Role: Co-Principal Investigator
Date: January 2019 - current
This project develops a real-time processing system capable of handling a large mix of sensor observations. The focus of this system is automation of the detection of natural hazard events using machine learning, as the events are occurring. A four-organization collaboration (UNAVCO, University of Colorado, University of Oregon, and Rutgers University) develops a data framework for generalized real-time streaming analytics and machine learning for geoscience and hazards research. This work will support rapid analysis and understanding of data associated with hazardous events (earthquakes, volcanic eruptions, tsunamis).
Earthquakes, tsunamis and volcanoes pose natural hazards on nearly unimaginable scales and compel geoscientists to find new ways to better understand the processes that cause them and to mitigate their effects on population and the built environment. Hundreds of millions of dollars invested by the NSF and other agencies for observing systems on land (EarthScope), in the ocean (Ocean Observatories Initiative-OOI) and in space (ESA Sentinel and the upcoming NASA NiSAR missions) are designed not only to advance scientific discovery but also to enable enhanced early warning and forecasting of short and long-term natural hazards events. The shear volume and complexity of the data from these data streams, coupled with the need to model, analyze and assess hazards in a matter of only moments, makes geophysical applications to hazards early warning a Big Data problem. In the non-academic world, widely adopted and proven open source software, developed by entities such as Google, Facebook, Linkedin and the National Security Agency, is routinely used for massive scale analytics. Due to barriers to data access and lack of expertise, geoscientists have only just begun to adopt some of these powerful new capabilities. GeoSCIFramework will unite computer scientists and geoscientists to develop a data framework for generalized real-time streaming analytics and machine learning for geoscience and hazards research. The architecture will be similar to the so-called SMACK stack (Spark, Mesos, Akka, Cassandra, Kafka) with components such as Kibana/Grafana, ElasticSearch, and NiFi as needed. The framework will be applied to use cases in the Cascadia subduction zone and Yellowstone in the study areas of the science team and where EarthScope and OOI have the greatest concentration of instruments. The architecture pattern will be transportable and scalable, running in a Docker environment on laptops, local clusters and the cloud. Integral to the project will be developing, documenting and training using collaborative online resources such as GitLab, Jupyter Notebooks, and utilizing XSEDE to make larger datasets and computational resources more widely available.
GeoSCIFramework proposes an innovative approach that looks at the world with a fly’s eye perspective where a composite of thousands of harmonized high-rate real-time GNSS, seismic, pressure and other sensors will continuously stream data into an integrated framework and combined with a background of satellite radar time series generated at a 100 meter pixel level across the globe and made available through XSEDE. Trained in this multi-data environment and informed by physical models, machine learning algorithms and spatio-temporal analyses will provide the capability to instantly recognize that a tsunamigenic earthquake has occurred or to identify longer term subtle motions of the earth's surface on previously unrealized scales. This approach is extensible to not just detection and characterization of earthquakes but also to the onset of other geophysical signals like slow-slip events or magmatic intrusion, expanding the potential for new scientific discoveries.
National Science Foundation - Award Number 1835692 - Rutgers University
National Science Foundation - Award Number 1835791 - UNAVCO
National Science Foundation - Award Number 1835566 - University of Colorado
National Science Foundation - Award Number 1835661 - University of Oregon