Thanks for the comments.

Roof runoff seems to be a much better idea that trying to transport it from a nearby stream. Although being able to adapt it for either source would seem to be a good system.

I was generally thinking of a system very much like the one Blast posted a diagram of, in the thread linked in his post above. His preflush tube is very much the like idea I had of using a settling tank.

I've read a bit on slow sand filters with bio layers, and my reading indicates they work well. That would be my ideal second step. However, I think I should probably try it out. I also have read about sand as just being a good filter, without maintaining the bio layer. Designing the system to backflush the sandfilter seems more complicated, although I imagine one could adapt the valves off of a pool filter to work. Pool filters might be the thing to adapt for any fitler in such a system.

My idea of adding a carbon/charcoal filter to the system would be to try to account for chemical contamination. I know I'd never be able to account for all possibly chemical contaminants. I've tried to think of a filter that would draw out a lot of chemicals, and carbon seems to do a good job of it, and with some effort, I should be able to create some charcoal, so I could renew that filter. I do not know how effective it might be, but trying to get stuff out is better than not trying at all and slowly poisoning oneself.

Boiling does seem like a good longterm option. Thanks for that suggestion. I could store enough chlorine to do the job for a while. However, I also know about some of the hazards of calcium hypochlorite (that is usually the chemical we call chlorine), and I do not want to store tons of it and risk runaway decomposition.

I've thought about distilling too.

Thanks.