The data distribution in space is shown in Figure 1, the distribution in time is shown in Figure 2 and the bins are described in Table 1. To see the results, click here
The SOOP /XBT network around Australia was established by the CSIRO Division of Marine Research, as a national contribution to the international Tropical Ocean Global Atmosphere (TOGA) and the World Ocean Circulation Experiment (WOCE) programs.
The goals of XBT sampling, under the TOGA program, were :
Three different approaches to sampling, were used :
Under the broadscale mapping approach, ships were selected to spread out, in order to map the temperature fields horizontally, throughout the basin. Broadscale mapping in relation to this project, is done by Neville Smith, at BMRC - Bureau of Meteorology Research Centre. Ideally, the sampling strategy was to cover each line in the TOGA/WOCE/IGOSS XBT Network, once per month, with an XBT drop every 6 hours ( or 150km ). This mode of sampling is sometimes called "low density".
Under the frequently repeated trackline approach, ships were selected to sail repeatedly along nearly the same route, in order to resolve the low frequency detail of temporal variability of temperature sections. Ideally, the sampling strategy was 18 voyages per year, and 6 observations per day. The sampling rates are based on the method of optimal interpolation ( Meyers et al., 1989; Phillips et al., 1990; Sprintall and Meyers, 1991; Meyers et al.,1991; Smith et al, 1991 ).
The high density sampling rate was 4 seasonal sections per year at a higher rate of spatial sampling ( 16-20 observations per day ) in order to obtain eddy statistics of temperature and geostrophic shear and transport fields.
The broadscale and frequently repeated routes are covered by volunteer merchant ships, which measure the temperature down to 450m or 750m, depending on the type of XBTs deployed on the ship.
This analysis is focused on the frequently repeated XBT lines, comprising initially the CSIRO SOOP XBT network. Later, it will be extended to additional lines in the TOGA/WOCE/IGOSS XBT network. Only lines with good temporal coverage were selected, usually with 12 - 18 sections per year.
All data have been quality controlled by A.Gronell and R.Bailey, according to the methods developed by Bailey et al ( 1994 ) for processing delayed mode XBT data. Only top and good quality XBT data are used in this study.
The XBT temperatures are binned and analyzed by the methods described below. The bins are 1 degree in latitude, in general, with the exception of the bins near the continental slope, which were adjusted on some lines, to represent coastal currents. Line PX-2 is an exception, because it runs east-west, across the Banda Sea. Overlapping 1 degree longitude bins, with half degree spacing, were used in this case, to partially resolve the thermal structure associated with the gaps between islands of the Indonesian archipelago.
