Saildrone Data Handling: Workflow from Drone to Desktop

Although the model of leasing an observing platform is not new, the concept of retrieving and delivering data to NOAA scientists by a third party is unconventional. In 2014, the Pacific Marine Environmental Laboratory and Saildrone Inc. entered into a Cooperative Research and Development Agreement to advance an unmanned surface vehicle, a Saildrone. Saildrone, in collaboration with PMEL scientists and engineers, furnishes these platforms with an array of up to 16 mission specific high quality sensors. PMEL scientists have access to near real-time data feeds as the drones are underway, while the unprocessed high resolution data are delivered after it is manually retrieved at the end of each mission. Over this three-year period, the handling of real-time data has evolved to a process where the data flows from the platform to the scientists using a series of data and network services.

Since 2017, the PMEL Science Data Integration Group worked with scientists to facilitate seamless data transfer to internal systems, the use of standards-based data formats, and the configuration of services that make the real-time data easily accessible and compatible with commonly used applications. Fine-tuning of the dataflow began with the data transfer format. Data are landed at PMEL via ftp as NetCDF files based on PMEL-developed templates. ERDDAP and Live Access Server (LAS) data servers were configured to facilitate data visualizations and interoperable data access. The placement of meteorological observations onto the GTS began with the 2017 missions, benefiting from an infrastructure developed for a GTS pilot project. 18 Eighteen Saildrone missions have now used this real time data delivery framework.

This presentation will cover details specific to the configuration of well established data services to facilitate the smooth data flow from the autonomous drone to the scientists’ desktop. Considerations for the reporting frequency and real time message delivery detail will also be discussed. We will also provide information on the mechanism used to place these data on the GTS.