The ocean science community trend toward data delivery via web services that follow the so-called FAIR data principles (findable, accessible, interoperable, reusable) has greatly assisted our development of real-time ocean forecast systems that assimilate diverse inputs of satellite and coastal in situ data. The further embrace of Climate-Forecast (CF) metadata conventions and the sub-setting facilities in THREDDS and ERDDAP make our software tools for data acquisition easily re-useable for new applications, requiring little more than the redefinition of the bounding box that encompasses the model domain. For some novel data streams such as sensors on lobster traps, fishing trawls and animal tags, the ability of ERDDAP to easily re-serve simple data formats such as Excel as a FAIR web service has further expanded the scope of data inputs.

To monitor the impact and quality of respective data sources on the system, we track the provenance of all observation platforms closely through all steps of the data assimilation. With ERDDAP search and subset capabilities, the quick browse slide sorter, and the versatile RESTful output interface, it is straightforward to construct simple dashboards for monitoring data availability, making model-data comparisons, and exporting data to other tools for more in-depth analysis.

The presentation will introduce the essentials of ERDDAP capabilities, and demonstrate how we deploy these to enable search, subset and export of inputs and outputs to the Rutgers University coastal ocean forecast system, and to browse the system performance in comparison to other models.