ERA5, MERRA2 and GNSS RO Water Vapor Comparisons and Implications
Presenter:
E. Robert Kursinski
PlanetiQ
PlanetiQ
Talk
We have extended our examination of the hydrological cycle and our understanding of it by comparing low latitude, free tropospheric specific humidity histograms from ECWMF ERA5 and NASA MERRA-2 reanalyses with those derived from GPS-RO. In comparison to ERA-Interim and the operational ECMWF analyses of 2007, ERA5 performance has improved incrementally across the free troposphere with the biggest improvement in the mid-troposphere. In comparison with the MERRA reanalyses, the MERRA-2 reanalyses have improved significantly in the lower free troposphere but are noticeably worse than MERRA in the mid and upper troposphere. The MERRA humidity results remain the closest to the error deconvolved GPS RO results in the mid and upper troposphere.
The observational constraints on global moisture analyses are dominated at present by information contained in satellite IR and microwave radiances. We suspect that the limited improvements in newer global moisture analysis/reanalysis product performance reflect the fact that coarse vertical information from radiances provide limited constraints on key hydrological processes and therefore rather limited guidance for model refinement.
At present, GNSS RO sampling densities are too sparse to have much impact on global NWP moisture analyses. These results suggest that the observational constraints from a 10x to 100x increase in GNSS RO sampling densities would enable major advances in our understanding and representation of the hydrological cycle. These would come both by forcing the analyses to agree with the dense GNSS observational constraints and by providing the modelers with dense, high vertical resolution constraints containing tell-tale signatures of key processes at work that are needed to refine how these processes are represented in models and drive model climatologies closer to reality. Presumably, the resulting improved process-representation would then enable NWP models to better hold onto analysis moisture increments and extend model forecasting skill farther into the future.
Presentation in PDF:
