There's a proof-of-concept test being started in which a big balloon will be used to tether the high end of a hosepipe, and water will be pumped up to be dispersed in a spray. It's a fair old engineering challenge, and the water is just a placeholder for speculative future compounds to cool the atmosphere.  I have a problem with the whole notion of geoengineering: we're already doing it.  We're very successfully engineering the atmosphere to be hotter and wetter, right now. 

There's an analogy with driving towards a precipice: our carbon emmissions are the accelerator; the cliff is, say, 4 degrees of warming (=catastrophe).  What do you do when you are driving towards a bad thing?  You let go the accelerator.  You do not really apply the brake as well... and you sure as hell don't apply a pedal labeled "probably a brake: untested".  There's a reason that the brake and accelerator are worked with the same foot.

I have yet to be swayed by geoengineering.  It strikes me that adding more inputs to a chaotic system in a transitional state is just asking for trouble.  I worry that the gee-whizz relief of being able to "do something" will make it attractive to the people who make such decisions.