Assessing individual and local scale forest vulnerability to mortality from the 2019 extreme drought in Central Europe

Despite the recognized effects of intense droughts and increased temperatures on tree mortality, the die‐off patterns observed worldwide are poorly reproduced by global vegetation models. This problem is exacerbated by our limited understanding of the drivers involved in drought-induced growth decline and mortality and the temporal and spatial scales at which they operate. In turn this limits our ability to identify which forest areas are most vulnerable to drought induced growth decline and mortality, information that is vital to plan for conservation of biodiversity and for effective forest management. Local variation in drought susceptibility of individuals within a species is strongly influenced by local scale environmental variation, patterns of local adaptation and, critically, legacies of past management and stress exposure on tree density and architecture. However, while we now recognise the importance of such factors in driving deviation from model predictions, we lack detailed local data to drive improvements in model development. There is, therefore, an urgent need to understand the structural and environmental factors that are associated with local within-population variability in tree response to drought to enable the next generation of bioclimatic models to be adequately parameterised. We will integrate high-resolution satellite imagery, UAV-mounted hyperspectral and LiDAR remote sensing data, and targeted field-based measurements to predict individual and local scale vulnerability to drought events. In this project we will: i) collect ultra-high resolution spectral and structural data using the headwall system from NERC Field Spectroscopy Facility to quantify tree canopy condition, derive structural measurements and create detailed terrain models; ii) combine UAV-based spectral and structural measurements with detailed field-based assessments to determine the structural and environmental correlates of individual and local vulnerability and iii) predict the vulnerability of forest stands using UAV and satellite remote sensing data in the wider region.