Fig 1.
Nonlinear relationship between climate variation and death rate of police violence.
The red, blue, purple, and green curves represent the fitting functions with respect to the high-order polynomial, natural cubic spline with three nodes, natural cubic curves with seven nodes, and B-spline, respectively. The light blue area indicates the confidence interval of the natural cubic spline with three nodes. The histogram above the horizontal axis presents the distribution frequency of the monthly average temperature.
Table 1.
The quantitative temperature effect on the death rate of police violence across different subgroups.
Fig 2.
Temperature effects on the death rate of police violence across different subgroups.
The upper, middle, and lower parts of the figure (divided by red dashed lines) represent the heterogeneity analysis with the introduction of interaction terms, the heterogeneity analysis of the binned models, and the displacement effects, respectively. The estimated effect of temperature is marked by the black solid circles, with grey lines illustrating their 95% confidence intervals.
Table 2.
The quantitative temperature effect on the death rate of police violence over different years.
Fig 3.
The effect of temperature on the death rate of police violence varies over different years.
The black dots represent the effect of temperature for a specific year, with the 95% confidence interval represented by the grey lines. The red horizontal dashed line indicates the average temperature effect over the entire time period, estimated based on all samples.
Fig 4.
The effect of temperature on the death rate of police violence varies over different months.
The blue dots represent the effect of temperature for a specific month, with the 95% confidence interval represented by the blue lines.
Fig 5.
The effect of temperature on the death rate of police violence varies over different states.
The color gradient approaching red indicates a stronger positive effect of temperature, whereas the gradient approaching blue reflects a stronger negative effect, with the black lines outlining states where the estimated effect is more pronounced relative to the rest of the country.
Fig 6.
The effect of temperature on the death rate of police violence varies over different climate regions.
The color gradient approaching red indicates a stronger positive effect of temperature, whereas the gradient approaching blue reflects a stronger negative effect, with the black lines outlining climate regions where the estimated effect is more pronounced relative to the rest of the country.
Fig 7.
Cumulative excess deaths of police violence due to expected temperature changes in the United States by 2050 under the SSP5-8.5 scenario.
The black line represents the median projection, and the yellow colored area shows the distribution of 30,000 Monte Carlo projections derived from resampled parameter estimates and 30 climate models. The box plot illustrates the median, interquartile range, and 95% confidence interval for the estimated cumulative excess deaths projected through the year 2050.