The main thing that brings weather balloons and airships to earth is loss of lifting gas (helium, hydrogen, hot air, etc.) What if the lifting gas could be replenished?
The idea is to collect a small amount of rainwater in the gondola, and use electrolysis to generate H2 via solar panels attached to the balloon. You'd "top off" the balloon with this generated gas. An active buoyancy control could make sure the balloon maintains the desired altitude.
One of the issues to solve is that as the lifting gas seeps out, oxygen and nitrogen will seep in through the membrane. Maybe a molecular sieve could be also used to filter out the large molecules.
Also, electrolysis doesn't work too well on rainwater due to lack of electrolyte (salt), although I can't find hard figures on this. Estimating about 1 g of H2 per amp per day via electrolysis, you wouldn't need so big a PV array. You could use a voltage regulator to keep it at the ~1.25 V required.
An alternative design would be to use the molecular sieve to fill the balloon with nitrogen, instead of hydrogen. This eliminates the inefficient electrolysis step, but you'd need a VERY big balloon.
Bonus Points: Attach the balloon to a glider, and when the balloon reaches maximum altitude it purges the hydrogen. It then glides in a direction of its choosing to a body of seawater. From there it harvests salty water, which it uses to re-inflate its balloon.-- RobotVoodooPower, Nov 02 2005 Ethermal Resting Place Ethermal_20Resting_20Place [theircompetitor, Nov 03 2005] SOLONG project http://www.acpropulsion.com/solong/Partially baked (solar UAV) [RobotVoodooPower, Jun 04 2008] H2 leaks faster than anything else (it's such a small molecule). Also, I thought these things rose above the clouds, so you won't have any water to play with.
Why not simply make it a hot air balloon, and heat the air?-- DrCurry, Nov 02 2005 Ah, so it's not redundant to Ethremal Resting Place-- theircompetitor, Nov 03 2005 Rainwater is very slightly salty due to sea spray (here in Australia there are salt deposits under the surface that are thought to have been produced by millions of years of low runoff, thus a gradual net gain of salt), so if you had a supply of salt to start with, I don't imagine you would run out as it's not used up much during electrolysis (unless you us a high voltage, then you start making chlorine gas, hypochlorite, hydroxide etc depending on precise conditions - you could use the denser than air chlorine outside the balloon (see air-cooling balloon) but I don't think that would work.) N2 etc. would diffuse in very much slower than H2 diffuses out, but in the spirit of Eterna you could flush out some of the contaminated gas with fresh pure H2.-- spidermother, Jan 26 2006 random, halfbakery