NOBLE-GAS CONSTRAINTS ON THE EVOLUTION OF THE ATMOSPHERE-MANTLE SYSTEM

We present a model on the evolution of atmosphere-mantle system assumi ng a significant degassing from the less depleted mantle, based on the recent results of the He-Ar systematics proposed by Matsuda and Marty (1995). The degassing fluxes of noble gases are represented by the co ncentrations in the mass flow, which follows the model of Porcelli and Wasserburg (1995a,b). However, we have not assumed the steady-state, and have assumed the mass transfer as an exponential function of time. The degassed amount from the less depleted mantle is related to the m ixed amount of mass between the less depleted mantle and the depleted mantle. The amount of mixing is constrained from the K abundances in e ach mantle reservoir. Consequently, the degassed amount from the less depleted mantle is also constrained. The following conclusions are obt ained from this model. The initial elemental ratio He-3/Ar-36 is likel y to be in cosmic abundances, and the initial concentration of Ar-36 i n the Earth is constrained to be similar to 1 x 10(12) atoms/g. The pr esent fraction of Ar-36 in the depleted mantle reservoir is calculated to be at most 2.4 X 10(-3) of the initial inventory. The Ar-40/Ar-36 ratio in this reservoir is estimated to be less than (1.3-1.4) x 10(5) . The less depleted mantle reservoir also degassed and the present amo unt of Ar-36 in calculated to be at most 1.3 X 10(-1) of the initial i nventory. The Ar-40/Ar-36 ratio in the less depleted mantle reservoir is estimated to be less than 3.7 X 10(4). To explain the difference be tween the neon isotopic ratios in the atmosphere and those observed in the mantle rocks, isotopic fractionation during the escape from the a tmosphere and/or significant contribution of late accreted materials p ossessing planetary Ne is necessary. In our model, Ne degassing from b oth mantle reservoirs would have provided at least 50-90 times the amo unt of Ne-22 in the present atmosphere. Therefore, the fraction suppli ed by late-accreted material would be negligible, which is contrary to the model proposed by Porcelli and Wasserburg (1995b). We conclude th at isotopic fractionation model during escape from the atmosphere prov ides the best fit to the observations. Copyright (C) 1998 Elsevier Sci ence Ltd.