A few scientific results already obtained with the ANDRO dataset can be downloaded here.
The following text Ollitrault_ColindeVerdiere.JPO2014.pdfwas presented during the Venice Argo workshop 27-29 September 2012, and the paper (unpublished yet) was also prepared for this workshop. A comprehensive and updated version (to be preferred) is given in:
The MATLAB program ANDRO_1000dbar_geopotential_reading.m allows you to read the file ANDRO_20121129_1000dbar_geopotential_2M.dat which contains the geopotential values (m2 s-2) on the mean 1000 dbar surface worldwide. The geopotential Φ of the 1000 dbar isobar is obtained by solving numerically the Poisson equation that results from taking the divergence of the geostrophic equations on the sphere, assuming Neumann boundary conditions. Details are given in Ollitrault_ColindeVerdiere.JPO2014.pdf.
This figure ANDRO_20121129_1000dbar_geopotential_2M.pdf shows the geopotential height (in cm) of the mean 1000 dbar surface, estimated as 102 Φ/g with g = 9.80665 ms-2.
This geopotential was obtained after 2M temporal iterations, starting from an initially flat surface. The convergence of the solution is exemplified in the following figure convergence_diagram_2M.pdf
Using the ANDRO_20121129.dat dataset, the following ASCII file ANDRO_20121129_UVT_circ150km_CM_950m_1150m_4d_17d.dat was generated prior to the estimation of the 1000 dbar geopotential (Ollitrault & Colin de Verdière, JPO, 2014).
It contains current and temperature averaged over 150 km diameter disks, centred every half integer degrees (74.5°S, 73.5°S, …, 79.5°N and 179.5°W, 178.5°W, …, 179.5°E). Only disks containing at least 30 days of float data provide averages. Only float subsurface displacements with representative pressures comprised between 950 and 1150 dbar, and subsurface drift durations comprised between 4 days and 17 days are considered. Averages are pondered with the drift durations and affected to the mass centre of the displacement positions found within the current 150 km diameter disk.
Information found in this file are: index of latitude (1 corresponding to 74.5°S), index of longitude (1 corresponding to 179.5°W), <u> in cm s-1, <v> in cm s-1, <u’2> in cm2s-2, <v’2> in cm2s-2, <u’v’> in cm2s-2, float-days, number of different floats selected in the current disk, <t>, <t’2>, <u’t’>, <v’t’>, float-days with temperature data, latitude of the mass centre, longitude of the mass centre.
The following plots show the mean currents near 1000 m depth over the whole world Ocean.
They are obtained from the ANDRO data set as averages over all displacements found in the [950,1150] dbar layer and within 150 km diameter disks centred on half integer latitudes and longitudes (e.g. at 45.5°N, 35.5°W).
In the plots shown, the origins of the arrows representing the resulting mean currents are centred on the center of mass of the individual velocity positions (the weights being the durations of the individual displacements considered).
Only displacement duration comprised between 4 days and 17 days were considered (see Ollitrault and Colin de Verdière, JPO 2013 for further details).
Blue, cyan, and yellow vectors are the 120-, 45- and 15-day equivalent displacements corresponding to velocities less than 1 cm s-1, between 1 and 5 cm s-1, and greater than 5 cm s-1 respectively.
Note that a mean velocity vector (or rather its corresponding displacement) is plotted only if there are at least 30 days of data within the 150 km diameter disk considered.
Error ellipses (0.39 probability inside) on the mean currents are given only for the yellow arrows (i.e. whenever the velocity is greater or equal to 5 cm s-1) and only if there are at least 90 days of data within the disk under consideration. Appendix A in Ollitrault and Colin de Verdière, JPO 2013, gives further details).
The 950-m isobath (white) and the 2000-m and 4000-m isobaths (black) are shown on the plots.