DART is created by the Data Assimilation Research Section (DAReS), a small group of data assimilation experts, software engineers and physical scientists. DAReS was founded as the result of NSF NCAR's Data Assimilation Initiative, a collaborative process that gauged the needs of researchers across NSF NCAR. The initiative established the necessity of developing methodologies, tools and software to enable data assimilation capabilities within models developed at NSF NCAR.

DAReS staff serve projects at NSF NCAR and within the university community by providing a bridge between teams with existing observational and modeling expertise. This aids model development by describing model error and enables fundamental discoveries by allowing observationalists and modelers to share their expertise by combining models and data using a rigorous assimilation framework.

In the years since DAReS’ founding, DART has grown beyond its original purpose and now supports data assimilation in models developed at NSF NCAR and by the broader geosciences community. DAReS staff have expertise in algorithm development, software design, software portability, large-model implementation and execution, observations and observation operators.