in volume, dense in time and space, and have many dimensions. We present the Geophysics Processing Toolkit (GPT), an application for processing geophysical survey data prior to interpretation and inversion. Initially developed as a processing toolkit for airborne electromagnetic (AEM) data, our application can be extended to beyond electromagnetics processing and inversion to incorporate multiple geophysical datasets such as gravity and magnetics. Interactive visualisation and signal processing tools make the process more efficient. We have developed the GPT using a cross-platform technology stack designed to work in a containerized environment in a Cloud and accessible via a web browser. This approach makes it intrinsically scalable and cost-efficient to operate. The toolkit architecture allows for a greater degree of extensibility offering a range of interactive visualisations and integration of a suite of signal processing tools for noise detection and removal. Modern visualisation technologies allow the software to run on a standard workstation or a laptop while efficiently delegating all computationally-intensive tasks to the accompanying Cloud-based processing unit. Decoupling of visualisation components from cloud compute and storage nodes allows on-the-fly substitution of analytical codes, e.g. forward modelling, inversions. This brings greater flexibility in experimental research through the ability to apply various numerical methods and compare results via elaborate visualisations and through the application of statistical methods. Delegation of computing tasks and storage requirements to a third-party cloud provider (a) minimizes procurement and maintenance costs of computing/storage infrastructure and (b) eliminates clients’ privacy concerns as data are stored and processed in an isolated cloud environment.