Why cirrus cloud seeding cannot substantially cool the planet

The net warming effect of cirrus clouds has driven part of the geoengineering research toward the idea of decreasing their occurrence frequency by seeding them with efficient ice nucleating particles. We study responses of cirrus clouds to simplified global seeding strategies in terms of their radiative fluxes with the help of the ECHAM-HAM general circulation model. Our cirrus scheme takes into account the competition between homogeneous and heterogeneous freezing, preexisting ice crystals, and the full spectrum of updraft velocities. While we find that the cirrus cloud radiative effect evaluated from our model is positive and large enough (5.7 W/m(2)) to confirm their geoengineering potential, none of the seeding strategies achieves a significant cooling due to complex microphysical mechanisms limiting their climatic responses. After globally uniform seeding is applied, we observe an increase in cirrus cloud cover, a decrease in ice crystal number concentration, and a decrease in ice crystal radius. An analysis of their respective radiative contributions points to the ice crystal radius decrease as the main factor limiting seeding effectiveness.