Satellites can now measure sea ice density throughout the Arctic Ocean throughout the year.

Traditionally, spacecraft struggle to determine the full state of the flies during the summer months because the presence of surface meltwater has confused their instruments.

But using new "deep learning" techniques, scientists have overcome this limitation to obtain reliable observations in all seasons. These advances have far-reaching implications.

Apart from the obvious advantage of ships, which need to know the parts of the Arctic that will be safe to navigate, there are significant advantages of climate and weather forecasting.

At the moment, there is considerable variation in estimates of when the polar oceans may become completely ice-free in a warming world.

Having an improved insight into melt processes in area and thickness should now sharpen the output from computer models during those critical months when flows are being reduced.

"Despite the excellent efforts of many researchers, these climate models predict when we will first see a completely ice-free Arctic Ocean in summer - they vary by more than 30 years," Dr. Jack Landy, 

UiT Arctic University of Norway told BBC News.

"We need to tighten up those predictions so that we're much more confident about what's going to happen and when — and how climate responses will accelerate as a result."