Operational Ocean Data Assimilation to Improve Upper Ocean Current Estimates for Global Ocean Monitoring, Coupled Climate Forecasts, and Coupled Hurricane Forecasts

Author:

Stephen Penny - (University of Maryland, College Park)

Co-Investigator(s):

James Carton (University of Maryland, College Park)

Co-I/Institutional PI (s):

Dr. David Behringer (University of Maryland, College Park)
Dr. Yan Xue (University of Maryland, College Park)

Collaborator(s):

Matthew Harrison (University of Maryland, College Park)
Dr. Guillaume Vernieres (University of Maryland, College Park)

Abstract:

This proposal supports the operational assimilation of all available altimetry data (including Jason 3), scatterometer surface winds, and near surface drifters from the Global Drifter Program (http://www.aoml.noaa.gov/phod/dac/index.php) to enhance estimates of upper ocean currents in an eddy-permitting 1/4º global ocean model. In addition, these data will be combined with in situ Argo profiling floats, moored buoy arrays, and satellite measurements of sea surface temperature (SST), which are all already assimilated in a new Hybrid Global Ocean Data Assimilation System (Hybrid-GODAS) that is in preparation for operational deployment at NCEP.

This project expands the scope of the OSCAR (http://www.oscar.noaa.gov) estimates of near surface currents by:

• Including more data sources: in addition to satellite altimetry and scatterometer data, we incorporate commonly assimilated observation data such as Argo profiling floats, moored buoys, and satellite SST.
• Providing uncertainty estimates: we will provide uncertainty estimates of the ocean surface currents as derived from ensemble statistics.
• Applying Lagrangian assimilation of surface drifters: we use real-world drifters to guide and constrain simulated drifters within the data assimilation, thus ensuring consistency between satellite and in situ measurements of surface currents (Ph.D. thesis of current student Luyu Sun)
• Incorporating a global dynamical model of the ocean: we use the GFDL MOM6 dynamics.
• Using a state-of-the-art data assimilation (DA) method: We are among the first to achieve a hybrid DA system for the ocean, using the Hybrid-Gain method of Penny (2014; MWR).
• Accounting for atmospheric surface forcing errors: we incorporate an ensemble of surface fluxes derived from an ensemble atmospheric analysis (e.g. the 20th century reanalysis for historical periods, or the operational ensemble analysis starting in 2012).
• Assimilating surface winds into the ocean: we have developed a strongly coupled DA approach (Sluka et al., 2016; GRL)
• Increasing the resolution: we produce our analyses on a 1/4 global grid.

We further the development of our advanced Hybrid-GODAS (Penny et al., 2015; MWR). A 21-year reanalysis study has shown that the Hybrid-GODAS provides significant improvement over the current operational 3DVar-based GODAS in many fields, including near surface current anomalies (in comparison to OSCAR), equatorial Pacific subsurface currents (in direct comparison to ADCP data), upper ocean heat content, and sea surface salinity anomalies, along with an overall global reduction in RMS deviations and biases between forecasts and observations.

Our Hybrid Ocean DA system is currently in research and development for three major operational applications at NCEP: (1) Global ocean monitoring, of primary interest to the CPC (now transitioning to operations), (2) Ocean initialization for a 3-way coupled (HWRF/WW3/HYCOM) hurricane forecast, and (3) The ocean component of NCEP’s next-generation Climate Forecast System (CFSv3). As the same DA system is shared by all of three of these applications, advancements made to the global ocean monitoring system will be immediately leveraged for ocean initialization in coupled hurricane forecasting and coupled climate forecasting as well.