Variations Trend Evaluation of Rainfall Using Mann-Kendall and Linear Regression in Khuzestan Province

Subject Areas : watere sciences

mostafa  ghasabfeiz ¹ (water engineering department, water science faculty, Islamic azad university, shoushtar branch, shoushtar)
hossein  eslami ² (water engineering department, water science faculty, Islamic azad university, shoushtar branch, shoushtar)

Received: 2017-05-29 Accepted : 2017-10-17 Published : 2017-11-22

Keywords: Regression, Khuzestan province, Mann-Kendall, Rainfall trend,

Abstract :

Rainfall trend could be evidence of climate change. There are different ways to review trends in rainfall time series are divided into two categories: statistical parametric and non-parametric. Linear regression method is from statistical and most common non-parametric method is Mann-Kendall. Considering the period of 33 years from 1980-81 until 2013-14, at 47 meteorological stations in Khuzestan province. Results showed that the slope of the regression line can vary from -13 to 0.8. More weather stations are a negative trend in the amount of rainfall and only Abdolkhan and Polezal stations have a significant positive trend. Mann-Kendall test for the existence of trends in rainfall at a confidence level of 5% was found that the negative trends are more stations and rainfall in the province is declining. Only stations of Polezal, Chamnezam, Harmaleh, Abdolkhan and Arabhasan have positive trend and Batvand shoor has Z value equal to zero with rainfall amounts are fixed in the period. Although all stations have a positive or negative trend, but only three stations Sousan, Sepid dasht sezar and Mollasani are significant at a confidence level of 5%. Although it cannot be concluded that climate change has occurred but due to the fact that most stations are reduced rainfall trend is alarming.