The range of emerging observation candidates includes, but is not restricted to: Raman lidar, DIAL lidar, densification of the ground-based GNSS network, crowd-sourced observations, drones, wind profiler refractivity to name but a few. Some systems are considerably more expensive per unit and harder to maintain than others, but inherently provide more information content. For some, the information content may great but a large effort is required in forward modelling the observation from model fields in order to be useful for assimilation. These challenges will be familiar to all in this field of endeavour.
In broad terms, the observing requirement expressed in terms of meteorological parameters is well characterised (the WMO OSCAR database being good evidence of this). However in a cost-constrained environment the methodology for optimally fulfilling those requirements is not. Factors such as the technical maturity, inherent information content, effort to extract that information content, the ‘receptiveness’ of the DA system to make use of the information and operational considerations such as full life cost all form part of the decision making process when investing in a new observation type.
This paper will discuss ongoing work in this area in the UK Met Office and seek to engage other organisations to share experiences and best practice.