Sie sind auf Seite 1von 15

Available online at www.sciencedirect.

com

Journal of Hydro-environment Research 1 (2007) 95e109


www.elsevier.com/locate/jher

Development of integrated watershed management schemes


for an intensively urbanized region in Korea
Kil Seong Lee, Eun-Sung Chung*
School of Civil, Urban & Geosystem Engineering, Seoul National University, San 56-1, Sillim-dong, Gwanak-gu, Seoul 151-744, Republic of Korea
Received 22 December 2006; revised 27 April 2007; accepted 12 July 2007

Abstract

A systematic, seven-step approach to integrated watershed management for sustainability was proposed and applied to upstream watershed of
the Anyangcheon in Korea, which experiences streamflow depletion, frequent flood damages, and poor water quality due to rapid urbanization.
To understand watershed components and processes, static and dynamic data were collected and synthetic hydrologic cycles generated by HSPF
(Hydrologic Simulation Program e FORTRAN) were simulated (STEP 1). To identify and quantify problems within the watershed, three indices
(following the pressureestateeresponse model) were employed: Potential Flood Damage (PFD), Potential Streamflow Depletion (PSD), and
Potential Water Quality Deterioration (PWQD). Composite programming, a method of multi-criteria decision-making, was employed to estimate
all indices and analytic hierarchy process are introduced to quantify the weighting values of all indicators (STEP 2). The primary goal of man-
agers is to maintain certain minimum levels of water for instreamflow requirement and total maximum daily load (TMDL). Therefore target
water quality and, instreamflow requirements (including low flow and fish flow) were specifically set (STEP 3). All possible management al-
ternatives were listed (STEP 4) and a few specific management options which are technically, economically, and environmentally feasible,
were selected (STEP 5). The ability of each feasible option to achieve the desired water quantity and quality criteria was analyzed and quantified
using the HSPF (STEP 6). Finally, an evaluation index was calculated using each of the proposed alternatives in order to rank the sustainability
and priority of alternatives (STEP 7).
Ó 2007 International Association for Hydraulic Engineering and Research, Asia Pacific Division. Published by Elsevier Ltd. All rights reserved.

Keywords: Sustainability; Integrated watershed management; Composite programming; HSPF

1. Introduction source pollution from urban regions decrease water quality.


Korea is no exception. Shim (2003) reported 543 of 3773
Many urbanized watersheds in the world suffer from stream- second-graded streams in Korea are in danger of depletion
flow depletion and poor stream quality, which often negatively and many researchers showed that urban runoff quality has be-
affect instream and near-stream ecologic integrity, as well as come a serious problem and may even have fatal consequences.
water supply. Such distortions in the hydrologic cycle mainly Therefore, single-purpose planning efforts, which do not con-
result from rapid increase of impermeable area due to urbani- sider other relevant factors, are not a desirable management
zation, decreases of baseflow runoff due to groundwater pump- tool because all hydrologic components in a watershed are
ing, and reduced precipitation inputs driven by climate forcing. closely inter-related.
As well, combined sewer overflows and increase of non-point In recent years, water management has become more com-
plicated, not only because of the increasingly sophisticated
ways humans use water, but also as a result of changing atti-
* Corresponding author. Address: 34-302, School of Civil, Urban & tudes towards sustainability, which has become a fundamental
Geosystem Engineering, Seoul National University, San 56-1, Sillim-dong,
Gwanak-gu, Seoul 151-741, Republic of Korea. Tel.: þ82 2 880 8345; fax:
concept in resource planning and management. Sustainable de-
þ82 2 877 8170. velopment was defined by the World Commission on Environ-
E-mail address: cool77@snu.ac.kr (E.-S. Chung). ment and Development as ‘‘development that meets the needs

1570-6443/$ - see front matter Ó 2007 International Association for Hydraulic Engineering and Research, Asia Pacific Division. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.jher.2007.07.004
96 K.S. Lee, E.-S. Chung / Journal of Hydro-environment Research 1 (2007) 95e109

