GLOBAL WATER CYCLE AND EARTH'S THERMAL EVOLUTION

Abstract

Convergent margin processes play an important role in the distribution of terrestrial volatile species. During subduction processes volatiles are filtered from the subducting package and are restricted to return to the mantle. Water is the most abundant volatile and plays an important role in these processes. There is a number of geochemical investigations to determine the subduction, regassing, and recycling fluxes as well as the regassing ratio of water. The latter describes the partition of subducting water by water that is regassed into the mantle and water that is returned to the surface in arc magmas. Here we present a geophysical-based modelling approach for the calculation of such fluxes and ratios in order to compare them with the geochemical data. In order to assess the recent values and the evolution of the subduction, regassing, and the recycling flux a simple parameterized thermal convection model with a water-dependent rheology and a constant continental growth model is applied. To test the sensitivity of the results different continental growth models were applied and the total amount of water in the system was varied as well as the initial distribution of water in the reservoirs. According to our estimations a value of 0.31 for the time independent regassing ratio of water, RH2O, is an acceptable upper bound. Lower values of RH2O give larger water reservoirs on the surface compared to the recent situation. Larger values of RH2O suggest smaller surface reservoirs of water and, therefore, seem to be unlikely. The model results show a relatively stable value for the regassing ratio of 0.31 by varying the initial conditions of the water distribution in the reservoirs (which are pretty much unknown at the present moment). But RH2O is very sensitive towards the total amount of water in the system. Altering the value of four ocean masses to ten we get values for the regassing ratio from 0.31 to 0.89. Nevertheless, as a result of all numerical experiments the recent subduction flux is stable and equal to 1.02x1015 g/a. The influence of the continental growth model on the results could be neglected. The calculated value for the recent subduction water flux fits the modern geochemical data very well while our value for RH2O is smaller. One possible reason could be that in our experiments RH2O remains constant and, therefore, represents an average value over Earth's history. In order to check this assumption we apply a simple exponential time dependence of RH2O. Here, the modern regassing ratio increases to 0.41. Therefore, based on a geophysical modelling approach in contrast to the geochemical investigations we suggest a smaller value for the modern regassing ratio of about 0.3 to 0.4.