Research to Operations is a Reversible Process
The launch of National Oceanic and Atmospheric Administration’s (NOAA’s) next-generation satellite, GOES-R, brings with it a whole new set of complex algorithms. These algorithms not only ensure legacy products continue uninterrupted, but they will also produce a few new products not available in the current system.
Going forward, all of these algorithms will continue to be updated and improved, providing the meteorological, oceanographic, and space community with a constantly improving data set to work with. The process of improving these algorithms, whether by the government or academia, is called research to operations (R2O).
Many people fail to see that an efficient R2O program needs—requires is probably a better word—an excellent operations-to-research (02R) program. Algorithms should be developed for easy insertion into research and development systems as well as operational systems. An operational algorithm that cannot be easily transferred to a research facility—O2R—will probably not be the basis for the next algorithm improvement—R2O. At best, future modifications will be difficult, time consuming, and costly to implement. At worst, the hypothetical wheel gets reinvented.
GOES-R industry partners Harris and Atmospheric and Environmental Research (AER) have adopted two methods for improving the ease of R2O2R that can work for developing the new algorithms needed with a new enterprise ground system:
Method #1: The Algorithm WorkBenchTM (AWB) – In the AWB method, algorithms are managed as components. A data model interface that is common among development, test, and production environments is used. Algorithms may be compatible with all the major algorithm languages, like C++, FORTRAN, and Python.
Method #2: The Algorithm Bridge Service (ABS) – The power of ABS is that R2O is facilitated as well as O2R. ABS enables new algorithms to be inserted into operations or research without software modification. Also, no special infrastructure requirements—other than what the original researcher designed—are required to transfer the same algorithm software from operations back to a research environment. In addition, this new methodology does not require the individual researcher to adopt expensive operational constraints required to ensure that latencies and data interactions are met.
By enabling their products for both research and operations, NOAA will be better positioned to improve service to its customers. The future is bright for GOES-R and the research community’s efforts to continuously and tightly work with the operational algorithms, providing an ever increasing quality of service to the American public and the world.
Read more about Harris’ involvement in the GOES-R mission.