Climate affects global patterns of COVID-19 early outbreak dynamics

Gentile Francesco Ficetola; Diego Rubolini

is the (which was not peer-reviewed) The copyright holder for this preprint . https://doi.org/10.1101/2020.03.23.20040501 doi: medRxiv preprint Figure 1 . Variation of Covid-19 growth rates in relation to climate, and spatial predictions for different months. Variation of confirmed Covid-19 cases growth rates for the first 5 days after reaching a minimum threshold of 50 cases (r 50 ) during the January-March 2020 pandemic outbreak in relation to A) mean temperature and B) mean absolute humidity of the outbreak month. Partial regression plots (37) of the best-fitting linear mixed models (LMMs) of r 50 in 5 relation to temperature and humidity, respectively (Tables S3-S5 ). The quadratic terms of both temperature and humidity were highly significant (temperature: F 1,87 = 14.4, P < 0.001; humidity: F 1,84 = 7.82, P = 0.006; full details in Tables S3 and S5). Shaded areas are 95% confidence band. C): global pattern of r 50 ; the size of dots is proportional to the observed r 50 values. The background shows the spatial prediction of growth rates according to mean March 10 temperatures (22) . Predictions are based on the best-fitting LMM of r 50 in relation to mean temperature of the outbreak month (Table S3 ). D-E: Spatial prediction of growth rates according to mean June and September temperatures (22) , highlighting that conditions for rapid disease spread appear in temperate regions of the Southern Hemisphere.