**5.1 Temporal change in CCl4 concentration in the aquifer**

The changes in CCl4 concentration in typical wells are presented in Table 4 and Fig.15. However, there is a general downward trend in concentration, CCl4 concentration in most of the wells increased in 2010. This may be due to the decrease in groundwater exploitation.


"n" is the number of the samples

Table 4. Summarizes of changes in the average CCl4 concentrations with time (μg/L)

The Fig. 15 shows that: (1) the CCl4 concentration in karst auifer has obvious seasonal variation. In general, CCl4 concentration of groundwater during the drought period from February to June is relative lower and during the rainy period from August to October is much higher. (2) CCl4 changes rapidly with time, which is notably different from the

Transport of Carbon Tetrachloride in a Karst Aquifer in a Northern City, China 447

common porous media aquifer. This may be due to the facts that: (1) The groundwater velocity in the aquifer is much higher than that in the porous media, which can reach 131.9m/h-3027.0 m/h. As a result, the advective flow dominates the movement of pollutant; (2) Local groundwater flow regime changed frequently. It was one of the important water supply source with over 80 wells owned by different departments for different purposes. Pump stopping and starting at different wells caused change in local flow field and subsequently cause change in CCl4 concentration; and (3) CCl4 transport channel is complex.

The non-parametric Mann-Kendall test was used to detect monotonic (increasing or decreasing) trends in time-series of CCl4 concentrations for the eight typical wells during the period 2004-2010. The Mann-Kendall test is widely used in environmental science for the detection of trends in time-series data. A 5% significance level was used to indicate statistically significant trends in the current study. The Mann-Kendall trend statistics (Z) indicates significant decreasing (Z <–1.96, p <0.05) and increasing (Z >1.96, p <0.05) trends. Table 5 presents that there are a highly significant decreasing trend in CCl4 in the karst aquifer and the decreasing trend in pollution source sub-area and the north sub-area area are more significant.

Well Time Z Trend

**5.3 Temporal change in CCl4 plume distribution** 

was higher than that in the wells outside of the center.

than in length.

X-49 2004.02-2010.09 -4.35532 Decreasing, Significant X-62 2004.02-2010.09 -5.72119 Decreasing, Significant X-47 2004.02-2010.09 -4.74661 Decreasing, Significant X-83 2005.02-2010.09 -2.35838 Decreasing, Significant X-43 2004.02-2008.12 -3.38838 Decreasing, Significant X-59 2004.02-2010.09 -2.80019 Decreasing, Significant X-74 2004.02-2010.08 -4.67132 Decreasing, Significant X-56 2004.02-2010.09 -5.11015 Decreasing, Significant Table 5. The Mann-Kendall trend statistic (Z, p<0.05) in CCl4 concentration in typical wells

The Fig. 16 gives the change in CCl4 plume over time. In almost ten years, there was very little change in the distribution of CCl4 in the range of 3-10μg/L or 10-50 μg/L. There was a major reduction in the volume of groundwater containing concentrations between 50 and 300μg/L. The plume extended westward and eastward in northern sub-area to in the flood

Based on observed trends in the development of a plume, plumes can be grouped as four categories: expanding, stable, shrinking and exhausted (Rice at al., 1995). In this study, the length of pollution plume of CCl4 in the water-bearing aquifer was found to be stable while concentration is shrinking, which indicates that the plume decreased faster in concentration

Dynamic groundwater flow field is one of the most important factors controlling the CCl4 plume (Han et al., 2006). CCl4 plume in the aquifer is similar with the groundwater flow field, which is shown in Fig. 17. CCl4 diffusion was confined by higher water level around the plume (Zhu et al., 2008). The scope of seriously polluted wells was similar with the center of cone of depression. CCl4 concentration in the center of the local cone of depression

period every year. It should be noticed that the plume expanded eastward.

**5.2 Mann-Kendall trend tests** 

Data sample collected

Fig. 15. The variation of CCl4 over time in typical supply wells

common porous media aquifer. This may be due to the facts that: (1) The groundwater velocity in the aquifer is much higher than that in the porous media, which can reach 131.9m/h-3027.0 m/h. As a result, the advective flow dominates the movement of pollutant; (2) Local groundwater flow regime changed frequently. It was one of the important water supply source with over 80 wells owned by different departments for different purposes. Pump stopping and starting at different wells caused change in local flow field and subsequently cause change in CCl4 concentration; and (3) CCl4 transport channel is complex.

