《Table 1–Regression models of visibility under different RH in DQL, June 2014–May 2015.》

  提示：宽带有限、当前游客访问压缩模式

本系列图表出处文件名：随高清版一同展现

《"Temporal variability of visibility and its parameterizations in Ningbo,China"》

1. 获取 高清版本忘记账户？点击这里登录

1. 下载图表忘记账户？点击这里登录

The obtained regression parameters in Eq.（3）were chosen as initial values of modeling fit.Multiple linear regression was conducted between the residue of prediction and other environmental parameters.Datasets with hourly resolution from June 2014 to May 2015 were used to develop the multiple nonlinear regression equations.An independent variable was added into the regression equation by a stepwise procedure based on importance.It demonstrated that for the first two RH categories，i.e.RH≤80%，RH is the common factor in addition to particle concentration for the variation of visibility，then the regression equations for these two levels were eventually combined together.After several circles of regression and iteration，the final modeling results considering main influencing factors besides PM2.5and RH within three RH ranges are listed in Table 1.It showed that the main contributors to visibility under different RH are different，and the influence of all variables on visibility was additive.Specifically，the independent variables in the model are PM2.5，and RH when RH≤80%，while O3is the major contributor to visibility（aside from PM2.5and RH）within RH of80–90%.The importance of O3in the model requires further investigation.Results presented in Table 1 also suggested that temperature can affect visibility when RH>90%.Likely，temperature affects visibility by influencing condensation of water vapor in the atmosphere.