Search for: All records

Abstract This manuscript shares the lessons learned from providing scientific computing support to over 600 researchers and discipline experts, helping them develop reproducible and scalable analytical workflows to process large amounts of heterogeneous data.When providing scientific computing support, focus is first placed on how to foster the collaborative aspects of multidisciplinary projects on the technological side by providing virtual spaces to communicate and share documents. Then insights on data management planning and how to implement a centralized data management workflow for data‐driven projects are provided.Developing reproducible workflows requires the development of code. We describe tools and practices that have been successful in fostering collaborative coding and scaling on remote servers, enabling teams to iterate more efficiently. We have found short training sessions combined with on‐demand specialized support to be the most impactful combination in helping scientists develop their technical skills.Here we share our experiences in enabling researchers to do science more collaboratively and more reproducibly beyond any specific project, with long‐lasting effects on the way researchers conduct science. We hope that other groups supporting team‐ and data‐driven science (in environmental science and beyond) will benefit from the lessons we have learned over the years through trial and error.