Figure 1

Mean impact of temperature on log cell per capita GDP change (\(\Delta \log y_{ir}\)) in China. The plot depicts the relationship between annual average temperature and changes in per capita GDP with 95% confidence intervals (\(N=1078*3\)) for quinquennial variations between 1985 and 2005. Empirical models include cell-fixed effects, year-fixed effects, and precipitation as controls (see Supplementary Methods). Lagged effects are not included. (a)-(c) show the mean impacts of temperature on economic growth, (a) modeling with random forests, (b) modeling with cubic regression (c) modeling with linear spline regression. (d) the kernel density of observations and the models’ fitted values