As a contribution to UK WOCE, seven PALACE floats (Profiling Autonomous LAgrangian Circulation Explorers) were deployed in the Irminger Basin in late October 1996. Four were CTD profilers (10587-90) and three were temperature-only profilers (10591-93). All were ballasted to drift at 1500 m, and to surface once every 14 days for position fixing and data transmission. They have performed satisfactorily. One float expired soon after launch (10591), believed due to damage sustained during launch. One was probably destroyed by ice in the East Greenland Current after 2 years (10589). With one exception, 10592, for whose demise no obvious reason exists, the other 4 all survived over 4 years, and one of these (10593) is still working over 5.5 years on.

We were particularly pleased with number 10587, our 'hero float', which had a torrid time from winter 1997 until spring 1998. In mid-November 1997, it was swept inshore while surfaced, then when it submerged, it was in only 100 m water depth. From then for the next 2 months until mid-January 1998, it spent its 'cruising' time grounded in depths between 100 and 1000 m. But that's not all . . . For the three months between early January until early April 1998, most of its Argos transmissions were either abbreviated or entirely absent. Given the season and the location - Northern Labrador Sea in winter - we think that the float was struggling to poke its head through gaps in the sea ice. A very robust float. The utility of such winter profile data from inaccessible regions is illustrated in Cuny et al. (2002).

We have been interested in determining the stability of the conductivity sensors on the floats. The floats were manufactured by Webb Research Corp. and fitted with Falmouth Scientific Inc. HACC conductivity sensors (the temperature sensors were made by Yellow Spring Instruments). The water masses of the Irminger Basin and adjacent areas are highly variable from year to year, so appealing to climatological data was not appealing. However, the floats' first operational year (1996-7) coincided with the intensive phase of WOCE North Atlantic measurements. By comparing float data with near-coincident high-quality research vessel CTD data, we found that the floats' conductivity sensors were stable to within FSI's specification - typical salinity drifts were 0.0005 per month. We can therefore apply reliable calibrations to float conductivities for at least the first 18 months' operation, which result in salinity measurements accurate to 0.005.

Sheldon Bacon S.Bacon@soc.soton.ac.uk

To read more about this work, see the following publications:

Bacon, S., L. Centurioni and W. J. Gould, 1998: Evaluation of profiling ALACE float performance. SOC Internal Document, No. 39, 72 pp. (available as a PDF)

Bacon, S., L. R. Centurioni, and W. J. Gould, 2001: The evaluation of salinity measurements from PALACE floats. J. Atmos. Oceanic Tech., 18 (7) 1258-1266.

Cuny, J., P. B. Rhines, P. P. Niiler and S. Bacon, 2002: Labrador Sea boundary currents and the fate of the Irminger Sea water. J. Phys. Oceanogr. 32 (2) 627-647.