A statistical overview and perspectives on data assimilation for marine biogeochemical models

作者：Dowd Michael^*; Jones Emlyn; Parslow John

来源：Environmetrics, 2014, 25(4): 203-213.

DOI：10.1002/env.2264

• data assimilation
• ecological modelling
• state space models
• dynamic models
• Bayesian hierarchical
• particle filters

Marine biogeochemistry refers to the processes associated with the planktonic ecosystem of the ocean. These are central to nutrient, carbon, and energy cycling, as well as providing the basis of the marine food chain. The field is being revolutionized by new data types and observing platforms, as well as by improvements in ocean modelling brought about by increasing computer power. To further our understanding of these systems, statistical estimation and inference are needed to combine the information in these data with dynamic models to provide improved estimates for the ocean%26apos;s biogeochemical (BGC) state and its parameters. Such methodologies are termed data assimilation (DA). This paper seeks to provide an overview of DA for the emerging area of marine BGC modelling. A statistical framework is offered, and DA methods that are applicable to the spatio-temporal dynamic models and data that define the BGC problem are reviewed. In addition to this primer on current BGC DA approaches, we offer our perspectives on the challenges and future work necessary to advance this field. This work emerged from a symposium on marine BGC DA that took place in Hobart, Australia, on 28-30 May 2013.

• 出版日期2014-6
• 单位CSIRO