This project will improve predictions of the impacts of Amazon deforestation on local and regional climate.
Deforestation in the Amazon causes large changes to the land surface, disrupting land-atmosphere interactions, warming the local climate and reducing regional rainfall (Spracklen et al., 2012; Baker and Spracklen, 2019). Deforestation projected over the next few decades may increase local temperatures by a few degrees and reduce regional precipitation by 40% (Spracklen and Garcia-Carreras, 2015), increasing the likelihood of major forest fires and potentially leading to further forest dieback (Butt et al., 2021). Increases in runoff caused by deforestation have the potential to compound the severity of major flood events.
Despite these important impacts, the complex biophysical changes that accompany deforestation are not fully understood. Climate model representation of these complex changes is poor and predictions of the impacts of deforestation on local and regional climate are very uncertain. For example, the UK climate model overestimates the albedo-driven cooling and underestimates the latent-heat driven warming impacts of deforestation, leading to a large cooling bias (Robertson, 2019). Many climate models are unable to reproduce the correct controls on Amazon evapotranspiration, with implications for regional precipitation projections (Baker et al., 2021). It is possible that models that are unable to accurately represent land-atmosphere interactions over the intact Amazon would be unlikely to accurately simulate the climate impacts of land-use change, but this has not yet been tested. Furthermore, trade-offs between climate model complexity and the computational costs required to run them drive a need to fully understand the benefits gained by increasing model resolution and including intensive processes such as explicit convection.
To address these issues, this studentship will provide a detailed evaluation of deforestation-induced changes to the land surface, surface fluxes, and land-atmosphere feedback pathways over the Amazon in global and regional climate models. The project will assess the credibility of the simulated land-use change contribution to local warming and regional climate change over the next century. New understanding from the project will support policies to protect and sustainably manage tropical forests.
1) How do climate models represent the climate response to tropical land-use change, with a focus on the Amazon?
2) How does the representation of land-atmosphere interactions depend on model resolution and complexity?
3) How will projected Amazon land-use change impact local and regional climate and how robust are model predictions?
This project will assess model representation of the biophysical changes that occur following land-use change in the Amazon, using models of different resolutions and complexity, and so inform our understanding of future deforestation-driven regional climate change. The research would exploit existing and planned simulations from global and regional climate models from the UK and Brazil, together with simulations from the latest generation of climate (CMIP6) models. Specifically, simulated changes in the land-surface state, surface runoff, fluxes of energy and moisture, and land-surface feedbacks on the atmosphere following deforestation would be examined, and compared with remote sensing observations, station data and state-of-the-art reanalyses.
The project will provide training in climate science, land use change, numerical modelling and analysis. During the project you will learn to use and analyse climate models and remote sensed data. The Met Office are supporting this studentship as a CASE Partner. This will provide opportunities to work closely with Met Office scientists and provides in-depth training with the UK climate model. We encourage and support our students to publish and communicate their work to a variety of stakeholders including non-governmental organisations (NGOs), businesses and relevant governments. You will become a member of the Biosphere Atmosphere Research Group (BAG) and join the Leeds Ecosystem, Atmosphere and Forest (LEAF) centre at the University of Leeds.
Potential for Impact
This studentship tackles a key environmental challenge – the impact of large-scale tropical deforestation on the Earth’s climate. New commitments made during COP26 to end deforestation by 2030 provide new urgency to better understand the impacts of deforestation on regional climate. The studentship will improve our understanding of this complex system and improve representation of land-atmosphere interactions in the UK Earth System Model, which in turn will help inform decisions to support climate-resilient economic development and social welfare. The project will inform our assessment of carbon budgets for climate goals, particularly of the effectiveness and impacts of land-based mitigation and adaptation. The project will inform the development of a range of models including JULES/UM/LFRic. The studentship aligns closely with CSSP-Brazil (Climate Science for Service Partnership – Brazil), a collaboration between UK scientific institutes (including the UK Met Office and the University of Leeds) and Brazil’s National Institute for Space Research (INPE), National Institute for Amazon Research (INPA) and the National Centre for Monitoring and Early Warning of Natural Disasters (CEMADEN). These links are critical to ensuring that the project helps deliver environmental solutions to tropical land use change. Close links with a range of Brazilian stakeholders, ranging from research institutes to government departments, will help ensure that new knowledge created by this project helps to craft new solutions to tropical deforestation.
A good first degree (1 or high 2i), or a good Masters degree in a physical, mathematical or geophysical discipline, such as mathematics, physics, engineering, environmental sciences or meteorology. Enthusiasm for environmental research is crucial and experience in programming, numerical modelling or data analysis (e.g., Python) is desirable.