Neuroemulation: definition and key benefits for water resources research

摘要

Neuroemulation is the art and science of using a neural network model to replicate the external behaviour of some other model and it is an activity that is distinct from neural-network-based simulation. Whilst is has become a recognised and established sub-discipline in many fields of study, it remains poorly defined in the field of water resources and its many potential benefits have not been adequately recognised to date. One reason for the lack of recognition of the field is the difficulty in identifying, collating and synthesising published neuro-emulation studies because simple database searching fails to identifying papers concerned with a field of study for which an agreed conceptual and terminological framework does not yet exist. Therefore, in this paper we provide a first attempt at defining this framework for use in water resources. We identify eight key benefits offered by neuro-emulation and exemplify these with relevant examples from the literature. The concluding section highlights a number of strategic research directions, related to the identified potential of neuroemulators in water resources modelling.