Data from: Using archived television video footage to quantify phenology responses to climate change

Predicting how the timing of cyclic life‐history events, such as leafing and flowering, respond to climate change is of paramount importance due to the cascading impacts of vegetation phenology on species and ecosystem fitness. However, progress of this field is hampered by the relative scarcity, and geographic and phylogenetic bias, of long‐term phenology datasets.
By taking advantage of archived television video footage, we here developed an innovative tool using previously unexploited records to build long‐term datasets of phenological responses. To demonstrate the potential of this method, we worked with broadcast archives of sport events and focus on one of the most famous professional road cycling races world‐wide, the Tour of Flanders.
After viewing >200 hr of film, we compiled 523 individual × year observations of leaf‐out and flowering of 46 individual trees and shrubs visible in four decades (1981–2016) of video footage. We detect surprisingly strong advances in the timing of tree leaf‐out and flowering in the footage: trees almost never had flushed at the time of the spring race in the 1980s while significantly more individuals had flushed in the video footage between 2006 and 2016 (probabilities of leafing and flowering increased by 19% and 67%, respectively). These shifts were most strongly related to January–March temperatures and growing‐degree hours (cumulative heat) in the preceding months.
We demonstrate that this technical advance offers key benefits to fill gaps in existing phenology time series and reveal that archived video footage can indeed be applied to determine species‐temperature relationships with high spatiotemporal resolution. Only by compiling more data from the past will we be able to further our understanding on the effects of climate change on species and ecosystems in the future.