analytics and modeling
This may have been covered at some point however it seems likely that satellite photos, experience, and powerful analytics should work well together and I wonder if anyone has knowledge?
There should be thousands of satellite photos of sea junk in the 40s with good documentation of local weather. For other purposes I would imagine that organizations interested in oceans would have developed an algorithm that looks for relationships in this data and should develop relatively distinctive snapshots from these combinations.
Lots of junk with 10 foot waves and clouds with sun 20 degrees above horizon looks very different from no junk with sun overhead etc. Feed in enough data and it would cover practically any combination.
Then you run images of a week old aviation crash, sunk ship or other specific event and it seems likely that the profile would be significantly different. The algorithm would effectively say "haven't seen anything like that before under random circumstances. ''
The reason I ask is that we do similar process with business and finance data and law seem to be doing similar things with gang groupings.
It seems logical that shipping opperations, environmental, science types would have ocean models as I have described.
If so is it possible that no one has thought of this approach or am I missing something obvious?