Skip to main content

Global navigation satellite system (GNSS) radio occultation (RO) has become critical to space-based climate monitoring, improving and stabilizing weather prediction, and atmospheric research over the past several decades, but it has suffered from inadequate cataloging, inconsistent data formats, and difficulty in accessing and downloading data. Led by principal investigator Dr. Stephen Leroy (Verisk Atmospheric and Environmental Research, Inc.), we have addressed these problems by reformatting and cataloging RO data as produced by multiple RO processing centers and making them available to the global public free of charge through the Registry of Open Data of Amazon Web Services (AWS).

Project Objectives

Our goals include the following:

  • Provide all GNSS radio occultation data to the global public in consistent data formats free of charge through the cloud
  • Produce tutorial demonstrations that illustrate how to use RO data in the cloud for common scientific problems
  • Conduct workshops that introduce RO data in the AWS cloud and cloud computing in general

This project makes all GNSS RO data available to the global public via the Registry of Open Data of AWS. As a remote sensing technique, GNSS RO has proven invaluable for climate monitoring, for atmospheric process studies involving the need for high vertical resolution, and for anchoring numerical weather prediction and atmospheric reanalysis. Research involving RO data, however, has been stunted due to data organization, inhomogeneous data formats, download bandwidth and data volume, and the inability to query and subset RO data. We address these issues by reformatting RO retrieved products to data formats independent of retrieval center, we do so for retrieval products from multiple retrieval centers, we post the data in the cloud for free access at high bandwidth, and we have created an online metadata database of all RO dating to the original mission, GPS/MET, in 1995. We have provided a Python application programming interface (API) that enables quick queries of RO data, sub-setting by critical RO metadata, and subsequent downloading of the data.

This project reformats all GNSS RO data produced by three independent RO processing centers into consistent data formats and catalogs metadata for each individual RO retrieval in a database. The lowest level product contains excess phase, amplitude, and precise orbits, considered a Level 1b collection of data for each occultation. The second level product contains bending angles versus impact parameter and high vertical-resolution retrievals of refractivity, typically considered Level 2a collection of data. The highest level product contains retrievals of temperature, pressure, and water vapor as functions of height, typically considered Level 2b collection of data. At this point the repository contains all RO mission data processed by the University Corporation for Atmospheric Research (UCAR), COSMIC-1, and Metop data as processed by the Radio Occultation Meteorology Satellite Application Facility (ROM SAF). Future holdings will include Level 1b data for all RO data processed by the European Organisation for the Exploitation of Meteorological Satellites (EUMETSAT), and Level 1b, Level 2a, and Level 2b data for all RO missions processed by NASA's Jet Propulsion Laboratory (JPL).

Several tutorial demonstrations have been composed and published on GitHub illustrating the power of cloud-based RO data. The tutorials take the form of Jupyter notebooks and they address analysis of RO distributions, inter-center comparison of retrievals, and atmospheric process studies. The project team will host at least two public workshops to demonstrate the power of cloud computing for scientific research. The first was held at the 9th meeting of the International Radio Occultation Working Groups in Austria in a two-hour session. The second will be held over a span of four days in Boulder, Colorado.

Major Accomplishments

  • All existing GNSS radio occultation data are now hosted in the AWS Registry of Open Data free of charge and at high bandwidth to the global public
  • Tutorial demonstrations in the form of Jupyter notebooks are available to illustrate how the AWS repository of GNSS radio occultation data can be exercised for scientific research
  • Public workshops are being held to guide (especially) early career scientists in the usage of cloud computing for purposes of scientific research

For More Information

Access the Earth Radio Occultation - Registry of Open Data on AWS.

Publications and Presentations

Leroy, S.S., McVey, A.E., Leidner, S.M., Zhang, H., & Gleisner, H. (2023). GNSS radio occultation data in the AWS cloud. Earth and Space Sciences, 2023EA003021, In Review.

Leroy, S.S., & McVey, A.E. (2023). GNSS radio occultation data in the AWS cloud: Utilities and examples. Zenodo. doi:10.5281/zenodo.7799039