A250-03
OSSE System Improvements at NOAA/AOML

Thursday, 17 December 2020: 04:10
Virtual
Sean Casey1, Lidia Cucurull2, Paulo Paz1, Andres Vidal1, Michael Mueller3 and George R Halliwell Jr2, (1)Cooperative Institute for Marine and Atmospheric Studies/Atlantic Oceanographic and Meteorological Laboratory, Miami, FL, United States, (2)NOAA Atlantic Oceanographic and Meteorological Laboratory, Miami, FL, United States, (3)Cooperative Programs for the Advancement of Earth System Science/Atlantic Oceanographic and Meteorological Laboratory, Boulder, CO, United States
Recorded Presentation
Abstract:
Under the auspices of the Quantitative Observing System Assessment Program (QOSAP), the Observing System Analysis group at NOAA/AOML has refined and expanded its systems for use in Observing System Simulation Experiments (OSSEs). The first component in an OSSE, a Nature Run, provides a complete record of the assumed true state of the atmosphere. One of the requirements of any Nature Run is that it should reproduce the main characteristics of the real atmosphere. The ECMWF Cubic Octahedral Nature Run (ECO1280) has been validated for the first three months that are being used to conduct OSSEs. This validation concluded that the main characteristics of a real atmosphere were present and that it possessed a model climatology that agreed with reanalysis datasets during that time period. With respect to observation simulation, the capabilities of the Consolidated Observing Systems Simulator (COSS) have been expanded by adding radiance simulations with error addition (Clear-sky and all-sky) over 26 different sensors. The validation of simulated radiances from the ECO1280 nature run has shown reasonable values and geographical distribution in preliminary assessments; however, some issues exist and the group is working to resolve or document them for future OSSEs. Explicit observation errors have been added in order for Finite-Volume Cubed Sphere Global Forecast System (FV3GFS) observation minus background (O-B) statistics to be similar to its real-world counterpart. In doing so, forecast skill has been shown to be comparable between FV3GFS predicting the real world vs. predicting the OSSE world. Finally, ocean OSSE work demonstrated that assimilating ocean profiles sampled by an array of gliders deployed over an oceanic region is more effective at correcting ocean mesoscale structure than assimilating ocean profiles sampled at fixed locations. The ability of individual piloted gliders to cross ocean gradients and map mesoscale structure in their vicinity is responsible for this enhanced effectiveness.
<< Previous Abstract | Next Abstract >>