A lot.
First, setting loose a process to remove CO2 could get away from us. If CO2 drops below 300ppm, agriculture could not support more than about half of us. Get it below 200ppm and everything dies. We have a much wider margin for error for higher CO2 than lower.
But just as bad, screwing around with albedo has even narrower margins, and we’re overdue for another glaciation anyway.
It is foolish to turn on any large scale tweaking of albedo that we can’t quickly turn off, because that is a positive feedback loop that has existed in the past, and it ends with 2 miles of ice over Detroit.
Tidewater News and Opinion Forum 
RE: “A lot [could go wrong].”
I agree. In fact, I’d say the risk that geoengineering involves is technically unmanageable.
The main reason is that the risk associated with climate change is uncertain. That is, we don’t know with certainty what the outcome of doing nothing would be. In the calculus of risk management, this means that we cannot determine the probability or even the cost of occurrence (of the risk event) and therefore have no way to measure the benefits of any proposed mitigation.
WSJ’s writer acknowledges this when he says, “The case for geoengineering begins with the argument that the alternative is insufficient,” but this misses the point. The case for geoengineering ends with the observation that climate change is not reliably predictable.
LikeLike
Even if you accept the CMIP-5&6 models, which are known to be overheated, the difference between business as usual and pursuing Net Zero is only about 1/3°C in 100 years.
So, what possible sense can there be in potentially triggering an early glaciation over an unmeasurable difference in temperatures?
LikeLike