CONVECTION BENEATH FREEZING LEADS - NEW OBSERVATIONS COMPARED WITH NUMERICAL-MODEL RESULTS

Citation
Rd. Muench et al., CONVECTION BENEATH FREEZING LEADS - NEW OBSERVATIONS COMPARED WITH NUMERICAL-MODEL RESULTS, J GEO RES-O, 100(C3), 1995, pp. 4681-4692
Citations number
18
Categorie Soggetti
Oceanografhy
Journal title
JOURNAL OF GEOPHYSICAL RESEARCH-OCEANS
ISSN journal
21699275 → ACNP
Volume
100
Issue
C3
Year of publication
1995
Pages
4681 - 4692
Database
ISI
SICI code
2169-9275(1995)100:C3<4681:CBFL-N>2.0.ZU;2-B
Abstract
Vertical distributions of temperature, salinity, and horizontal and ve rtical current speed were measured along the edges of a number of rapi dly freezing leads in the southern central Beaufort Sea during March-A pril 1992. These observations were restricted to cases having sufficie ntly small ice water relative speeds that brine-driven convection was expected, based on scaling arguments, to dominate shear turbulence gen erated by the relative ice water motion. The observed salinity and cur rent features were consistent with an existing conceptual model for su blead convection consequent to brine rejection from ice formation, and with the results of a numerical model which assumes two-dimensionalit y in the cross-lead direction, and uses a brine input consistent with the field observations. These models predict convection of brine-enric hed plumes beneath the lead downward to the pycnocline, where they the n spread outward away from the lead. Volume continuity is satisfied by horizontal inflow of water to the lead at the surface. Observed mean downward convection speeds varied from 0.2 to 0.7 cm s(-1) and had max ima up to about 2 cm s(-1), consistent with the numerical model. Salin ities associated with the convecting parcels were about 0.001 practica l salinity units (psu) above ambient, which was an order of magnitude less than predicted by the model, while plumes of water flowing away f rom the lead on the pycnocline had salinities as high as 0.004 psu abo ve ambient. These similarities and discrepancies are discussed within the context of a steady state model versus a nonsteady field situation , and it is concluded that the steady ''convection cell'' concept is p robably valid if viewed as a time-averaged sequence of discrete convec tive plumes or ''thermals.''