Carbon dioxide, climate and silicate weathering

The Earth’s climate has changed considerably over the course of it’s 4.5 billion year history. Researching the hows and whys of variations in climate through time has given us an insight into the unprecedented speed and scale of human-driven change, but may have also offered up a solution to help remediate some of the damage we have caused: silicate weathering.

Carbon dioxide (CO₂) in the atmosphere dissolves in water, forming a weak acid which can react with silicate minerals in rocks, forming a solution of calcium (and/or magnesium) cations and bicarbonate anions, among other things. Rivers then carry this solution to the oceans where some of the constituents recombine to form calcium carbonate minerals which may then be buried on the ocean floor.

This process of silicate weathering and carbonate formation draws down and locks away CO₂ from the atmosphere. While this process occurs on longer-than-human-timescales of approximately 240,000 years, this is but a fraction of geological time and so we might expect this process to have kept the Earth’s climate tightly regulated through time. Yet, a lot of research has shown that the Earth has experienced many global warming (and cooling) events in its past. So, was this silicate weathering feedback on CO₂ sometimes less (or more) effective in the past, and why?

The Middle Eocene Climatic Optimum

In January 2020 I joined Philip Pogge von Strandmann’s research group at University College London, to provide modelling insights into some climate change occurrences that they had been investigating using lithium isotopes. One event, and which is the focus of the accompanying paper, was the Middle Eocene Climatic Optimum (MECO for short) – a global warming event which occurred around 40.5 million years ago, lasting for roughly 400,000 years. For those of us who look at past climate change the MECO is an exciting period of Earth history because it’s a bit weird. Much like other episodes of global warming, CO₂ in the atmosphere and global temperatures increased during the MECO, the oceans also became more acidic and carbonate minerals on the ocean floor began to dissolve. But the length of these other events was much shorter, over the duration of the MECO the higher CO₂ and temperatures should have caused more silicate weathering to occur, drawing down the CO₂ and leading to more carbonate mineral formation, rather than dissolution.

Lithium - not just useful for batteries!

So far, so strange, but why use lithium isotopes to investigate the MECO? Well, as it turns out, they are a useful tracer of silicate weathering processes because secondary minerals such as clays, which can form after silicate weathering, attract the isotopically light lithium cation (⁶Li⁺), leaving the solution and thus rivers enriched in the heavier isotope (⁷Li⁺). This occurs primarily in places with large floodplains, like along the Ganges river, where lots of clay minerals are forming. However, in areas with high erosion rates, such as mountain ranges, there is little clay formation due to a short water-rock reaction time and thus the isotopic composition of rivers is low. In areas of intense weathering, such as rainforests, thick soils overlie the bedrock and therefore pre-formed clays are weathered, releasing isotopically light lithium.

This relationship between the lithium isotope value of rivers and the weathering regime, which we call ‘weathering intensity’ and is a ratio between chemical weathering and the sum of physical and chemical weathering, gives a classic ‘boomerang’ shape.

Philip had analysed marine carbonates for lithium isotopes from several deep ocean drill cores across the globe and looked at samples of the cores which spanned the MECO timeframe. What he found was incredibly exciting – the first positive lithium isotope excursion identified for a global warming event! Other warming events, such as Oceanic Anoxic Event 2 and the Paleocene-Eocene Thermal Maximum have been found to have negative lithium isotope excursions, so to find a positive one here made the MECO even more unusual.

Insights from modelling the Earth system

This is where I entered the picture. I built a biogeochemical model of the Earth system, which looked at the carbon, silicon, lithium and osmium cycles and would enable me to test various hypotheses about what might have been going on during the MECO. We had some geological evidence to suggest that a rise in volcanic activity might have been the initial trigger and I knew that this had to have occurred on the land, rather than the oceans, because a rise in hydrothermal activity would likely have given us a negative lithium isotope excursion instead of a positive one. What my modelling suggested is that before the MECO, on a global average, the land was mainly covered by thick soils which were being slowly weathered, with a limited amount of clay formation taking place. The increase in volcanic activity not only provided CO₂ to the atmosphere, but also a fresh supply of silicate minerals to the Earth surface, which were easily weathered under the warm climate, allowing for the formation of new clay minerals.

The problem is that it’s not just lithium cations which are attracted to clays, but also calcium and magnesium, and a reduced supply of these cations to the oceans potentially means less carbonate minerals forming to lock away carbon for a long time. Bicarbonate anions would still have made their way to the oceans, but evidence in the geological record and my modelling hints that clay formation in the ocean (called reverse weathering) was also enhanced and this process can use up bicarbonate anions plus even more cations, producing CO₂ in the process.

So, we think that the MECO possibly lasted for such a long time because clay formation on the land and in the oceans increased, hitting the Earth with a double whammy of more CO₂ production and less carbonate burial. The MECO might have only ended when volcanic activity slowed, reducing the amount of silicate weathering and thus new clay formation taking place, allowing other feedbacks to then kick in.

Possible implications

Circling back to the present day. There are a number of projects which aim to speed up the process of silicate weathering and remove enough CO₂ from the atmosphere to try and limit the impact of climate change. Could the processes, suggested by our study, have an impact on these climate change mitigation methods? We’re currently unsure, but rest assured lots more research on silicate weathering and clay formation is underway.