Briefly, satellites can do a lot for you, but they can't do everything that most people think they can do.
The current system that does provide limited wide area surveillance of the Arctic is RADARSAT-2. It provides all-weather capabilities 24 hours a day, seven days a week.
The planned deployment of our RADARSAT constellation for an additional three satellites will boost the coverage or exposure time, the ability to revisit areas. It basically provides wide area surveillance to identify something going on, such as a ship moving out there. The technology does not now provide the capability to provide a specific, narrow “what is this vessel?” answer. It may be able to provide that it is a vessel, but you need to send in something else for narrow surveillance or reconnaissance purposes.
Whether these should be drones or existing aircraft or new aircraft—and I'm not considering the F-35 here—is an interesting question from a cost perspective. There's an assumption that drones are cheap. Well, they're not cheap: you will still need infrastructure and alternative training. It requires significant investment.
We have an existing capability, which is an air breathing and pilot-driven capacity for reconnaissance—very limited—relative to investing in the large numbers of drones that you're going to need and the infrastructure that goes along with it. I can't answer on the trade-off, but you still need a supporting system at the end of the day, a north warning system.
One thing that's important to remember about satellites, and I'll conclude here, is that the north warning system was modernized in the 1980s. We are now 30-plus years away from that. Satellites don't last 30 years. You have to replace them in five to ten years, depending on how lucky you are relative to the harsh environment in outer space. The 1967 Outer Space Treaty has no bearing on this issue at all.