HYDRAS combines data at different spatial scales, frequencies and of different plant organs, soil compartments and atmospheric conditions. The drone imaging and belowground ERT time series data are a unique source of information. To leverage these data to its full potential, we develop a hybrid modelling approach, where data-driven sub-models can be connected to process-based plant models that link soil and plant hydraulics to plant growth and development. This enables the quantification of dynamic traits, the disentanglement of complex plant responses and estimation of cultivar-specific model parameters. Consequently, it improves the understanding why varieties perform differently in different conditions.