Description:
Earth observation is brilliant at bringing fantastic high-resolution imagery to your desktop within a few hours - so much so that in many cases, explaining whats going on in an instantaneous snapshot from space can be a whole PhD! With increasing amounts of data and new tools and techniques using machine learning can we agree the best way to handle these owonderful instantaneous snapshots from space? How snapshots work together with time-integrated gridded fields that are often preferred?
Whats the best way to mange upscaling and downscaling of the unique information content held within a snapshot sample from space? How much ergodicity exists in our ocean measurements from space? What are we throwing away when we time-integrate and grid our data? Can machine learning AI help us to address these issues? Do we need a vocabulary and dictionary of snapshots to help validate numerical simulations, and to reveal complex interactions?
Have your say at resolving the Earth Observation Time/Space sampling conundrum!

Speakers:
Bertrand Chapron (IFREMER, Brest, France)
Johnny Johannesen (NERSC, Bergen, Norway )
Fabrice Collard (Ocean Data Lab, Brest, France)