of the present generation without compromising the ability of indices and analytic hierarchy process are introduced to quan-
future generations to meet their own needs’’. Water used to tify the weighting values of all indicators (STEP 2). The pri-
be regarded as a readily available natural resource which did mary goal of managers is to maintain certain minimum
not need protection, and single-purpose planning efforts were levels of water for instreamflow requirement and total maxi-
considered successful if they achieved their specific perfor- mum daily load (TMDL). Therefore target water quality
mance criteria. To achieve sustainability, however, all aspects and, instreamflow requirements (including low flow and fish
of resource benefit and effects should be considered simulta- flow) were specifically set (STEP 3). All possible management
neously. An integrated approach is essential in order to accom- alternatives were listed (STEP 4) and a few specific manage-
plish this. For watershed resources to be sustainable, the long- ment options which are technically, economically, and envi-
term stability of social and economic systems as well as natural ronmentally feasible, were selected (STEP 5). The ability of
resources should be simultaneously considered. Heathcote each feasible option to achieve the desired water quantity
(1998) said that integrated water management can be consid- and quality criteria was analyzed and quantified using the
ered effective when the four following criteria are satisfied. HSPF (Hydrologic Simulation Program e FORTRAN)
(STEP 6). Finally, an evaluation index was calculated using
1. It allows an adequate supply of water that is sustainable each of the proposed alternatives in order to rank the sustain-
permanently. ability and priority of alternatives (STEP 7).
2. It maintains water quality at levels that meet government
standards and other social water quality objectives. 2.2. Sustainable development indicator
3. It minimizes flood damage.
4. It allows sustainable economic development over the short Because sustainability is a function of various economic,
and long term. environmental, ecological, social and physical goals and ob-
jectives, water resources management must inevitably involve
Hence, to ensure sustainable development, decisions about multi-objective tradeoffs in multi-disciplinary and multi-
water management are now widely recognized as being multi- participatory decision making process. Therefore, various
ple-objective problems. ways to measure sustainability have been developed. One
Integrated watershed management (IWM), is defined as the way is to express relative levels of sustainability as separate
management of a watershed system with sustainable technolog- or weighted combinations of reliability, resilience, and vulner-
ical options, which ensures the sustainability of land, agriculture ability measures; these various criteria contribute to human
and forestry practices and will conserve natural resources, while welfare and vary over time and space.
allowing adequate institutional and economic options. Multi- There are several frameworks around which indicators can
Criteria Decision Making (MCDM) techniques are therefore be developed and organized. There is no unique framework
gaining importance as useful tools for complex real-world prob- that generates sets of indicators for every purpose. A frame-
lems because of their inherent ability to assess alternative sce- work may also change over time as scientific understanding
narios and identify the best option. Connell et al. (2000) also of environmental problems increases and as societal values
suggested MCDM as one of the approaches for developing evolve. In the context of the work of the Group on the State
sustainable water resources planning and management. of the Environment, the Pressure-State-Response (PSR)
This study develops a formal procedure of integrated water- framework has been used (Fig. 2). The PSR model considers
shed management schemes to accomplish sustainable develop- those human activities which exert pressures in the environ-
ment using MCDM technique. It further demonstrates a master ment and affect its quality and the quantity of natural re-
plan to rehabilitate the distorted hydrologic cycle of urbanized sources (state); society responds to these changes through
watershed in Korea. environmental, economic and sectoral policies and through
changes in awareness and behavior (societal response). The
2. Methodologies PSR model has the advantage of highlighting these links, help-
ing both decision makers and the public recognize environ-
2.1. Integrated watershed management mental and other issues as interconnected. This PSR model
was used to quantify the hydrologic risk in STEP 2 and the
A general, systematic, seven-step process (Fig. 1) was in- effectiveness of feasible alternatives in STEP 7.
troduced to organize an integrated approach to watershed plan-
ning and management. To understand watershed components 2.3. Composite programming
and processes, static and dynamic data were collected and syn-
thetic hydrologic cycles generated by continuous simulation Composite programming (CP), which is a multi-level/multi-
model (STEP 1). To identify and quantify problems within objective programming method, was introduced as an empirical
the watershed, three indices considering sustainability concept technique to resolve a geological exploration problem by Bar-
were employed: Potential Flood Damage (PFD), Potential dossy and Bogardi (1983). A general multi-objective problem
Streamflow Depletion (PSD), and Potential Water Quality De- can be transformed to a single objective problem. This transfor-
terioration (PWQD). Composite programming, a method of mation is done via a step-by-step regrouping of a set of objec-
multi-criteria decision-making, was employed to estimate all tives into a single objective.
K.S. Lee, E.-S. Chung / Journal of Hydro-environment Research 1 (2007) 95e109 97

Step 1 Understanding watershed components and processes

1. Watershed components
- Static: bedlock geology, soils, landuse, plant and animal
communities, social and economic system
- Dynamic: climate, runoff, streamflow, groundwater, water quality
2. Watershed processes
- Water quantity simulation
- Water quality simulation

Step 2 Identifying and ranking problems to be solved

1. Potential Streamflow Depletion (PSD)


2. Potential Flood Damage (PFD)
3. Potential Water Quality Deterioration (PWQD)

Step 3 Setting clear and specific goals

1. Target quantity: instreamflow requirement =


Max (Low flow, Environmental Flow from PHABSIM)
2. Water quality standard and total maximum daily load (TMDL)

Step 4 Developing a list of management options

Step 5 Eliminating infeasible options

Criteria: Technical, Environmental, Economic Feasibility

Step 6 Testing the effectiveness of remaining feasible options

1. Change of drought flow (Q355) and low flow (Q275)


2. Change of pollutant concentration and daily loads duration
curves

Step 7 Developing the final options

Evaluation Index
1. Increase of 275th and 355th values of flow duration curve
2. Decrease of average pollutant concentration and total daily
loads

Fig. 1. Integrated watershed management procedure.

