Weather/Climate Data Collection for Large-Scale Phenotype Predictability in the Midwest

A consistent increase of crop yields in the Midwest reflects the role technology plays, through the development of new crop varieties, irrigation technologies, and improvement of management practices. New proximal and remote sensing technologies may also lead to sustain such increase in productivity. However, the multidimensionality, heterogeneity and volume of data pose serious challenges and opportunities to the breeding community and their path to predict phenotypes. The Genomes to Field experiment is leading a regional-scale effort to improve crop performance in response to environmental changes and their interactions with crop genomes. The present work aims to define the conceptual framework to collect, store, manage and use weather/climate data from local to global datasets to forecast plant phenotypes. In particular, we will describe the development and performance of an architecture of software based on a plugin concept for the collection of multi-source data that can be eventually transformed via genetics-by-environment analytics into crop performances.