Extended Data Fig. 4: Sensitivity of crop yields to unit changes in different climate variables under SAI estimated using the per-grid-cell MLR method.
From: Solar geoengineering can alleviate climate change pressures on crop yields
T, RD, RI, P and RH stand for temperature, direct radiation, diffuse radiation, precipitation and relative humidity, respectively. T*RH and RH*P indicate the interactions between T and RH and between RH and P, respectively. The regressions are refitted with log change in RD, RI and P so that their coefficients can be easily converted and applied to percentage changes in these variables and because these variables show log linear relationship with yield. a, Global average responses to unit changes applied uniformly to all grid cells (% indicates the native unit (percent) for relative humidity, but relative changes in the radiation terms). b, Global average response to a standard deviation (sd) of a variable for each crop grid cell under the scenario SAI – RCP8.5 during 2020–2099 (that is, sd is the local variability of climate change induced by SAI). Error bars indicate the 2.5th to 97.5th percentile confidence interval of the global average response from Bootstrap resampling and spatial aggregation.
