Of course the increased greenhouse gases would help here as well; the ocean’s surface ice would melt on its own, after all, as soon as the air stayed regularly above 273 K. But as the conference proceeded, more and more problems with the greenhouse-gas plan were being pointed out. It entailed another huge industrial effort, almost the equal of the metanat monster projects, like the nitrogen shipments from Titan, or the soletta itself. And it was not a onetime thing; the gases were constantly destroyed by UV radiation in the upper atmosphere, so they had to overproduce to reach the desired levels, and then continue producing for as long as they wanted the gases up there. Thus mining the raw materials, and constructing the factories to turn those materials into the desired gases, were enormous projects, and necessarily a largely robotic effort, with self-guided and replicating miners, self-building and regulating factories, upper-atmosphere sampler drones— an entire machine enterprise.

The technical challenge of this was not the issue; as Nadia pointed out to her friends at the conference, Martian technology had been highly robotic from the very beginning. In this case, thousands of small robotic cars would wander Mars on their own, looking for good deposits of carbon, sulfur, or fluorite, migrating from source to source like the old Arab mining caravans on the Great Escarpment; then when new feedstocks were found in high concentrations, the robots could settle down and construct little processing plants out of clay, iron, magnesium, and trace metals, providing the parts that could not be constructed on-site, and then assembling the whole. Fleets of automated diggers and carts would be manufactured to haul the processed material in to centralized factories, where the material would be gassified and released from tall mobile stacks. It wasn’t that different from the earlier mining for atmospheric gases; just a larger effort.

But the most obvious deposits had already been mined, as people were now pointing out. And surface mining couldn’t be done the way it used to be; there were plants growing almost everywhere now, and in many places a kind of desert pavement was developing on the surface, as a result of hydration, bacterial action, and chemical reactions in the clays. This crust helped greatly to cut down on dust storms, which were still a constant problem; so ripping it up to get to underlying deposits of feedstock materials was no longer acceptable, either ecologically or politically. Red members of the legislature were calling for a ban on just this kind of robotic surface mining, and for good reasons, even in terraforming terms.