Building Trusted Enhanced-Resolution Satellite Passive Microwave Data Sets: Lessons Learned Producing Interoperable Data for Hydrologic and Cryospheric Applications

Abstract:

Beginning in 1978, the satellite passive microwave data record has been a mainstay of remote sensing of the cryosphere, providing twice-daily, all-weather, near-global spatial coverage for monitoring changes in hydrologic and cryospheric parameters. Historical versions of the gridded passive microwave data sets were produced as flat binary files described in human-readable documentation, using polar-aspect map projections with idiosyncracies that proved to be difficult for contemporary software packages to understand. In the spirit of Findable, Accessible, Interoperable, and Reusable (FAIR) data concepts, we have produced a new, modern version of the data that leverages the EASE-Grid 2.0 definition, self-describing data content and provenance conventions, machine-readable geolocation, and transparency in data processing parameters. Data production was planned for automated ingest into the NASA DAAC metadata system, and now comprises more than 100 data years of gridded passive microwave data from multiple sensors, including SMAP. We present our experience as a case study, including how we designed data content and formatting, and how we used multiple standards organizations to establish credibility and ensure usability for users of the data sets we produced. Users of our data sets can use netCDF Command-line Operators (NCO) power tools for unlimited control on spatio-temporal subsetting and concatenation of files. The GDAL tools understand the CF metadata and produce fully-compliant geotiff files from our data. ArcMap can then reproject the geotiff files on-the-fly and work with other geolocated data such as coastlines, with no special work required. We share lessons learned in using this combination of standards and a guiding objective of interoperability to significantly improve the user experience, now and in the future.