At the peak of the risk spectrum, Kim Jong Un could decide one day — surely not now, as alert levels south of the 38th Parallel are too high — to mount a full-scale invasion with the North’s million-man army. If he were to “roll the iron dice” (the phrase that German Chancellor Theobald von Bethmann-Hollweg used to describe going to war in 1914), the North’s nuclear capacity could come into play in two ways. The first would be limited to trying to deter a counter-blow by American-led forces, something akin to the amphibious “left hook” that Douglas MacArthur mounted at Inchon in September 1950. Indeed, it would take a very steely resolve in Washington to send large numbers of troops to fight in what might become an irradiated battle zone — where millions of Korean civilians would be among the most vulnerable.

The second, and far riskier, option Kim could choose would be to detonate a nuclear weapon at the outset of an invasion of the South. Not on the ground, but at very high altitude. Nobody would be killed, but the explosion would generate a highly disruptive electromagnetic pulse. This would fry or block almost all South Korean and American communications, and many other computer systems, greatly diminishing defensive capacities for days, probably weeks. It would greatly complicate efforts to deal with a blitz from the North, an assault perhaps to be led by Kim’s roughly 100,000 special forces, many of them infiltrating via tunnels or on radar-evading Colt II biplanes. The chance Pyongyang would be taking — that President Obama would refrain from nuking the North in retaliation — is enormous. But, given that a high-altitude detonation would not have killed anyone on the ground, the American response might well be limited to conventional counterattack. If so, the great early advantage enjoyed by invaders less dependent on the kinds of communications that would be disrupted — and the latent threat of holding South Koreans hostage to nuclear attack — just might give the North a chance to prevail.