Using novel simulations to quantify human influence on global climate variability

This exciting PhD project tackles an important topic in the field of climate change research. Observed climate change signals are modulated by internal climate variability. Internal variability is an innate property of the climate arising from complex and nonlinear interactions between different climate system components (e.g. the atmosphere and ocean) and manifests globally across a wide range of timescales.

It has been proposed that man-made climate change has altered the nature of internal climate variability, including key modes like the El Nino Southern Oscillation (Cai et al 2021), Atlantic Multidecadal Variability and Pacific Decadal Variability (Smith et al 2016). This is a timely and topical project as recent high profile research has suggested that the current emergence from persistent La Nina conditions into a strong El Nino year could be connected to climate change (Wang et al 2023).

However, there are many open questions regarding the mechanisms for such connections and whether any changes in climate variability due to human influence could be detected in observations. Major advances are needed to pin down the interactions between forced climate change and internal climate variability to support the interpretation of observed climate records and to support the production of informative projections of future climate change.

Aims and objectives

The overarching goal of this exciting PhD project is to isolate and quantify the effects of external climate forcings on internal climate variability. The student will achieve this by exploiting a new, unique dataset comprising state-of-the-art climate model simulations and applying advanced data science methods to identify signals.

- What are the human fingerprints of external climate forcings (e.g. greenhouse gases, aerosols, natural factors) on global modes of internal climate variability? - Are there fingerprints of human influence on internal climate variability that could be detected in observations?

- How will internal variability change under future climate change and which forcings are key to these changes? How do these changes affect future projections over the coming decades? Methods

The project will exploit the brand new Large Ensemble Single Forcing Model Intercomparison Project dataset (LESFMIP; Smith et al 2022). This provides simulations from multiple climate models in which the effects of specific external forcings are isolated. A large initial condition ensemble is performed for each case, whereby multiple realisations are performed with small differences in initial conditions creating different manifestations of internal variability (i.e. the "butterfly effect").

This design is ideal for examining internal variability because it provides detailed sampling of climate phase space. Machine learning will be used to identify altered patterns of climate variability associated with different external forcings. The student will use signal processing methods to ascertain when the signals of augmented internal variability will be detectible in observations.

They will work alongside a co-supervisor at the UK Met Office and have the opportunity to visit and engage with Met Office scientists.