End-to-end design and development of GOES-R level 2 algorithms: Cloud pathfinder algorithm case study

GOES-R is the next generation of the National Oceanic and Atmospheric Administration's (NOAA) Geostationary Operational Environmental Satellite (GOES) System, and it represents a new technological era in operational geostationary environmental satellite systems. GOES-R will provide advanced products, based on government-supplied algorithms, which describe the state of the atmosphere, land, and oceans over the Western Hemisphere. The Harris GOES-R Core Ground Segment (GS) Team will provide the ground processing software and infrastructure needed to produce and distribute these data products. As part of this effort, new or updated Level 1b and Level 2+ algorithms will be deployed in the GOES-R Product Generation (PG) Element. In this work, we describe the general approach currently being employed to migrate these Level 1b (L1b) and Level 2+ (L2+) GOES-R PG algorithms from government-provided scientific descriptions to their implementation as integrated software, and we present an illustrated case study of this process for a single "Pathfinder" algorithm. The ABI Cloud Mask algorithm was selected for this purpose based on its maturity and level of complexity. In general, GOES-R L2+ algorithms ingest calibrated and geo-located GOES-R L1b data and other ancillary/auxiliary/intermediate information to produce L2+ products such as aerosol optical depth, rainfall rate, derived motion winds, and snow cover. In this work, we describe the ABI Cloud Mask algorithm's journey from its scientific description in an Algorithm Theoretical Basis Document (ATBD) to a working engineering solution, and we highlight milestones in the software development life cycle. This process begins with requirements analysis and continues on through design and implementation to reproducibility testing, with final integration into the functional PG test environment. Throughout this process, the ways in which an algorithm interacts with the common algorithm framework and the larger PG infrastructure are defined and refined. This presentation outlines the software development process approach, refinements that were developed during this "Pathfinder" exercise, and provides a discussion of lessons learned about translating scientific specifications into software design with follow-on implementation.