Meteorological factors correlate with transmission of 2019-nCoV: Proof of incidence of novel coronavirus pneumonia in Hubei Province, China

Jianfeng Li; Linyuan Zhang; Zhihua Ren; Caihong Xing; Peihuan Qiao; Bing Chang

A variety of statistical methods were used for mathematical modeling to further reveal the potential laws. Kim et al. believed that the incidence and mortality of diseases followed the Poisson distribution and used Poisson function as the link function to establish a generalized linear model (GLM) (Kim et al., 2016) . Lim et al. used the cross design of Poisson GLM and temperature-matched case. Through case cross design, personal characteristics and long-term and seasonal trends can be controlled (Lim et al., 2012) . GLM was also used in the studies of Onozuka et al. and Souza et al. (Souza et al., 2012; Onozuka et al., 2009) . For seasonal viral infectious diseases, Gurgel et al. and Yusuf et al. used time series analysis to reduce model deviations (Gurgel et al., 2016; Yusury et al., 2007) . Chan et al., du Prel et al., Lin et al., Varela et al. use the autoregressive integral moving average (ARIMA) model to control the autocorrelation of time series data (Chan et al ., 2002; du Prel et al., 2009; Lin et al., 2009; Varela et al., 2004) . Carreras et al. used a generalized additive model (GAM) using a quasi-Poisson distribution as link function.