Modelling the impact of climate change on the occurrence of frost damage in Sitka spruce (Picea sitchensis) in Great Britain

Published: 09-11-2020| Version 2 | DOI: 10.17632/4vvdtr9597.2
Anastasia Atucha Zamkova,
Andy Smith,
Katherine Steele


Climate change is predicted to increase temperature and seasonal temperature variance in Great Britain (GB). Sitka spruce (Picea sitchensis (Bong.) Carr) is the most important tree species used in commercial plantations throughout Europe and GB. Frosts that occur outside the winter dormancy period can negatively affect trees, since they happen after dehardening. Damage can be especially severe at bud burst, before emerging needles mature and form protective barriers. Here, we modelled the impact of climate change on frost sensitivity in Sitka spruce with temperature data from the UKCP09 climate model HadRm3 that uses emission scenario SRESA1B for the years 2020-2099, the global and downscaled versions of the UKCP18 HadGem3 model that uses the emissions scenario RCP8.5, and the global model CMCC-CM for the RCP 4.5 and RCP 8.5 emissions scenarios, and compared the results with gridded historical data for the period 1960-2015. Three indicators that assessed the frost sensitivity of Sitka spruce were explored: the total number of frosts during the growing season (Index 1); the total number of frosts after bud burst (Index 2); the number of days with minimum temperatures below the resistance level (backlashes) during the growing year (Sep-Aug) (Index 3). The indices were validated with historical data for frost damage across GB, and Index 1-3 °C and Index 3 were shown to be significantly correlated. The frequency of all frosts and backlashes is expected to decrease with climate change, especially under higher emissions scenarios. Post bud burst frosts have been historically very rare in GB and remain so with climate change. Downscaled regional climate models detect geographic variability and improve prediction of overall trends in frost damage in comparison to global climate change models for GB. Code and results available here. The public datasets used to calculate these indices are available at the following links: Climate change data: Polygons that define shape of provenance region and seed zones: Historical dataset: