The current state of understanding of upper ocean processes such as fronts, upwelling, plumes, blooms and anoxic zones, among others are poorly understood at the causal level. This means that at its core, we have a heuristic approximation as to the causation of many of these naturally occurring phenomenon yet have poor to no conception of being able to even attempt to model them over space and time.
Typically we observe such phenomenon either opportunistically, and/or with remote sensing capabilities; point measurements via traditional ship-based observations, then provide an approximation to hypothesis driven models which may or may not be symptomatic of actual causation.
While traditional methods have proven to be accurate in many respects, the dearth of data using time consuming and expensive ship-based methods lead to the paucity of data, especially for time-series. Studying the onset of climate change requires science to establish a solid basis of understanding of processes which can both describe the past, and hope to provide a model for what to expect in the (near) future. With no established means to sample the range of oceanographic phenomenon, over the global range, we have little to no recourse but to use remote sensing and/or couple with new autonomous robotic platforms to provide a better perspective of at least the photic zone (upper 75m). And at least in coastal zones. Yet this too has proven to be challenging at best.
Robotic vehicles such as AUVs (autonomous underwater vehicles), ASVs (autonomous surface vehicles), UAVs (Unmanned aerial vehicles) are now proving to be more robust with longer observational capacity with a range of sensors. Coupled with advances in space borne platforms and sensing, including very cost-effective small satellites, it gives scientists the potential capability to obtain data at finer scales, systematically over longer periods of time.
Yet with such new platforms available to us, the fundamental issue of quantified resources still remains for our asking: “when and where (and how) we should sample” and model the water column.
To answer these challenging questions, we propose to bring together a highly inter-disciplinary group of people across fields which typically do not mix and meet: namely Statistics and Sampling Theory, Oceanography, Control, AI, Ocean Optics, Remote Sensing among others.
Topics of discussion (but not limited to):
Oceanographic modeling
Sampling
Information Theory
Control Theory (including deliberative AI-based control)
Machine Learning
With such an event, we aim to:
promote a better understanding of the fundamental problems in upper water-column ocean observation across disciplinary lines
determine tools/techniques/process already in place across disciplines which could be applied towards such an observational capability
promote the mixing of such ideas in a very relaxed and informal setting in Norway — an ocean facing nation which takes its fiduciary responsibility seriously
encourage and promote the dissolution of rigid disciplinary thought and action by encouraging people to write collaborative proposals across national boundaries
and to do in a relatively small, cozy and intimate event where discussions are well motivated and driven by a desire for societal impact