Examining chemical interaction of stream and aquifer near a river
barrage area by factor analysis
YUN-YEONG OH1, SE-YEONG HAMM1*, HANG-TAK JEON1
CHUNG-MO LEE1, MINGLIANG WEI 1
1Division of Earth Environmental System
Pusan National University, Busan 46241
REPUBLIC OF KOREA
Abstract: - A statistical approach has been applied in order to evaluate chemical interaction between
groundwater and river water by using electrical conductivity (EC) in the Changnyeong-Haman river barrage
(CHRB), Korea. The EC values in groundwater have been decreased an average of 50 μS/cm and 160 μS/cm.
According to factor analysis using the EC data, the EC in groundwater has been disturbed by the CHRB
construction as well as dredging of the Nakdong River during the construction work and with time ealpse the
groundwater quality became stabilized. It is also demonstrated that hydrologic environment has been changed
due to the CHRB construction, comparing the 1st period and 4th period. Hence, a long-term monitoring is
required in order to reveal the change of hydrologic environment that can adversely affect plant growth.
Key-Words: River barrage, Stream-aquifer interaction, Electrical conductivity, Factor analysis
1. Introduction
The chemical property of groundwater depends
on many factors such as the mineralogy of aquifers,
the chemical compositions of the precipitation and
surface water, climate, topography, and
anthropogenic activities (Edmunds et al., 1982).
Surface water-groundwater interaction can be
revealed by using physical and chemical
components between stream and aquifer.
Comprehensive understanding of groundwater-
surface water interactions will also identify
migration of pollutants and potential impact to
ecological systems within the aquatic environment
(Jorge et al., 2015).
Surface water–groundwater interaction has been
studied using hydrochemical and isotopic end-
member mixing analysis to describe the influence
of water origins (e.g., Christophersen et al., 1990;
Hooper et al., 1990; Ladouche et al., 2001). The
surface water–groundwater interactions in the Jialu
River basin using major ion chemistry and stable
isotopes were studied by investigating temporal and
spatial variations in water chemistry affected by
humans and by characterizing the relationship
between surface water and groundwater in the
shallow Quaternary aquifer (Yang et al., 2012).
Pierre et al. (2012) revealed a clear evolution of
river water chemical composition that could be
related to the increasing amount of limestone
aquifers inside the Eau Blanche River basin. Hinkle
et al. (2001) collected isotopic and chemical data
from shallow hyporheic zone wells and
demonstrated interaction between regional ground
water and river water. Dixon-Jain (2008) studied
groundwater-surface water interaction and
implication for nutrient transport to tropical rivers.
USGS (2003) reported surface water and
groundwater interaction of the Spokane River and
the Spokane Valley-Rathdrum Prairie aquifer, by
means of physical properties and major ions, trace
elements, and stable isotopes.
This study aimed to evaluate chemical
interaction between groundwater and river water by
using factor analysis with electrical conductivity
(EC) data in the Changnyeong-Haman river barrage
(CHRB) area which is located in downstream of the
Nakdong River, Korea and mostly used for
agriculture. The study area (Figure 1) having an
elevation of 6–10 m above mean sea level is mostly
utilized by agricultural field among which rice
field occupies more than 84%. Groundwater is
importantly used for irrigation and its usage is
dependent to the agricultural cycles of rice farming
from May to August and greenhouse cultivation
from December to April. The CHRB was
constructed during the period from June 2011 to
September 2012. By the context of the Four Major
River Restoration Project (4MRRP) from 2008
until 2012, 16 river barrages have been constructed
on the four major rivers (the Han, Nakdong, Geum,
and Yeongsan) in South Korea.
2. Change of EC in the groundwater
and river
The electrical conductivity (EC) values in
MOLECULAR SCIENCES AND APPLICATIONS
DOI: 10.37394/232023.2022.2.2
Yun-Yeong Oh, Se-Yeong Hamm,
Hang-Tak Jeon, Chung-Mo Lee, Mingliang Wei
E-ISSN: 2732-9992
5
Volume 2, 2022
groundwater have been collected at the total of 33
monitoring well (HAM-004 – 065) from Jul. 1,
2011 to Jun. 30, 2015 as well as the EC values of
the Nakdong River at the Chilseo purification plant
from Jul. 1, 2 012 to Jun. 30, 2015 (Table 1). For
the 1st period (Jul. 1, 201 1 – Jun. 30, 2012 ), the
average EC value in groundwater was 678 µS/cm
with a range of 161 (at HAM- 062) to 1836 µS/cm
(at HAM-047). For the 2nd period (Jul. 1, 2012 –
Jun. 30, 2013), the average EC value in
groundwater was 621 µS/cm with a range of 160
(at HAM- 062) to 15496 µS/cm (at HAM-047). For
the 3rd period (Jul. 1, 201 3 – Jun. 30, 2014 ), the
average EC value in groundwater was 614 µS/cm
with a range of 149 (at HAM- 062) to 1483 µS/cm
(at HAM-047). Finally, for the 4th period (Jul. 1,
2014 – Jun. 30, 2015 ), the average EC value in
groundwater was 615 µS/cm with a range of 149
(at HAM- 062) to 1443 µS/cm (at HAM-047).
