There were, however, a few exceptions to this situation. The Noachian Age had been followed by a brief “warm wet period” in the late Hesperian, a period perhaps as short as a hundred million years, when a thick warm CO2 atmosphere had allowed liquid water to run on the surface, carving some river channels down the gentle tilts of the plateau, between crater aprons diverting them this way and that. And these watercourses had of course remained after the atmosphere had frozen out, empty arroyos gradually widened by the wind. These fossil riverbeds, like Nirgal Vallis, Warrego Valles, Protva Valles, Patana Valles, or Oltis Vallis, were narrow sinuous canyons, true riverine canyons rather than grabens or fossae. Some of them even had immature tributary systems. So efforts to design a macro-watershed system for the south naturally used these canyons as primary watercourses, with water pumped to the head of every tributary. Then there were also a number of old lava channels that could easily become rivers, as the lava, like the water, had tended to follow the path of least resistance downhill. And there were a number of tilted graben fractures and fissures, as at the foot of the Eridania Scopulus, that could likewise be turned to use.

In the conference, big globes of Mars were marked up daily to display different water regimes. There were also rooms full of 3-D topo maps, with groups standing around different watershed systems, arguing their advantages and disadvantages, or simply contemplating them, or fiddling with the controls to change them, restlessly, from one pattern to another. Nadia wandered the rooms looking at these hydrographies, learning much about the southern hemisphere that she had never known. There was a six-kilometer-high mountain near Richardson Crater, in the far south. The south polar cap itself was quite high. Dorsa Brevia, on the other hand, crossed a depression that looked like a ray cut out from the Hellas impact, a valley so deep that it ought to become a lake, an idea that the Dorsa Brevians naturally did not like. And certainly the area could be drained if they cared to do it. There were scores of variant plans, and every single system was strange looking to Nadia. Never had she seen so clearly how different a gravity-driven fractal was from impact randomness. In the inchoate meteoric landscape, almost anything was possible, because nothing was obvious— nothing except for the fact that in any possible system, some canals and tunnels would have to be built. Her new finger itched with the desire to get out there and run a bulldozer or a tunnel borer.