Earth rotation and the ocean's circulation
Zusammenfassung der Projektergebnisse
The goal of this project was to derive knowledge about the oceanic angular momentum and mass distribution anomalies by combining geodetic observations and numerical models of the oceans with assimilation based techniques. The OMCT ocean model was extended to generate a parallel ensemble of ocean simulations. The cross-ensemble covariances were exchanged with a Kalman assimilation framework which was coupled to the model. This setup was expanded to incorporate the error sources for the atmospheric forcing. To account for the huge computational costs of an ocean ensemble a subspace Kalman-Filter approach was chosen. The applied SEIK filter exclusively operates on the dynamical important fraction of the model’s state space. The error information for the initial ensemble, the forcings and the observations were derived and tested. The final assimilation setup was able to assimilate Earth rotation parameters as well as GRACE mass field anomalies. The reproduction of the geodetic observations with our assimilation technique was successful. RMS-misfits between model and observation could be reduced by 40-80%. The project’s assimilation of GRACE and Earth rotation observations were unprecedented in the field of oceanography and were published for the first time in journal papers and conference contributions. The impact of the assimilation on the ocean state is significant and gives important information about unobserved ocean quantities like the meridional overturning circulation or the total ocean mass. The relevant changes in the assimilation simulation could be connected to the altered atmospheric forcings. The deviations with respect to the unaltered forcings contained information about the errors of atmospheric reanalyses and about the physical mechanisms which generate the assimilated observations in the ocean. We found that the dominant wind stress errors reside in the high latitudes while the prominent precipitation errors are located in the tropics and the high latitudes. In general, it can be stated that the forcings distributed by the reanalysis centers are not optimal for the generation of oceanic mass distributions and mass transports as observed by GRACE and Earth rotation products. When the Kalman-Filter was not allowed to change the forcing fields only small improvements in the model trajectory were possible. The corrections to the atmospheric forcings which can now be calculated with the developed assimilation approach could be of great value for the oceanographic community as well as for the centers which generate the atmospheric forcings. As for the physical mechanisms, the Length of Day changes could be linked to total ocean mass changes which has for geometrical reasons no effect on Polar Motion. Polar Motion itself is generated by wind forcing via the geostrophic balance. The resulting spatial mass distributions have likewise only little impact on Length of Day. The wind forcing is to a large extent responsible for the sub-seasonal ocean bottom pressure anomalies as observed by GRACE. The respective physical link could be identified as the wind driven oceanic Ekman transport. In addition, the assimilation successfully removed the artificial stripes from the GRACE observations.
Projektbezogene Publikationen (Auswahl)
- Assimilation of Earth rotation parameters into a global ocean model: length of day excitation. Journal of Geodesy, p 1-7
Saynisch, J., Wenzel, M. and Schröter, J.
(Siehe online unter https://doi.org/10.1007/s00190-010-0416-0) - Numerical simulations of short-term non-tidal ocean mass anomalies, In: Flechtner, F. Gruber, T., Güntner, A., Mandea, M.,Rothacher, M., Schöne, T., Wickert, J. (eds.), System Earth via Geodetic- Geophysical Space Techniques, Springer, Heidelberg, 119-129
Dobslaw, H. and Thomas, M.
(Siehe online unter https://doi.org/10.1007/978-3-642-10228-8_10) - Seasonal polar motion excitation from numerical models of atmosphere, ocean, and continental hydrosphere. Journal of Geophysical Research
Dobslaw, H., Dill, R., Grötzsch, A., Brzezinski, A. and Thomas, M.
(Siehe online unter https://doi.org/10.1029/2009JB007127) - Short-term Polar Motion Forecasts from Earth System Modeling Data. Journal of Geodesy, 84, 9, 529-536
Dill, R. and Dobslaw, R.
(Siehe online unter https://doi.org/10.1007/s00190-010-0391-5) - Assimilation of Earth rotation parameters into a global ocean model: excitation of polar motion. Nonlin. Processes Geophys., 18, 581-585
Saynisch, J., Wenzel, M. and Schröter, J.
(Siehe online unter https://doi.org/10.1007/s00190-010-0416-0) - Ensemble Kalman-Filtering of Earth rotation observations with a global ocean model. Journal of Geodynamics, GEOD1089
Saynisch J. and Thomas, M.
(Siehe online unter https://doi.org/10.1016/j.jog.2011.10.003) - Assimilation of GRACE-derived oceanic mass distributions with a global ocean circulation model. Journal of Geodesy, February 2015, Volume 89, Issue 2, pp 121–139
Saynisch J., Bergmann-Wolf, I. and Thomas, M.
(Siehe online unter https://doi.org/10.1007/s00190-014-0766-0)
