Temporal and spatial decorrelation scales in the North Atlantic Ocean
In order to perform optimum interpolation ( OI), which is further used to present climatological maps or to implement network design, the covariance statistics of the field of interest have to be estimated. Subjective choices have to be included in the analysis, whether e.g. using the structure func...
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2007
|
| Subjects: | |
| Online Access: | https://oceanrep.geomar.de/id/eprint/3280/ https://oceanrep.geomar.de/id/eprint/3280/1/Dipl.%202007%20Hummels,%20R.pdf |
| Summary: | In order to perform optimum interpolation ( OI), which is further used to present climatological maps or to implement network design, the covariance statistics of the field of interest have to be estimated. Subjective choices have to be included in the analysis, whether e.g. using the structure function or directly estimating the covariance or which analytical fit will be applied to the data correlation values. These subjective choices are evaluated in a test case forming a method to determine the covariance field and the resulting scales of decorrelation. Different data sets having individual premises are analyzed and partly compared. Previous analysis of this kind build on other data sources formed the network design for the ARGO project, which was started in year 2000 and is nearly fully established consisting of a global array of autonomous profilers. ARGO serves as the main data set for this analysis and the covariance field and the resulting scales can now be determined from the data provided by the project itself. Zero-crossing scales for temperature, salinity and the mixed-layer-depth are estimated, whereas temperature and salinity provide similar zero-crossing scales of up to 350 km, 8° latitude, 16° longitude and 23 months. Scales for the mixed-layer-depth are slightly smaller, but it is found, that the scales for the mixed-layer crucially depend on the criterion chosen to define the mixed-layer-depth out of temperature and salinity data. The N orth Atlantic Ocean, as defined here ( 45°N-81 °N, 70°W-20°0), has not been object to this kind of analysis before. As it is a region strongly influenced by topographic features, ideas are presented how the method could be modified taking the topographic influence into account. |
|---|