Posted by Prof Dr Hans von Storch, System Analysis and Modelling
My PhD student Meng Zhang defended yesterday successfully her thesis The statistics of travelling eddy variability in the South China Sea and its external forcing at the University of Hamburg.
Her work is significant for two reasons:
- She has constructed a climatology of travelling eddies in the South China Sea, and derived various statistics of these eddies, such as the distribution of travel times and lengths. Previously such climatologies have been derived, for shorter times, from satellite data, however without proper regard of the limitations of a high noise level of daily satellite-based data of sea surface height.
- She found the variability of a number of parameters describing the statistics of eddies strongly contaminated by „noise“, i.e., by local processes unrelated to the large-scale state. Thus, she shed light on the unavoidable stochastic character of smaller scale dynamics in regional oceans. This fact ist almost always disregarded by regional oceanographers, who conduct numerical experimentation by simply comparing two simulations and attributing the differences to the change implemented in the simulations. Obviously, ensembles of simulations are needed in extended simulations, and the assessment of numerical experimens needs statistical assessments (such as significance tests).
My little workgroup is pursuing this issue, with first demonstrating the presence of noise and now assessing the scale-dependence of signal-to-noise ratios (with the signal stemming from atmospheric forcing, and the noise reflecting local variations unrelated to external factors). (See also “Noise” in climatologically driven ocean models with different grid resolution). This work is done in cooperation with the Ocean University of China (中国海洋大学)