Oh dear, what have I started... Anyway, as I see it there are lots of feedbacks in this vein, but the effects of just about all of them are individually minimal:
Positive:
- Soot production
- Direct mechanical destruction of ice sheets by icebreakers
- Indirect mechanical disruption of ice and facilitation of warming by the wake of ship traffic
- Disruption of refreezing by the wake of ship traffic
- Local heat production of ship traffic and new arctic stations, e.g. drilling platforms, floating nuclear power plants, expanded harbours etc.
Negative:
- Less carbon-intensive shipping (no great effect, as there may well be more shipping overall, and the icebreaker escort requirement cuts into the saving).
Such minimal impact as this would have would mostly be when the ice has retreated enough to allow ship traffic, so in the months of June to November or so. If I were to speculate, I'd reckon it might have more of an impact on stopping the annual refreeze than on facilitating the melt (thinking about the effects of the turbulence caused by the ships passage), hence contributing very slightly to how late in the year the refreeze begins. Even the
1.5 million or more tons of cargo that might be shipped this year and the ships required to carry that are pretty small potatoes compared to the size of the ocean they're in. But they may well keep the northern sea route itself open a little longer, albeit without generally impacting the rest of the ocean.