Study of the Effect of NAO on Precipitation in Iran Using Network Analysis Aproach from November to April of the 1979-2016

Displaying and quantifying how large-scale climatic signals affect the regional and global climate is one of the issues of interest to researchers. New approch of Network Analysis enables us to study a complex system such as the climate more effectively. In this study, the effect of NAO on precipitation in Iran was investigated using the method of Network Analysis. The data used are daily precipitation and NAO daily indices obtained from ERA-Interim data base. The time period covers the cold months (November, December, Jenury, Februry, March, April) of the years 1979-2016. In order to smooth the daily precipitation data, a weekly moving average was applied to data. A modified Pearson’s cross correlation coefficient  was employed for correction of the bias of traditional Pearson correlation due to many zero-precipitation data. The years of the strong NAO negative and positive phases, NAO- and NAO+, were determined by the frequency of the weekly moving avereage NAO index (WMANAOI) greater that 2 standard deviations ( ), where Sd is the standard deviation. The precipitation network was established based on a constant link density of 0.05 for which vertex degree distribution of network is considered. The later statistics is the covariance between precipitation time series at grid points that show indirectly the effect of the variation of precipitation due to large-scale climate signal. Results demonstrated that Network Analysis method can display and quantified the effect of NAO-and NAO+ on the precipitation of Iran. Degree distribution over the whole duration shows that the most degree distribution is in the northwest, west, southwest and east of Iran. The deserts of Iran that are located in the center of the country have the least degree distribution, and so the least similarity with each other, that are clustered in a group. Also, threshold of network, i.e. , is 0.42 and the histogram of all  has the right skewness. The  is a threshold for selecting the statistically significant  between each two griod points over Iran. The degree distribution for noraml years, with less NAO activities, displays that is 0.42 and degree distribution totally is less than those of NAO+ and NAO-phases. During NAO-, degree distribution is increased in the northwest and southwest of Iran. The  in NAO- is 0.46. During NAO+, the degree distribution is increased in the south of Iran while the threshold of network remain 0.43. Also, the histogram of all  has a positive skewness at NAO+ and NAO-that indicate the strong correlations has positive values. Minimum values of  for the whole period, normal years, NAO+ and NAO-are -0.11, -0.14, -0.3 and -.033, respectively.