Climate change during Cenozoic inferred from global carbon cycle model including igneous and hydrothermal activities

Abstract

This paper discusses climate change in the Cenozoic by constructing a global carbon cycle model which is based on the GEOCARB-type model. Major improvements over previous models in this study are as follows. Previous models have not considered CO 2 behavior at subduction sufficiently. They do not distinguish at subduction zones between the CO 2 degassing from a back-arc basin (BAB) and that from an island-arc, although their degassing mechanisms may be different and should be treated separately. Also, previous models might overestimate the effect of silicate weathering in the regions of the Himalayan and Tibetan Plateau (HTP) on CO 2 drawdown. Recent studies have revealed that this is smaller than previously presumed. Thus, we deal with these two kinds of degassing independently, and estimate the contribution of silicate weathering based on their studies. They are incorporated into the model. The model results indicate that: (1) the contribution of silicate weathering in the HTP region is small; (2) the warming from late Oligocene to early Miocene might be due to the CO 2 degassing from the BAB; (3) the cooling event in the middle Miocene (15 Ma) is caused by a large amount of the organic carbon burial; (4) the CO 2 variation is well consistent with the other studies of CO 2 estimate, especially with the recent studies of CO 2 estimate from the Miocene which indicate a relatively low level of CO 2; (5) the age discrepancy between the CO 2 peak in our model and the period of the Miocene Climatic Optimum might be attributed to the uncertainty of estimate of the BAB production rate, as a result of which the estimate should be revised. The model does not reconstruct the cooling trends from the middle Miocene (about 15 Ma) and the cooling event at the Eocene/Oligocene (E/O) boundary. This mismatch cannot be explained by the uncertainty of parameter estimates. Concerning the former event, the CO 2 increase of the middle Miocene in our model may be consistent with increased volcanic activity resulting in a large amount of CO 2 degassing. Therefore, it could be indicated that the coolings from 15 Ma and at the E/O boundary are attributed to the change of latitudinal temperature distribution caused by the variations of the oceanic environments, because the current carbon cycle models are not able to consider factors other than the CO 2 greenhouse effect in estimation of surface temperature. However, the change of the temperature distribution on the surface itself does not affect the global averaged temperature. This assumption might not be precise because of the effects of in-depth heat exchange and other climatic factors. Nevertheless, a numerical test of the latitudinal sea surface temperature distribution based on observed δ 18O records indicates that these effects are not critical at least for the coolings at the E/O boundary and from 15 Ma. Thus, the result of our model still reflects the reasonable mean temperature at least as a global one, and at the same time, the accurate reconstruction of the above two coolings might be beyond our carbon cycle model.