Paradox lost: silicon 32 and the global ocean silica cycle

Abstract

The 32Si Paradox is that the GEOSECS measurements of 32Si specific activity in silica collected on ferric hydroxide-coated fibers are essentially uniform throughout the deep water of the global oceans [Somayajulu et al., Earth Planet. Sci. Lett. 85 (1987) 329–342; 107 (1991) 197–216]. Peng, Maier-Reimer, and Broecker have argued that 32Si specific activities in Indian and Pacific deep water should be 3–5 times lower than in the deep Atlantic, because the dissolved SiO 2 concentrations are higher than in the Atlantic by this factor, and because cosmogenic 32Si should be essentially confined to the ocean basins in which it falls due to its short half-life relative to mixing times for water interchange between the oceans. Thus these authors proposed that the entire GEOSECS 32Si data set “may be flawed”. The resolution of the 32Si Paradox is straightforward. Silica collected on the acrilan fibers is a two-phase mixture of biogenic particulate SiO 2 (opaline tests of diatoms and radiolaria) and silica scavenged chemically from dissolved SiO 2 in ocean water. Particulate silica is the high-activity component in this mixture, and dissolved SiO 2 is the low-activity end-member. Thus the mixing trajectories on ‘Cornucopia plots’ of specific activity vs. reciprocal SiO 2 recovered weights overlap in specific-activity range, regardless of the different concentrations and specific activities of dissolved silica in the deep waters. The specific activities of dissolved SiO 2 in the Pacific, Indian, and Atlantic oceans, as deciphered from the two-component total activity data, are ∼0, 2.6, and 4.5 dpm/kg SiO 2. The atmospheric production rate of 32Si has been calculated and is found to be 0.72 atoms/m 2 s. This value is much lower than in previous calculations, which were based on the Lal and Peters plots of stratospheric fallout that incorrectly use geomagnetic latitude for scaling the fallout patterns. Correcting these curves to scale by geographic latitude, which controls the stratospheric ‘dumping’ pattern, we show that the 32Si concentrations in Indian rains represent the total fallout from both stratosphere and troposphere, rather than only tropospheric fallout as was previously assumed. The new value of the atmospheric production rate is consistent with the low activities of the dissolved silica in the three oceans, which are modulated to some extent by radioactive decay of 32Si during the sequestering of particulate silica in sediments before regeneration in bottom waters.