NON-PARAMETRIC TREND ANALYSIS ON KONYA CITY PRECIPITATION

Abstract: The decreasing or increasing tendency of a random variable depending on time is called “trend”. The investigation of the trend for the hydrological and meteorological data, i.e. the determination of decreasing or increasing tendency of water amount, has great significance in terms of planning and managing of water resources. Parametric and non-parametric tests are used to determine the trends. Beside the parametric tests give best results when the data is in accordance with the normal distribution, non-parametric tests are independent from the distribution of the population’s random variable and parameters. Since the distributions are not normal in general, the results obtained by the non-parametric tests are better than those obtained by the parametric ones. Therefore, the parametric trend analysis methods of Mann-Kendall Test, Modified Mann-Kendall Test, Spearman Rho Test, Sen’s t-test and Sen’s Trend Test were used to perform the trend analysis of the total precipitation data of the precipitation gauging station (No.17244) of Konya Region between years 1929-2006. The data set obtained by the gauging station is considered as monthly total precipitation, annual total precipitation and all the data set, then they were investigated considering the significance level of 0.05 whether they have any meaningful trend or not. Additionally, the slope directions obtained by the trend analyses were checked by the aid of the Sen’s Trend Slope method used to determine the slopes of the trends. As a result of the trend analyses, it is concluded that the trend analysis methods used in this study presented test results in good agreement with each other, and their slope directions were obtained as the same as the slope directions of Sen’s Trend Slope Method.