CP employs a double-weighting mechanism. One set of determined, the first step is to normalize the basic values
weights are indicators which articulate the decision-maker’s (transposing them into the range of 0e1). This is undertaken
preferences regarding the relative importance of each indica- to make all indicators comparable to each other, thereby
tor. The other set are balancing factors given to groups in avoiding differences in units. Given the ideal value (sideal ),
which any numbers of indicators are involved. Unlike weights, and the worst value (sworst ), the normalized value (si ) of an
balancing factors are associated with groups rather than with actual indicator value (si ) can be calculated as follows:
each indicator. While the choice of weights emphasizes the
relative importance of the indicators to each other, selecting sideal  si
the balancing factors identifies how larger deviations in groups si ¼ ð1Þ
sideal  sworst
of indicators may affect the process. The purpose of high bal-
ancing factors is to give more emphasis to the indicators which
have large negative values (Goicoechea et al., 1982). where the choice is made to ensure that the si to be used in the
Once the relevant indicators, associated boundary values following equation represents the relative position with respect
(ideal and worst values), actual values and weights are to the best value. The next step is to calculate second-level
98 K.S. Lee, E.-S. Chung / Journal of Hydro-environment Research 1 (2007) 95e109

Fig. 2. Concept of Pressure-State-Response Model (OECD, 1993).

composite distances for each second-level group of basic indi- ecosystem integrity, recreation, scenery, etc. In this study,
cators using the following equation: only the minimum flow for fish habitat (called the fish flow)
!1=bj is considered and compared with the hydrological low flow.
X
Nj
b
Lj ¼ wij sijj ð2Þ 2.4.1. Low flow
i¼1
Average low flow is the main index used in Korea. The
where i is the sequential number given to a basic indicator, j mean values of daily flow exceed historical low flow 97% of
the sequential number of a certain group of basic indicators, the time (across annual sampling periods). However, average
sij the value of the basic index si within the second-level group low flow has to be estimated by using the data of gauged ba-
j, Lj the distance from the ideal point in second-level group j, sins or the results of hydrologic models for ungauged basins.
Nj the number of basic indicators in a second-level group j, wij Methods of low flow estimation in ungauged basins include
the weights expressing the relative importance of the Nj basic the drainage-area method, regional regression method, base-
indicators in group j, the sum of weights in any group being flow correlation method, etc. (Riggs, 1972; Stedinger and
equal to one, bj the balancing factor, which is equal or greater Thomas, 1985; Vogel and Kroll, 1992). Regional regression
than 1, among indicators within the group j. The consecutive method was used in this study.
computations of higher-level composite indices are made in
the same manner until a final composite distance for a system 2.4.2. Fish flow
is reached. Lj will be values of PFD, PSD, and PWQD. The PHABSIM (physical habitat simulation system; US Geo-
additional information can be obtained by Hartmann et al. logical Survey, 2001) is a specific model designed to calculate
(1987), Lee et al. (1991), Lee et al. (1992), Hagemeister an index to the amount of microhabitat available for different
et al. (1995) and Yurdusev and O’Connell (2005). life stages of fish at different flow levels. PHABSIM has two
CP uses indicators from different categories to calculate major analytical components: stream hydraulics and life
a composite distance, which identifies the distance of the ac- stage-specific habitat requirements. PHABSIM calculates the
tual system from the ideal state. Hence, schemes with small relationship between weighted usable area (WUA) and dis-
composite distances are closer to the ideal state than those charge by combining hydraulic modeling for different flow
with large composite distances (Yurdusev and O’Connell, levels and habitat modeling that focuses on habitat suitability
2005). criteria of the target species. After obtaining the relationship
between WUA and discharge, the discharge for the maximum
2.4. Instreamflow requirement WUA is selected as a fish flow to preserve ecosystem integrity.

