Talking about river formation tonight in chat brought up some questions. One person would like to have the river form from a glacier, and I think that's a good starting point, but as I pointed out, it could be the only thing that forms the glacier -- which he understood. Unfortunately, the pace of chat isn't a really good place for long discussions like that, I think.
I've been thinking about why a glacier couldn't be the only source, though. It would seem like a natural one, wouldn't it? I think it finally all came together.
Yes, I really do think about these things. I've read a few books on geology, mountain formation, and climate. (Odd things to know -- river formed valleys have a V shape and glacial formed valleys have a U shape.)
In order for a glacier to be the primary (as in largest, not first) source for a river, it would have to be very large. For it to be that large, I would think it would have an impact on the climate around it... that is everything would be very cold where it is. If it is huge and cold, it would slow the amount of ice melt from the glacier, and that would mean it couldn't have enough to form a river at the source. In fact, if it were that large, it might actually grow rather than retreat. Actually, many mountain-based glaciers do grow each year as new snowfall puts pressure on the upper areas, getting it to flow downward. Ice under that kind of pressure does flow, like water, but very, very slowly.
The same things would be true if you spilt the glacier up into a number of smaller glaciers, though you could spread their area of influence out farther. The only problem here is that all the glaciers would have to be 'pointed' in the same direction and have no impediments to their melt off that would stop them from joining into one large river. But... it could work. If they were all part of what once had been a large glacier, now melting back with ground areas showing between, they would likely all have a common spot where they join. I've stood at the top of a pass in the Rocky Mountains and watched the little brooks of snow melt fall off on both sides of me, one head west toward the Colorado River basin, and the other to the east, to the Missouri-Mississippi River basin, so glaciers couldn't just be added on mountainsides and all expected to meet up.
However, the more I consider this, a sudden change in climate could attack a huge glacier and start the melt. After all, it obviously has in the past. But would this sudden change last long enough for a city to be built at the mouth of the river, dependent already on that melt? Sudden, in ecology terms, doesn't mean something that happens in days, after all. Or would the melt be enough, even after the glacier shrunk or broke into several smaller ones?
No matter what, such an area -- either large glacier, or many small ones -- would also have a lot of moisture from evaporation on top of the glacier, not just melt at the base. So rain and snow -- and snow melt -- would play a part in the river's formation. So would all the climates it passes through from the glaciers to the ocean. Everything from the ecology (how much of the rain water is retained, for instance, in the foothills?), to the human made irrigation systems in the desert will affect the river.
All in all, I think it's a really fascinating subject. I've studied moraines and glacier lakes -- even stood by one. Lovely, pure clear water. I have pictures somewhere... I'll have to find them. (Print, before the digital cameras.)
I think I'm going to use this. I know about the caves under some glacier bases, about how something trapped in a glacier at the summit can take hundreds (or was it more?) of years to reach the base. It's all interesting stuff...