On the other side, 2013, the average EC values
in the river displayed 269 µS/cm, 263 µS/cm, and
234 µS/cm from the 2nd to 4th period, respectively.
The average EC values both in the groundwater and
the river decreased from Jul. 1, 2011 t o Jun. 30,
2015, indicating the change of hydrologic
environment due to the construction of the CHRB.
Table 1. Average electrical conductivity values from Jul. 2011 until Jun. 2015.
Wells
Average EC of GW (μS/cm)
Wells
Average EC of GW (μS/cm)
2011-
2012
2012-
2013
2013-
2014
2015
2011-
2012
2012-
2013
2013-
2014
2014-
2015
HAM-004EC 489 532 551 559 HAM-042EC 457 372 373 387
HAM-005EC 359 355 362 344 HAM-043EC 1547 1406 1401 1398
HAM-007EC 492 645 658 661 HAM-045EC 778 660 634 635
HAM-008EC 387 351 353 365 HAM-046EC 1027 201 185 180
HAM-010EC 1158 1260 1261 1256 HAM-047EC 1836 1549 1483 1443
HAM-011EC 1001 998 996 990 HAM-048EC 435 379 366 357
HAM-013EC 215 215 218 221 HAM-051EC 730 724 729 738
HAM-014EC 808 780 778 775 HAM-056EC 718 623 615 615
HAM-015EC 582 564 563 565 HAM-057EC 273 249 260 283
HAM-019EC 690 702 703 707 HAM-059EC 880 897 899 903
HAM-021EC 1455 1364 1335 1337 HAM-060EC 494 472 469 462
HAM-022EC 819 452 342 377 HAM-061EC 293 347 356 365
HAM-023EC 703 715 731 742 HAM-062EC 161 160 149 149
MOLECULAR SCIENCES AND APPLICATIONS
DOI: 10.37394/232023.2022.2.2
Yun-Yeong Oh, Se-Yeong Hamm,
Hang-Tak Jeon, Chung-Mo Lee, Mingliang Wei
E-ISSN: 2732-9992
6
Volume 2, 2022
HAM-035EC 606 535 550 559 HAM-063EC 402 403 401 401
HAM-038EC 422 540 560 588 HAM-064EC 344 286 271 288
HAM-040EC 948 795 757 767 HAM-065EC 477 418 413 420
HAM-041EC 376 536 549 467
July 2011 –
June 2012
July 2012 –
June 2013
July 2013 –
June 2014
July 2014 –
June 2015
Maximum EC
of average
values
1836
(HAM-047EC)
1549
(HAM-047EC)
1483
(HAM-047EC)
1443
(HAM-047EC)
Maximum EC
of average
values
161
(HAM-062EC)
160
(HAM-062EC)
149
(HAM-062EC)
149
(HAM-062EC)
Mean EC of
average values 678 621 614 615
Skewness EC
of average
values
1.289 1.177 1.100 1.073
Kurtosis EC of
average values 1.574 0.889 0.655 0.558
3. Factor analysis for electrical conductivity
Factor analysis is a kind of multivariate techniques
that simplify large data sets into smaller number of
groups and make useful generalizations that
provide meaningful insight (Lawrence and
Upchurch, 1982; Suk and Lee, 1999; Shim et al.,
2000; Hamm et al., 2000; Jeong, 2001; Lee and
Woo, 2003). Thus, the factor analysis can identify
groups of highly correlated original variables
(Papatheodorou et al., 2007). In this study, R-mode
factor analysis which estimates the relationships
between variables by extracting eigenvalues
and eigenvectors from a co variance or
correlation matrix, was carried out in order to
identify the EC characteristics of the groundwaters.
The factors were extracted using principal
component analysis using the procedure of
orthogonal varimax rotation and were determined
for eigenvalues higher than or equal to 1 ( Davis,
2002), using SPSS ver. 23 (IBM Corporation,
2015). The R-mode factor analysis of the EC
values of the 33 monitoring wells was carried out
for the four periods of one year unit (Jul. 2011 –
Jun. 2012, Jul. 2012 – Jun. 2013, Jul. 2013 – Jun.
2014, and Jul. 2014 – Jun. 2015). Total three
factors were identified by the R-mode factor
analysis.
For the first period (Jul. 2011 – Jun. 2012),
Factor 1 showed higher positive loadings to HAM-
005, 008, 013, 015, 040, 047, 051, and 056 (>0.70)
with a negative loading to HAM-007, 010, 01 9,
046, 048, and 064 which accounted for 42.68% of
the total variance with an eigenvalue of 14.04. This
factor is related to the influence of the Nakdong
River and tributaries (Fig. 1a). Factor 2 showed
higher positive loadings to HAM-011, 022, and 065
(>0.70) with a negative loading to HAM-014 and
060 which accounted for 22.05% of the total
variance with an eigenvalue of 7.28. This factor
may be related to the influence of the agricultural
activity (Fig. 1a). Factor 3 showed high positive
loadings to HAM-035 (>0.70) which accounted for
9.04% of the total variance with an eigenvalue of
2.98. This factor may reflect the influence of the
natural groundwater (Fig. 1a).