## **5.2 Mann-Kendall trend tests**

446 Pesticides in the Modern World - Risks and Benefits

CCl4 (μg/L)

2004-2-19

CCl4 (μg//L)

04-2-19

2004-6-30

2004-10-19

05-2-21

2005-4-14

2005-5-31

2005-7-18

Fig. 15. The variation of CCl4 over time in typical supply wells

2005-8-22

2005-10-13

2005-12-6

2006-2-8

2006-4-7

2006-6-26

2006-10-10

2007-3-28

Data sample collected

2007-7-11

2007-9-24

2008-1-23

2008-6-13

2008-9-17

2009-1-19

2009-9-28

2010-5-7

2010-5-26

2010-7-2

2010-8-10

2010-9-17

2010-12-7

CCl4 (ug/L)

2004-3-11

2004-6-30

2009-9-25

2004-11-29

2005-2-16

2005-3-22

2005-4-28

2005-6-11

2005-7-15

2005-8-11

2005-9-6

2005-10-18

2005-11-30

2006-1-11

2006-3-2

2006-4-11

2006-7-18

2006-9-27

2007-2-13

2007-5-12

Data sample collected

2007-8-5

2007-9-17

2007-11-28

2008-3-24

2008-7-1

2008-9-5

2008-11-27

2009-4-21

2009-10-27

2010-5-4

2010-5-19

2010-6-8

2010-7-6

2010-8-6

2010-9-3

2010-9-30

2010-11-12

X--56

2010-12-17

CCl4(μg/L)

04-2-19

2004-7-16

2004-11-1

05-2-21

2005-4-14

2005-5-31

2005-7-21

2005-9-1

2005-10-27

2005-12-21

2006-2-22

2006-4-18

2006-8-1

2006-11-14

2007-4-27

Data sample collected

2007-8-13

2007-10-17

2008-3-14

2008-7-11

2008-10-17

2009-3-20

2009-11-22

2010-5-12

2010-6-8

2010-7-13

2010-8-20

2010-9-27

Well X-59

2010-11-17

2004-6-10

2004-9-13

2004-12-15

2005-2-28

2005-3-28

2005-5-17

2005-7-5

2005-7-29

2005-8-25

2005-10-8

2005-11-22

2006-1-6

2006-2-22

2006-4-11

2006-6-1

2006-9-27

2007-2-13

Data sample collected

2007-5-12

2007-8-13

2007-9-24

2007-12-11

2008-5-13

2008-8-1

2008-10-17

2009-2-20

2009-8-28

2010-4-28

2010-5-14

2010-6-4

2010-7-2

2010-7-30

2010-9-3

2010-9-30

Well X-61

2010-11-12

CCl4 (μg/L)

2004-2-19

2004-4-29

2004-7-8

2004-9-16

2004-11-25

2005-2-3

2005-4-14

2005-6-23

2005-9-1

2005-11-10

2006-1-19

2006-3-30

2006-6-8

2006-8-17

2006-10-26

2007-1-4

2007-3-15

2007-5-24

2007-8-2

Data sample collected

2007-10-11

2007-12-20

2008-2-28

2008-5-8

2008-7-17

2008-9-25

2008-12-4

2009-2-12

2009-4-23

2009-7-2

2009-9-10

2009-11-19

2010-1-28

2010-4-8

2010-6-17

2010-8-26

Well X-47

Well X-49

2010-11-4

The non-parametric Mann-Kendall test was used to detect monotonic (increasing or decreasing) trends in time-series of CCl4 concentrations for the eight typical wells during the period 2004-2010. The Mann-Kendall test is widely used in environmental science for the detection of trends in time-series data. A 5% significance level was used to indicate statistically significant trends in the current study. The Mann-Kendall trend statistics (Z) indicates significant decreasing (Z <–1.96, p <0.05) and increasing (Z >1.96, p <0.05) trends. Table 5 presents that there are a highly significant decreasing trend in CCl4 in the karst aquifer and the decreasing trend in pollution source sub-area and the north sub-area area are more significant.


Table 5. The Mann-Kendall trend statistic (Z, p<0.05) in CCl4 concentration in typical wells
