The results showed that the temporal variation of standard ice thicknesses in EW and NS directions were very similar. Standard ice thickness in EW direction was obviously larger than that in NS direction. The annual mean standard ice thicknesses in EW direction and NS direction were 43.44 mm and 38.83 mm. The maximum standard ice thickness happened in 2008 with the maxima of 154.03 mm and 116.74 mm in EW direction and NS standard respectively. The yearly icing days varied from 33 days in 1999 to 84 days in 2012. Morlet wavelet analysis showed there were quasi 3 years, quasi 6 to 8 years, quasi 11 to 12 years and quasi 25 years periods in standard ice thickness. Based on the Pearson type III and Gumbel extreme value distributions, standard ice thicknesses in different recurrence period were calculated. Using Pearson type III extreme value distributions, the calculated standard ice thicknesses in EW direction and NS were 69.57 mm and 60.86 mm for the 15-year recurrence period, respectively. For 50-year recurrence period, calculated standard ice thicknesses increased to 109.07 mm (EW) and 87.43 mm (NS). The calculated standard ice thicknesses in EW direction were larger than those in NS direction for different recurrence period and the similar result was also found using Gumbel extreme value distribution. The north wind in winter, which made more supercooled rain and cloud droplet collide with EW direction metallic line leaded to thicker ice in EW direction. Meteorological characteristics were compared between the icing days and the non-icing days from 1 December 2012 to 4 March 2013. The mean air temperature of the icing days was 7.8 ℃ lower than that of the non-icing days. The mean relative humidity of the icing days was larger than that of the non-icing days. Both the mean visibility and wind speed of the icing days were smaller than those of the non-icing days. The dominant wind direction was north when icing happened and south wind was usually found in the non-icing days.