Timing application of dormancy breaking products is essential for adequate bloom and yield in cherry. Available temperature-based chill accumulation models often fail to predict correct spray timing, particularly during years with unusual temperature patterns recently driven by climate change. One potential cause for the unreliability of the current models is their dependence on a single climatic variable, air temperature, to estimate the impact of weather on tree physiological processes. We hypothesized that using tree temperature, rather than air temperature, will increase the accuracy in predicting chill accumulation, especially in warm and sunny winters. Hence, our objective was to develop a framework to predict cherry tree temperature based on easily available environmental data. For three consecutive seasons and in three commercial cherry orchards across California, including the southernmost US cherry production region, we measured main climatic parameters at half hour intervals. At the same time, we monitored tree temperature with T-type thermocouples inserted below the bark of main branches. We developed predictions of tree temperature as a function of meteorological variables obtained from public weather stations using generalized additive models. Trees were, on average, 10 ºF warmer than the air during clear days, with differences being up to 20-25 ºF. Tree chill accumulation was about 8-12 chill portions lower than air chill accumulation. The difference was year- and site-specific, reflecting the importance of including diverse environmental parameters to estimate chill accumulation precisely. The ‘TreeChill model’ predicts tree temperature based only on environmental parameters easily achievable from public weather stations with a coefficient of determination of 0.956 resulting in only 0.4 chill portions difference between measured and predicted tree chill. This model will enable growers to implement tree temperature in their management decisions, including dormancy breaking agent applications, cultivar selection, pest control etc, increasing California cherry industry resilience to climate change. In the future, we plan to adapt the model to different crops and locations.
