Spread of SARS-CoV-2 Coronavirus likely to be constrained by climate

Miguel B. Araujo; Babak Naimi

Building on the concept of ecological niche, we develop projections of monthly changes in the climate suitability for SARS-CoV-2 outbreaks. Projections are obtained from an ensemble of 10 familiar machine learning ecological niche models (11) , each with 20 replications generated by repeating four times a 5-fold cross validation that accounts for and enables the quantification of variability across initial conditions (12) . Models were trained using the distribution of all recorded local transmissions of SARS-CoV-2 Coronavirus (excluding imported cases) with data compiled from John Hopkins University Mapping 2019-nCoV portal (13) . Predictors were monthly mean temperature, interaction term between monthly minimum temperature and maximum temperature, monthly precipitation sum, downward surface shortwave radiation, and actual evapotranspiration. Climate data refer to a period including January, February and March 2009-2018, with data downloaded from the updated high-resolution global climatology database -TerraClimate (14) . Models were then projected monthly for the rest of the year.