Validity limits of carbon tetrachloride as an ocean tracer

We investigate non-conservative behavior of carbon tetrachloride (CCl4) in the ocean by evaluating concurrent data of this tracer and of the chloroflu orocarbon CFC-12 from three zonal sections in the South Atlantic( similar t o 30 degreesS, 19 degreesS, 11 degreesS; METEOR cruises M22/5, M15/3, M28/1 ) and from two sections in the western Weddell Sea (POLARSTERN cruises ANT XIII/4, ANT XV/4). The issue is of interest biogeochemically and for uses o f CCl4 as a transient ocean tracer. For the South Atlantic a simple model i s employed that simulates the meridional tracer transfer into the Central W ater and Antarctic Intermediate Water from their southerly outcrops. From a joint fit for the three sections we deduce a CCl4 depletion rate of approx imately 22% per year for temperatures exceeding 13 degreesC, which confirms a previous estimate and exceeds rates due to hydrolysis by up to about 50- fold. A tracer utility of CCl4 in warm ocean waters thus hardly exists. How ever, below similar to 13 degreesC the decomposition rates decrease sharply , and they become negligibly small below about 3 degreesC (rate < 0.1% per year, compatible with rates due to hydrolysis). In the Weddell Sea we do no t find positive evidence of a CCl4 destruction at depth (upper limit 1% per year), in keeping with the South Atlantic result. In the western Weddell S ea deep waters we deduce a apparent CCl4/CFC-12 ratio age of about 30 years . We confirm a previous claim of a CCl4 deficiency in newly formed Weddell Sea deep and bottom waters, which we deduce to amount to approximately 32% relative to CFC-12. We ascribe this deficiency to CCl4 loss within the vent ilated source waters (possibly due to interaction with sea or shelf ice), c ombined with a slower gas transfer from the atmosphere into the upper water s (contribution <similar to> 12%). It is argued that CCl4 deficiencies rela tive to more stable tracers should be common in newly formed deep and botto m waters, and that assessing such initial deficiencies is a prerequisite fo r using CCl4 as a tracer. An open question is a CCl4 instability at reduced oxygen concentrations, although the critical oxygen level appears to be lo wer than reported previously (Tanhua et al., Mar. Chem. 54 (1996) 159). Mor eover, temperature might only be a proxy for the real agent that governs CC l4 destruction. The actual mechanism of decomposition remains unknown, but judging from an Arrhenius plot a first-order chemical reaction can be exclu ded. It is estimated that the ocean contributes roughly 8% to the total env ironmental destruction of CCl4. (C) 2001 Elsevier Science Ltd. All rights r eserved.