Carbon export by erosion of the biosphere: The role of earthquake-triggered landslides

Does mountain building influence the global carbon cycle and carbon dioxide (CO2) concentrations in the atmosphere, and hence modify Earth's climate? This question is at the heart of a long-standing debate as to how erosion and weathering act to draw down CO2, countering the input of CO2 from volcanoes over millions of years. Earthquakes offer a direct way of removing CO2 from the atmosphere, but how this happens and how large the impact might be has remained poorly understood. Ground shaking during earthquakes can trigger tens of thousands of landslides, which strip large amounts of carbon from soil and plants in mountain forests. Depending on the amount of carbon involved, and what happens to it, earthquakes could be a way in which mountain building drives the global carbon cycle. If the carbon removed from mountain forests reaches rivers, it can be transported in muddy waters downstream and into a long-term sink (storage) of CO2 following burial in sediments.

Our research aims to quantify, for the first time, whether large earthquakes could increase the erosion and river export of carbon from mountain forests. Until now, it has proved very challenging to measure the impact of these extreme and unpredictable events in river systems. This is because we need samples collected from rivers before and after earthquakes, in order to use geochemical measurements to fingerprint the carbon source and measure the carbon flux. The research team has recently been involved in the only case where this has been done, after the 2008 Wenchuan earthquake in China which triggered over fifty thousand landslides. There, we found that the carbon flux in a mountain river increased significantly in the four years which followed the earthquake and the associated catastrophic landsliding. Our work from Wenchuan sets the foundation for this research proposal, demonstrating that carbon mobilised by earthquake landslides does reach rivers. However, we expect the impacts to last for ten years, or even hundreds, and Wenchuan is just a single example. It is clear we need other data, and a new approach in order to fill this research gap.

We will study multiple large earthquakes, and make measurements over decades to centuries before and after each earthquake. To do this, we will combine landslide maps from historical earthquakes around the world, with some of the best studied records of sediment export following large earthquakes in lakes. The well-dated and well-understood lake sediment records come from the western Southern Alps, New Zealand. They record the response of the mountain landscape to four large earthquakes over a thousand years. We will use geochemistry techniques to fingerprint and track the carbon sourced from vegetation and soil (rather than eroded from bedrock) and combine these with measurements of sediment flux to calculate carbon flux. This can be done for four earthquakes in two lakes, as well as during the 'background' periods before and long after the catastrophic landsliding. The datasets will allow us to confidently quantify how earthquakes increase carbon export from forests from a 'background' state.

We will use the new data and our understanding of the main processes operating to build a model to allow us to assess the role of earthquakes for carbon flux in mountains on longer timescales. For the first time, we will be able to apply the model around the world to mountains which experience earthquakes. We will account for changing earthquake size and how often they happen, and the amount of carbon in the forest and the rate at which it is degraded in landslide deposits. We will also consider how different rainfall patterns in river catchments can change the flux of carbon. With these novel insights, we will be able to quantify how earthquakes impact the carbon cycle, CO2 and the evolution of Earth's climate.

There is a growing awareness of the importance of understanding the prolonged geomorphic hazard associated with earthquakes. The research outputs from this proposal will include a data-driven numerical model which will allow us to quantify the impact of earthquakes on carbon (and fine sediment) transfer in river networks, grounded in our new datasets from multiple earthquakes around the world. These outputs will allow us to constrain the longevity of impacts and the patterns of impacts upstream. These assessments have the potential to benefit agencies tasked with better understanding of the response of river networks to large earthquakes in mountain belts around the world. Suspended sediment load response has important implications for hydro-engineering and reservoir management, and can have negative impacts on floodplain, freshwater, estuarine and coastal resources and ecosystems downstream for hundreds of kilometres downstream.

In New Zealand, which has a long history of large earthquakes, we have identified the key agencies which are likely to benefit from our research, including the Crown Research Institutes the Geological Nuclear Sciences (GNS), National Institute of Water and Atmospheric Research (NIWA), and stakeholders including the agricultural industry, regional councils, and the Department of Conservation. To allow stakeholder input, a workshop in the first year will develop discussion on their data needs and the research questions and deliverables which will most directly benefit these users. Following this workshop, we will be in a strong position to engage efficiently as our measurements become available. Face-to-face meetings with stakeholders will bookend the project and allow us to refine and adapt our academic outputs to provide bespoke resources and solutions which benefit these wider users.

To help guide our research outputs and engage beneficiaries during the research grant programme, we will also organise a three day workshop in Year 2 of the grant. Entitled 'Erosion and the Carbon Cycle', we will invite world-leaders to present and discuss their research, considering geomorphological, geochemical, and modelling perspectives, while advertising widely amongst the UK, EU and international community. The aim will be to guide the research outputs to best meet the user's needs during the project, while fostering new collaborations. Wider academic dissemination of the research findings will be through Open access, peer-review publication of the research. In addition, results from each of the objectives will be presented at two international conferences during the project (Year 2 European Geosciences Union Annual Meeting 04/2018, Year 2/3 American Geophysical Union 12/2018). A themed session will be organised and chaired at AGU to promote discussions, initiate future collaborations and identify future research aims and projects. The PI and Co-Is have a strong track record in these activities.

Public interest in the carbon cycle has increased over recent decades. This research addresses an important component of the natural carbon cycle, which is poorly-constrained at present, so the findings should be of interest to this wider non-academic group. In addition, the broad range of academic beneficiaries, crossing fields of geochemistry and geomorphology, will also generate interest outside the formal outputs and deliverables. For these reasons, we will create a project based webpage. This will explain the research team expertise; publicise papers with non-specialist summaries; allow wider access to conference presentations; and explain the fieldwork and ongoing research (and when appropriate preliminary findings). In addition to the webpage, we will make use of social networking outlets to reach a wider audience (e.g. @geoCcycle). Durham University and Otago University promote research-led teaching and the findings will be incorporated into undergraduate/postgraduate courses.