Data Assimilation for the Coupled Earth System (Invited Presentation)

The field of data assimilation has made great strides since the early days of numerical weather prediction (NWP). The operational pull of improving prediction of the atmosphere on time scales from hours to days has led to considerable advances in data assimilation for initialization, reanalysis, multiple satellite data streams, and observing system simulation experiments. Following the lead of the atmosphere, ocean and land surface assimilation has made significant progress over the past several decades. However, coupled data assimilation across the atmosphere-ocean-land system has lagged. This presentation will provide some historical perspective on assimilation of observations of the atmosphere, ocean, and land surface and discuss the untapped potential for coupled data assimilation of the Earth System. The talk will conclude with other emerging avenues of data assimilation such as the role of machine learning, data from drones and autonomous vehicles, hyper-local observations, the urban environment, and the Internet of Things, as well as software infrastructure that can scale with complexity, partners, and innovations.