Hot as You Like It: Models of the Long-term Temperature History of Earth Under Different Geological Assumptions

The long-term temperature history of the Earth is a subject of continued, vigorous debate. Past models of the climate of early Earth that utilize paleosol contraints on carbon dioxide struggle to maintain temperatures significantly greater than 0°C. In these models, the incoming stellar radiation is much lower than today, consistent with an expectation that the Sun was significantly fainter at that time. In contrast to these models, many proxies for ancient temperatures suggest much warmer conditions. The surface of the planet seems to have been generally free of glaciers throughout this period, other than a brief glaciation at ~2.9 billion years ago and extensive glaciation at ~2.4 billion years ago. Such glacier-free conditions suggest mean surface temperatures greater than 15°C. Measurements of oxygen isotopes in phosphates are consistent with temperatures in the range of 20-30°C; and similar measurements in cherts suggest temperatures over 50°C. This sets up a paradox. Models constrained by one set of geological proxies cannot reproduce the warm temperatures consistent with another set of geological proxies. In this presentation, we explore several potential resolutions to this paradox. First, we model the early Earth under modern-day conditions, but with the lower solar luminosity expected at the time. The next simulation allows carbon dioxide concentrations to increase up to the limits provided by paleosol constraints. Next, we lower the planet's surface albedo in a manner consistent with greater ocean coverage prior to the complete growth of continents. Finally, we remove all constraints on carbon dioxide and attempt to maximize surface temperatures without any geological constraints on model parameters. This set of experiments will allow us to set up potential resolutions to the paradox, and to drive a conversation on which solutions are capable of incorporating the greatest number of geological and geochemical constraints.