Temporal variability of visibility and its parameterizations in Ningbo, China

Abstract Simultaneous and continuous measurements of visibility, meteorological parameters and air pollutants were carried out at a suburban site in Ningbo from June 1, 2013 to May 31, 2015. The characteristics of visibility and their relationships with air pollutants and meteorological factors were investigated using multiple statistical methods. Daily visibility ranged from 0.6 to 34.1 km, with a mean value of 11.8 km. During the 2-year experiment, 43.4% of daily visibility was found to be less than 10.0 km and only 9.2% was greater than 20.0 km. Visibility was lower in winter with a frequency of 53.4% in the range of 0.0–5.0 km. Annual visibility had an obvious diurnal variation, with the lowest and highest visibility being 7.5 km at approximately 06:00 local time and 15.6 km at approximately 14:00 local time, respectively. Multiple correspondence analysis (MCA) indicated that the different ranges of visibility were significantly affected by different levels of pollutants and meteorological conditions. Based on the analyses, visibility was found to be an exponential function of PM 2.5 concentrations within a certain range of relative humidity. Thus, non-linear models combining multiple linear regressions with exponential regression were subsequently developed using the data collected from June 2014 to May 2015, and the data from June 2013 to May 2014 was used to evaluate the performance of the model. It was demonstrated that the derived models can quantitatively describe the relationships between visibility, air quality and meteorological parameters in Ningbo.