A fundamental challenge to the production of climate-resilient crops is how to measure dynamic yield-relevant responses to the environment, such as growth rate, at a scale which informs mechanistic understanding and accelerates breeding. The timing, duration and architectural characteristics of inflorescence growth are crucial for optimising crop productivity and have been targets of selection during domestication. We report a robust and versatile procedure for computationally assessing environmentally-responsive flowering dynamics. In the oilseed crop, Brassica napus, there is wide variation in flowering response to winter cold (vernalization). We subjected a diverse set of B. napus accessions to different vernalization temperatures and monitored shoot responses using automated image acquisition. We developed methods to computationally infer multiple aspects of flowering from this dynamic data, enabling characterisation of speed, duration and peaks of inflorescence development across different crop types. We input these multiple traits to genome- and transcriptome-wide association studies, and identified potentially causative variation in a priori phenology genes (FLOWERING LOCUS C, EARLY FLOWERING3) for known traits and in uncharacterised genes for computed traits. These results could be used in marker assisted breeding to design new ideotypes for improved yield and better adaptation to changing climatic conditions.