CRTF Telecon - 09-23-2015

Hybrid GODAS

Steve Penny, UMD/NCEP

A 21 year historical ocean reanalysis (1991-2011) has been produced at NCEP using the Hybrid 3DVar/LETKF Global Ocean Data Assimilation System (Hybrid-GODAS) and the 3DVar GODAS similar to that in the operational GODAS. Compared to the 3DVar-GODAS, the Hybrid-GODAS has much smaller bias and RMSD of observation minus forecast in both temperature and salinity. Validated against observations and eight operational ocean reanalyses around the world, it is shown that the Hybrid-GODAS is superior to the 3DVar-GODAS in many aspects including improved thermocline and halocline, sea surface salinity, surface currents and Equatorial Undercurrent variability. However, the comparison with the AMOC measurement at 26N shows an spin up in the meridional transport in the Hybrid-GODAS although the seasonal cycle is well simulated. Since the climatology river runoff is included, SSS variability near the Amazon outflow region is underestimated. In addition, there are large mean biases in SSS in various regions in the Atlantic and near the Antarctica, suggesting the climatology fluxes from R2 might be biased in those regions. A plan is to shift toward 1/4^ox1/4^o global MOM6 with 2m resolution near the surface considering that high resolution model provides much larger ensemble spread needed to correct model errors, and to implement Hybrid-GODAS into operation as a replacement for GODAS. Experiments with a simple coupled model with assimilation of atmosphere data only show that strongly coupled data assimilation leads to 40% reduction of RMSE in the ocean.  So there is a plan to shift toward a strongly coupled data assimilation system including all the components in CFSv3 (atmosphere, wave, ocean, sea ice, land and aerosol).

• Discussion topics
  
  • Relaxation back to climatology time scales, does it affect salinity assimilation?
  • Tropics; what is the spread caused by?
    • SW forcing – temperature
    • Thermocline – wind stress
  • What is causing the bias sign change around Antarctica?
    • R2 bias
    • Salinity relaxation

Planning for CPO Technical Report on NCRTF May Workshop.

Planning for CPO Report

• Too late for BAMS (4 months after workshop) and EOS (2 months after workshop)
• Major factor to consider: how to resolve the problems that are identified in the reanalyses

Lisan: Maybe useful to summarize work.  

Spin-off from CPO report to a BAMS publication Inbox, map room, on the Task Force.

Xiquan - more advertising for group to “make noise”, finish report and then BAMS pub about progress.

Action item: Everyone who attended workshop, please send any written notes. Deadline 2 October