CBER Seminar - Dr Tadhg Carroll, UCL

Abstract: I will present insights from a ‘UCL Climate Crisis Grand Challenges’ working group –comprising a cross-disciplinary team of ecologists, statisticians and climate scientists – aimed at integrating skills and knowledge required to improve scientific inference on effects of climate on biodiversity under substantial limitations in available data. Effective management of the ecological impacts of climate change will require a detailed understanding of how climate interacts with other environmental drivers (e.g., human land use), relying upon the integration of data from ecology, earth systems sciences and related fields. I will discuss the types of limitations inherent in such data – including gaps, biases and uncertainties which vary widely across space, time, and data types and sources – and resulting inferential challenges which arise (e.g., problems around missing data, measurement error, and prediction in ‘non-analog’ environmental space). I will outline a workflow for robust inference which comprises: the identification of ‘actionable evidence’ required to inform management and policy, the precise specification of inferential targets, and a principled consideration of the range of data generating processes that produced data underlying an analysis. Finally, I will present an empirical example attempting to apply some of these principles to assess effects of climate/land-use interactions on the bird species occurrences.