For the second period (Jul. 2012 – Jun. 2013),
Factor 1 showed higher positive loadings to HAM-
021, 022, 040, 045, 048, and 064 (>0.70) with a
negative loading to HAM-004 which explained
MOLECULAR SCIENCES AND APPLICATIONS
DOI: 10.37394/232023.2022.2.2
Yun-Yeong Oh, Se-Yeong Hamm,
Hang-Tak Jeon, Chung-Mo Lee, Mingliang Wei
E-ISSN: 2732-9992
7
Volume 2, 2022
31.62% of the total variance with an eigenvalue of
10.43. This factor may reflect the influence of the
Nakdong River and tributaries (Fig. 1b). Factor 2
showed higher positive loadings to HAM-008, 013,
019, 038, and 059 (>0.70) with a negative loading
to HAM-047 and 060 which accounted for 22.32%
of the total variance with an eigenvalue of 7.37.
This factor may reflect the influence of the
agricultural activity (Fig. 1b). Factor 3 showed high
positive loadings to HAM-051 (>0.70) which
explained 10.82% of the total variance with an
eigenvalue of 3.57. This factor may reflect the
influence of the natural groundwater (Fig. 1b). The
EC values and factors for the second period
indicate the disturbance of groundwater quality by
the construction of the CHRB.
For the third period (Jul. 2013 – Jun. 2014),
Factor 1 presented higher positive loadings to
HAM-004, 008, 013, 019, 038, 040, 042, 059, and
064 (>0.70) with a negative loading to HAM-005,
011, 041, 046, 047, and 060 which explained 44.29%
of the total variance with an eigenvalue of 14.62
(Fig. 1c), reflecting the higher influence of the
Nakdong River and tributaries than the second
period. Factor 2 presented high positive loadings to
HAM-014 (>0.70) with a negative loading to
HAM-021 which accounted for 16.49% of the total
variance with an eigenvalue of 5.44. This factor
may reflect the influence of the agricultural activity
(Fig. 1c). Factor 3 showed high positive loadings to
HAM-007, 010, a nd 048 (>0.70) which explained
9.47% of the total variance with an eigenvalue of
3.13. This factor may reflect the influence of the
natural groundwater (Fig. 1c). This pattern of
factors designates stabilization of groundwater
quality with time elapse since the completion of the
CHRB construction.
Finally, for the fourth period (Jul. 2014 – Jun.
2015), Factor 1 presented higher positive loadings
to HAM-013, 015, 019, 021, 035, 038, 040, 05 1,
062, and 064 (>0.70) with a negative loading to
HAM-010, 011, 014, 045, and 048 which explained
43.87% of the total variance with an eigenvalue of
14.48 (Fig. 1d), reflecting the influence of the
Nakdong River and tributaries. Factor 2 presented
high positive loadings to HAM-007 (>0.70) which
explained 15.72% of the total variance with an
eigenvalue of 5.19. This factor may reflect the
influence of the agricultural activity (Fig. 1d).
Factor 3 s howed high positive loadings to HAM-
061 (>0.70) which explained 15.72% of the total
variance with an eigenvalue of 3.57. This factor
may reflect the influence of the natural
groundwater (Fig. 1d).
According to the pattern of factors for the total
periods, the EC values have been stabilized with
time elapse since the completion of the CHRB
construction, with indicating the effect of the water
quality disturbance by the CHRB construction and
dredging of the Nakdong River during the
construction work.
4. Conclusions
This study interpreted the chemical interaction
between the stream and aquifer by using factor
analysis of electical conductivity (EC) in
groundwater for the four periods with year nuit
from July 2011 to June 2015, associated with the
effect of the construction of Changnyeong-Haman
river barrage (CHRB). Based on the factor analysis
and the EC variation with time elapse, it is
concluded that the EC values has been disturbed by
the CHRB construction and dredging of the
Nakdong River during the construction work and
then the groundwater quality became stabilized. In
addition, it is demonstrated that hydrologic
environment has been changed due to the CHRB
construction, comparing the 1st period and 4th
period. Hence, a long-term monitoring of chemical
constituents in groudnwater is required in order to
reveal the change of hydrologic environment that
can adversely affect plant growth.
ACKNOWELEGEMENT
This study was supported by the Basic Science
Research Program through the National Research
Foundation of Korea (NRF) funded by the Ministry
of Education (NRF-2013R1A1A2058186) and also
financed by the research project “Advanced
Technology for GW Development and Application
in Riversides (Geowater+)” in the “Water
Resources Management Program (code 11
Technology Innovation C05)” of the Ministry of
Land, Infrastructure and Transport (MLIT) and the
Korea Agency for Infrastructure Technology
Advancement (KAIA).
MOLECULAR SCIENCES AND APPLICATIONS
DOI: 10.37394/232023.2022.2.2
Yun-Yeong Oh, Se-Yeong Hamm,
Hang-Tak Jeon, Chung-Mo Lee, Mingliang Wei
E-ISSN: 2732-9992
8
Volume 2, 2022