Caldeira Lab

Simulated long-term climate response to idealized solar geoengineering

Long Cao, Lei Duan, Govindasamy Bala, & Ken Caldeira

Cao, L., L. Duan, G. Bala, and K. Caldeira ,2016: Simulated long-term climate response to idealized solar geoengineering. Geophys. Res. Lett., 43, doi:10.1002/2016GL068079.


Solar geoengineering has been proposed as a potential means to counteract anthropogenic climate change, yet it is unknown how such climate intervention might affect the Earth's climate on the millennial time scale. Here we use the HadCM3L model to conduct a 1000 year sunshade geoengineering simulation in which solar irradiance is uniformly reduced by 4% to approximately offset global mean warming from an abrupt quadrupling of atmospheric CO2. During the 1000 year period, modeled global climate, including temperature, hydrological cycle, and ocean circulation of the high-CO2 simulation departs substantially from that of the control preindustrial simulation, whereas the climate of the geoengineering simulation remains much closer to that of the preindustrial state with little drift. The results of our study do not support the hypothesis that nonlinearities in the climate system would cause substantial drift in the climate system if solar geoengineering was to be deployed on the timescale of a millennium.


Figure 1. Model-simulated 1000 year time series for climate variables of (a–c) TOA net flux, (d–f) surface air temperature, (g–i) precipitation, (j–l) precipitation minus evaporation (P − E), (m) sea ice area, (n) maximum strength of Atlantic meridional circulation (AMOC), and (o) net primary production (NPP) in the simulation of CTR, 4 × CO2, and SRM. A 20 year running averaging is applied to all variables.


Figure 2. (a–c) Model-simulated zonal mean distribution of changes in surface air temperature, precipitation, and precipitation minus evaporation (P − E) in the 4 × CO2 and SRM simulations for the period of years 60–100 and years 960–1000, respectively. Equal area is used in the Xaxis. (d–f) Temporal evolution of RMS difference in surface air temperature, precipitation, and precipitation minus evaporation (P − E) for the 4 × CO2 and SRM simulations. RMS differences are calculated relative to the CTR simulation using equation (1) over the globe.