True, although in this case “large” means in excess of 100,000 tonnes each. I believe one suggestion was oil tanker sized ships built with cobalt hulls and housing the bomb in their holds.
What if we “duck & cover” ?
Exactly. Three Mile Island did in fact have a complete core meltdown. However, because it also had a well-designed containment structure, it didn’t release any significant amount of radioactive contamination.
Chernobyl is just about the worst case scenario imaginable. The core melted, set the graphite moderator on fire, flashed all of the coolant to steam, and blew the top off of the whole building. As Chronos mentioned, the design was positively idiotic. It had what is referred to as a “positive temperature coefficient of reactivity,” which meant that it was inherently unstable. As the temperature of the coolant increased, the reactor’s power also increased. U.S. reactors are designed with a negative temperature coefficient of reactivity, and are designed to use the coolant as the moderator necessary to keep the reaction going. If the coolant is lost, the reaction automatically stops. In such an instance, the residual “decay heat” can cause a meltdown, but the actual nuclear reaction will not still be occurring.
Even in the case of the area surrounding Chernobyl, which has been evacuated for the last 24 years, reports are that wildlife is doing great, including large animals such as deer. Apparently the minimal amount of residual radioactivity still present is less of an impediment for the wildlife than the presence of people.
It is difficult to imagine any situation that could result in the “whole east coast of the U.S. [being] a radioactive wasteland for thousands of years” from a nuclear accident. Radioactive materials are either short-lived (giving off large amounts of radioactivity) or long-lived (giving off smaller amounts of radioactivity). So after a relatively short amount of time following a hypothetical nuclear accident (or a weapon, for that matter), the worst of any radioactive contamination has decayed away.
Cobalt-59 is a stable metal. When irradiated with neutrons (in a nuclear reactor or a weapon), you can convert to metal to Cobalt-60. Cobalt-60 is one of the worst possible radioactive materials there is, because it has a half-life of 5.27 years. Unlike something with a half-life of minutes or seconds, it will persist in the environment for years. Unlike something with a half-life measured in thousands or millions of years, it’s relatively short half-life indicates that it is fairly radioactive and therefore quite dangerous.
Nevertheless, after five half-lives, you would have just 3.125% the radioactivity as you did when you started. This is a non-insignificant 26 years, but it’s not forever, nor is it “thousands of years.” So a nuclear weapon salted with cobalt could indeed make a region uninhabitable for years or decades, but not much longer than that.
Thanks everyone for your informative replies. I find the topic fascinating. A couple of supplementary questions if that is OK. I’ve heard before that the concept of “Nuclear Winter” was an exaggeration at best, is this true? More to the point, what short/medium/long term effects would such a conflict have on climate, the weather?
No, I think Nuclear Winter is very much on the table. Here (warning pdf)is a fairly recent study from Rutgers University. Their conclusions? In a full scale Russia/USA exchange, expected to produce up to 150Tg of smoke, we might experience immediate cooling of No. America by about 20° C and Eurasia of about 30° C. World wide temperatures could drop by 8° C. And a decade later temperatures would still be depressed by at least half these amounts. In smaller exchanges the effect would be significantly reduced. But even an exchange producing 5 Tg of smoke (like say a India/Pakistan exchange) would produce a effect similar to the Little Ice Age.
Depending on the size of the exchange imagine the last hard frost in Iowa is around late June and the first hard frost is in mid August. In fact growing seasons would be cut by 60-120 days. No more grain production in Iowa. And no more citrus in California and Florida. Short term they also predict large disruptions in the water cycle, which would among many things would send No. America into a severe, decade long, continent wide drought. So nothing good if you like to eat or drink.
Animals do benefit from shorter lifespans. If you only live 10 years in the wild, getting cancer after 20 years is not such a big issue.
And the whole term “uninhabitable” is really more a measure of modern human standards (or squeamishness) than the actual amount of radiation. If you don’t mind dying of cancer in your 40s, even the area immediately around Chernobyl is habitable for humans. Given that human lifespans were below 40 in the middle ages, calling Chernobyl uninhabitable is like saying the 1300s were uninhabitable. Not that I want to downplay the issues of radiation, but irrational fear has really driven public opinion on the issue.
<insert nitpick about average lifespan>
Even if an area were habitable in the sense that you could live in a house or work in a factory there, the land might be unfarmable because of radioisotopes getting into the food chain.
From what I have heard, it is not so much that nuclear winter isn’t real, it is that its effects would get lost in the noise. If your agricultural production is 1% of pre-war levels, nuclear winter is like burning someone already on fire.
There is if the US is ahead in the mineshaft gap 
High infant and child mortality doesn’t mean everyone lives to be 40. It means that 40% of everyone dies by age 12 and the rest can live up to nice and old. Growing up in a radiation zone means everyone gets tumors and dies young-ish. Two very different scenarios. A guaranteed early death by cancer? No thanks. I’d rather take my chances in the 14th century.