Real-time GNSS Quality in the Network of the Americas

Abstract:

The high-rate (1sps), real-time streaming data of the Network of the Americas (NOTA), part of the NSF funded GAGE program, consists of ~950 continuously operating GNSS stations distributed across western North America, the Caribbean and northern South America. This enhanced data from the NOTA represents a significant step in the evolution from networks originally designed to download 15 second interval GPS observations once per day to generate a 24-hour average position estimate. Each of these higher rate real-time station nodes consists of one of several combinations of receiver and antenna hardware generating one second epoch encapsulated BINEX format records streamed via TCP/IP to a central Network Operations Center at UNAVCO Inc. These data are re-distributed via NTRIP in BINEX and RTCM3 format, and used for precise point position (PPP) estimates. Multiple hazard and remote sensing communities, including Earthquake/Tsunami warning and space weather modeling, have increased their inclusion and dependence on these high-rate real-time and near real-time data products. On-going investigations categorize and characterize the performance of these stations, comparing metrics including data product latencies, completeness, receiver signal to noise ratios, estimated signal multi-path and cycle slips. These station performance metrics are considered relative to the entire NOTA streaming network, relative to their individual lower rate metrics, and relative to some of the available defined specifications of the end-users. We continue these investigations using temporal, spatial, and constellation dependencies, internally motivated by the need for improved station health monitoring for optimal operations and externally targeting improved network health visibility and station integrity for end users.

In addition, these new high-rate low-latency data sets are encouraging network operators to consider new and innovative ways to handle these data, including Kafka-type data management systems combined with modern postgres databases optimized for time-series data sets. Besides enhanced data management efficiency, integrating these modern messaging and database stacks could enable novel research across multiple domains in a variety of fields including AI/ML investigations.