Improvements in GRACE Release-05 over the oceans and implications for ocean dynamics

Type: Presentation

Venue: AGU Fall Meeting 2012


Katherine J. Quinn; Rui M. Ponte; Christopher G. Piecuch (2012) Improvements in GRACE Release-05 over the oceans and implications for ocean dynamics. AGU Fall Meeting 2012, San Francisco, CA.

The Gravity Recovery And Climate Experiment (GRACE) is unique in its ability to measure changes in ocean bottom pressure (OBP) over the entire globe at large scales. Improvements to the latest release (RL05) of the GRACE data promise better understanding of ocean dynamics. We evaluate the GRACE RL05 data versus the previous release (RL04) over the oceans using OBP estimates produced by the ECCO (Estimating the Circulation and Climate) project, for six years from 2005 to 2010. Uncertainties in the GRACE data can be approximately estimated by comparing GRACE and ECCO and assuming that the true signal and errors are uncorrelated. We find that RL05 has generally lower errors than RL04. Over the open ocean RL04 has errors of ~1-2cm for 750km Gaussian-smoothed data, whereas RL05 has errors of ~0.5-1.5cm. For both RL04 and RL05 the errors are dominated by leakage of strong land signals into the ocean, even though some effort has been made to reduce this leakage with post-processing techniques. This is a particular concern near Greenland, the Amundsen Sea sector of the West Antarctic Icesheet, and Alaska, where large trends in the ice masses lead to large trends in the gravity signal. Regions showing large positive correlations between GRACE and ECCO, large fraction of ECCO variance explained by GRACE, and large signal-to-noise ratios include the northwestern Pacific Ocean, broad areas of the Southern Ocean, the South China Sea and the Arafura Sea. These regions are dominated by annual and sub-annual variability but there is also considerable interannual variability. The demonstrated ability of GRACE to measure interannual OBP variability on a global scale is unprecedented and has important implications for assessing deep ocean heat content and ocean dynamics. The improved GRACE data will also provide new valuable constraints for ocean models.