Lightweight Data Systems
An oceanographer walks into a bar and says to the bar tender ‘I’ve become so adept at generating data that I can’t keep up with it! What can I do?’ I recently co-hosted and led a workshop at the University of Washington on this topic, one that will lead to follow-on workshops along the west coast and eventually all over the world. This meta-tech work is intended to present many paths to solving this fundamental and recurring problem on the scale of a small research budget and a limited amount of time to get a working solution that includes data sharing.
Parker MacCready at the University of Washington has modified the Regional Ocean Modeling System (ROMS) computational fluid dynamics model to run every 24 hours and produce a 72 hour ocean state forecast. At Microsoft Research I have built prototype cloud middleware that translates this forecast into easily-consumable information; and as well we have built example Clients. The ultimate aim is to provide shellfish growers and others in the marine food industry of Washington State with actionable information on a regular basis.