Analysis of temporal trend of groundwater quality using nonparametric Mann-Kendall and Sen’s methods (Case study: Yazd-Ardakan Plain)

As one of the vital components of life, water is essential for many economic activities. The assessment and monitoring of groundwater quality has always been a major challenge associated with special problems. Taking into account the decline in the level of groundwater in most of Iran’s plains including Yazd-Ardakan plain, and its effects on groundwater quality, the present study uses a non-parametric Mann-Kendall test to investigate the trend of changes in the vriables of groundwater quality in Yazd-Ardakan plain and to estimate the slope of trend line by Sen’s slope estimator. 

Yazd-Ardakan Plain is one of the widest plains in Iran. In this research, the statistics and information about 25 sources including 12 wells, 12 qanats and 1 spring during a 10-year statistical period (2004-2014) has been used. To study the trend in time series, many parametric and nonparametric statistical methods have been developed and widely used by various researchers. Mann-Kendall method was initially developed by Mann (1945), and then by Kendall (1970). The null hypothesis in Mann-Kendall test implies randomness and lack of trend in data series, and confirmation of hypothesis one (rejection of the null hypothesis) indicates the presence of a trend in the data series. A very useful index in Mann Kendall test, Sen’s slope estimator shows the linearity of trend. 

Groundwater is considered as one of the important biological parameters; therefore, the monitoring and evaluation of groundwater quality change is a significant issue. Overall, of all the studied parameters, the parameters of total dissolved salts, sodium absorption ratio and electrical conductivity had very large positive slopes compared with other parameters. This represents a significant increase in the concentration of sch parameters during the period of investigation. 

In this study, the trend of changes in some groundwater quality parameters in Yazd-Ardakan plain in Yazd province was studied by using Mann-Kendall’s nonparametric test and the magnitude of this trend was measured by Sen’s slope estimator. The results show that the quality of groundwater in this plain has declined over the studied period. The biggest slope changes in the trend line (micromhos on cm  EC= 550)  was related to the electrical conductivity indicating groundwater salinity, Moreover, the noticeable changes in electrical conductivity in recharge and discharge months indicate drastic changes in the groundwater quality due to precipitation. It can be argued that rainfall (total monthly rainfall of 10.8 mm in the month of discharge and total rainfall of 76 mm in the month of recharge) has greatly influenced the  groundwater quality .