GRAFIIR – An Efficient End-to-End Semi-Automated GOES-R ABI Algorithm Performance, Analysis, and Implementation Verification System
GRAFIIR is the GOES-R Analysis Facility for Instrument Impacts on Requirements. It is a facility being established to leverage existing work on AWG proxy data, GOES-R science algorithms, visualization, Risk Reduction, sensor trade-offs, and calibration/validation. Its mission is to “connect the dots” between instrument systems development and mission science requirements. GRAFIIR uses existing validated science software, as well as developmental software glue components and applications to provide a flexible experimental environment. Real and proxy sensor measurements are generated with specified perturbations to assess impact on GOES-R mission science. These perturbations can take the form of sensor noise, navigation and registration errors, diffraction, et cetera. The end result is distilled to summary reports (including pass/fail criteria and trouble area identification) and product datasets which can be used for further analysis.
One of the newly developed components is called GLANCE. This software tool is built to assess and evaluate many of the GOES-R data and products (i.e. imagery, clouds, derived products, soundings, winds, etc.) in a semi-automated way which is consistent and traceable to mission science requirements. It is intended to ensure the instrument effects on the products can be fully accounted for, characterized and that algorithm implementation and product performance can be automatically analyzed. It is part of a stable of software components which are being built to provide functionality including:
• ensemble testing tools capable of large scale cluster runs using GEOCAT and GOES-R Algorithm Integration Team (AIT) framework software,
• sensor proxy data with noise permutation tools using FORTRAN and MATLAB,
• data format and content (NetCDF / HDF / AREA / ancillary) management tools,
• data verification reporting automation using Python and conforming to web standards,
• ensemble statistics to depict how the outcome is changed by the variety of ensemble inputs,
• verification statistics with visualization templates and scripts for McIDAS-V (JYTHON/JAVA), and
• time series comparisons and automatic report generation in HTML pages with embedded diagrams and McIDAS-V integration,
As GRAFIIR proceeds with development, demonstrations of this feature set will also be highlighted and made available to the GOES-R development community as a functional and effective end-to-end system supporting NOAA's GOES-R ABI project.