Growth of alpine grassland will start and stop earlier under climate warming (dataset)

Data used for a study that investigated how an experimentally prolonged growing season affects growth dynamics in typical alpine grassland. Despite more than doubling the time available for plant growth, these alpine plants mostly stopped growth and entered senescence after a seemingly pre-defined time. Such cessation of seasonal activity (above- and belowground) falsifies projections of higher productivity based on future prolonging of the growing season.

We provide data and code for statistical analyses (data and scripts.zip) as well as the trained weights of the neural network and the annotated images used for training the network (CNN_model.zip). Update 03/2025: Added the photos and annotations used for validation of the model (CNN_validation_images.zip). Please feel free to use our trained model weights for transfer learning or the annotated source images as a base to train your own network.

Roots were segmented with an adapted version of U-Net (Smith et al., 2020). Please refer to this publication for instructions on how to setup the code:

Smith, A. G., Petersen, J., Selvan, R., & Rasmussen, C. R. (2020). Segmentation of roots in soil with U-Net. Plant Methods, 16(1), 13–27. https://doi.org/10.1186/s13007-020-0563-0