Identifying drivers of oak damage during Storm Arwen with Dr Kate Halstead
In late November 2021, Storm Arwen brought northerly winds exceeding 40 metres per second to the UK, with the worst damage recorded in Northumberland and Scotland. The storm caused widespread damage to homes, infrastructure and woodlands, with oak trees among the most affected and many notable and veteran specimens lost. A second storm, Barra, followed soon after, though its impacts were far less severe.
Together these events raised an important question: what makes some oak trees more vulnerable to storm damage than others?
Windthrown oak at the Wallington Estate
My recent research, published in Agricultural and Forest Meteorology, set out to address this question. Working with colleagues from Newcastle University, Forest Research and Northumbria University, we combined fieldwork, tree-ring analysis, remote sensing and statistical modelling to identify the main drivers of storm damage to oak during Arwen and Barra.
The study focused on three sites in the North East of England, one of the regions worst affected. At Wallington Estate, Cockle Park Farm and Gosforth Nature Reserve we recorded oaks that had been damaged alongside those that remained intact. By comparing damaged and undamaged trees at the same locations, we aimed to identify the factors that made the difference. Tree structure was measured using terrestrial laser scanning, canopy health was assessed from satellite imagery, growth patterns were analysed from tree-ring cores, and field inspections recorded visible disease symptoms and structural defects.
L-R. 1. Damaged oak tree cores from Wallington, used to measure ring widths and calculate average growth rates over a 15-year period
2. 3D point cloud of storm damaged oak, taken using a GeoSLAM ZEB Horizon LiDAR scanner, from which measurements of tree structure e.g. DBH, height, crown area, were extracted
3. Another windthrow oak at the Wallington Estate
The results showed that individual tree characteristics were much more important than broader site features in explaining storm damage. Structural defects such as weak forks, cavities and old pruning wounds were strongly associated with failure, and trees with visible disease symptoms were also more likely to be damaged, both because infection and decay weaken the wood itself and because they can reduce growth and overall vigour over time.
Growth rates provided another important signal, with damaged trees having grown more quickly in the 15 years before the storms, on average around 23 per cent faster than those that remained undamaged. Although rapid growth can reflect favourable conditions, it is often linked to lower wood density, which in turn reduces mechanical strength and increases vulnerability to breakage. By contrast, wider site conditions such as slope, elevation and soil wetness, which are frequently highlighted in commercial conifer wind risk studies, appeared less influential for the scattered oaks in this work. This is likely to reflect the relatively small scale of the study as well as the different growing contexts of oak compared with plantation species.
We also compared two modelling approaches. Random Forest, a widely used machine-learning method, ranked variables according to their predictive strength, whereas Structural Equation Modelling, a hypothesis-driven approach, allowed us to test both direct and indirect pathways between factors. Both approaches identified structural defects and disease as key predictors of storm damage, but the Structural Equation Models revealed additional relationships, such as the way disease symptoms were linked to reduced growth, which in turn increased vulnerability. Capturing these indirect links provides a more complete understanding of how multiple stressors combine to affect storm risk.
The implications of these findings are significant. Oaks are of exceptional ecological and cultural importance in the UK, supporting thousands of associated species and holding centuries of history in our landscapes. Hence, protecting them against the growing risks of climate-induced storms is a priority. Our results suggest that careful monitoring of tree health and structure is particularly important, as identifying oaks with visible defects or signs of disease may help managers anticipate which individuals are most vulnerable. While storms cannot be prevented, improved understanding of risk factors can support more informed conservation and management decisions. The next step is to aid the development of wind risk models such as ForestGALES so that they better represent oak, particularly in parkland settings. Our results indicate that structural defects, disease symptoms and growth rates are stronger indicators of vulnerability than broad site features, and incorporating these factors, as well as accounting for non-prevailing storm tracks like Arwen, would strengthen predictions for oak across parkland, hedgerows and other open-grown settings.
“Oaks are of exceptional ecological and cultural importance in the UK”.
Structural Equation Model showing which tree characteristics are positively (green) or negatively (red) related to storm damage in oaks_ numbers on the lines show the strength of each relationship
For managers and owners, the findings underline the importance of practical monitoring, especially of very old trees of interest such as ancient and veteran oaks. Inspections that record defects such as weak forks, cavities or pruning wounds, alongside visible signs of disease or decline, provide the clearest indication of which trees may be most at risk. Focusing attention on these vulnerable individuals offers a realistic way to anticipate storm impacts and to plan management that better protects oak across sites.
Storm Arwen was described as a once-in-a-generation event, yet under climate change such storms are becoming more frequent and severe. By identifying the factors that predispose oak trees to damage, this research provides a foundation for more effective modelling and for practical steps that can strengthen resilience. The damage from Arwen shows that storm impacts are not determined by location alone; the history of each tree, its growth patterns, its health and the structural weaknesses it carries all contribute to how it responds to extreme weather. Recognising and understanding these factors will be key to helping oak withstand the storms of the future.
This research was carried out as part of a 3.5-year PhD project at Newcastle University, funded by Action Oak and supported by Forest Research.
If you are a site owner or manager and have experienced storm damage to oak trees (whether from the storms studied here or more recently), we would be very interested to hear about your experiences. For more information, please contact Dr Kate Halstead at kate.halstead@forestresearch.gov.uk