Instreamflow requirement in STEP 3 is defined as the min- 2.5. Hydrological simulation program e FORTRAN
imum flow which must flow in a stream. Generally, instream-
flow requirements are the maximum value of either low flow HSPF (Hydrological Simulation Program e Fortran; Bick-
or environmental control flow. Environmental control flow nell et al., 2001) is a watershed model that simulates runoff
can be composed of multiple factors such as water quality, and non-point pollutant loads leaving a watershed and
K.S. Lee, E.-S. Chung / Journal of Hydro-environment Research 1 (2007) 95e109 99

performs the fate and transport processes in streams and one- digital elevation model (DEM), stream network, and storm
dimensional lakes. HSPF is comprised of three main modules sewers, it was divided into 9 sub-watersheds (OJ, WG, DJ,
(PERLND, IMPLND, and RCHRES) and five utility modules. SB, HU, SA, SB1, SS, and the remaining sub-watershed;
For simulation with HSPF, the watershed has to be represented shown in Fig. 3). The watershed area, in which approximately
in terms of land segments (pervious and impervious lands) and 0.98 million people reside, is 127.13 km2 (population density:
reaches. The PERLND module represents hydrology and wa- 7708 persons per km2). Primary land cover types within the
ter quality processes that occur on pervious land segment, watershed (as of the year 2000) consist of 38.3% urbanized,
while IMPLND may be used for impervious surface areas 51.6% forest and 7.5% agricultural areas. Characteristic of
where little or no infiltration occurs. The RCHRES module all sub-watersheds were shown in Table 1.
simulates processes that occur in a single reach of an open
channel or well-mixed impoundment.
4. Application

2.6. Pollutant loadings (PLOAD) 4.1. Understanding watershed components


and processes (STEP 1)
PLOAD (Edwards and Miller, 2001) is a GIS-based model
used to calculate pollutant loads for watersheds. PLOAD esti- 4.1.1. Watershed components
mates non-point sources (NPS) of pollution on an annual aver- Water movement in a system is affected by many physical,
age basis for any user-specified pollutant. The user may chemical, and biological components (or features) and processes.
calculate the non-point source loads using either the export An understanding of these components and processes is an essen-
coefficient or the EPA’s simple method approach. tial first step to assess the condition of a watershed and the impacts
of management actions on the system (Heathcote, 1998).
3. Description of study watershed Understanding watershed components generally requires
static and dynamic data. Static data are changeless or slowly
The upstream watershed of the Anyangcheon (AYC) was changeable; dynamic data are gradually or rapidly changeable.
selected. AYC is the first tributary of the Han River in Korea. Static data consist of physical features, soils, population, flood
The study stream has a length of 17.91 km. Based on the damage, land-use, plant and animal systems and valued features

Fig. 3. Location of upstream watershed of the Anyangcheon and its sub-watersheds.


100 K.S. Lee, E.-S. Chung / Journal of Hydro-environment Research 1 (2007) 95e109

Table 1
Characteristic of the sub-watersheds
Name of Area (km2) Length of Slope Landuse (2000) (%)
sub-watershed stream (km) Urban Forest Agricultural
Entire 127.13 17.91 1/250 38.3 51.6 7.5
WG 3.78 3.82 1/50w1/60 6.96 81.51 9.43
OJ 4.26 2.85 1/30w1/60 7.65 77.48 11.38
DJ 5.35 4.02 1/270 57.09 27.73 11.21
1/140w1/180
SB 10.29 4.32 1/160 40.34 48.49 9.15
HU 44.58 9.26 1/410 22.82 57.47 15.7
1/140w1/160
SA 8.07 6.49 1/90w1/40 18.90 72.61 5.12
SS 13.17 5.74 1/120 7.88 83.21 8.16
1/20w1/50
SB1 4.59 2.76 1/60w1/100 11.52 68.91 11.5

and activities. The dynamic data are streamflow, groundwater, (TN), and total phosphorus (TP) per unit area for all study
and water quality (BOD, COD, SS, pH, temperature, .). sub-watersheds and the entire study watershed. Urbanized wa-
tershed such as DJ, SB, HU, SA, and SB1 emit higher pollutant
4.1.2. Watershed processes loads to their streams than the others, which suggests that these
Understanding watershed processes generally requires con- sub-watersheds require best management practices to improve
tinuous data for both water quantity and quality. Since it is im- water quality.
possible to get such data for every important site, hydrologic
simulation models are usually necessary to identify the water 4.2. Identifying and ranking problems (STEP 2)
quantity and pollutant values.
All hydrologic cycles, biological oxygen demand (BOD), The study watershed simultaneously suffers from stream-
concentrations and BOD total daily loads of study sub-water- flow depletion, frequent flood damages, and poor water qual-
sheds were derived from simulation results of HSPF which ity. Composite programming was applied to calculate PFD,
was calibrated and verified in Lee (2007). Results are shown PSD, and PWQD and the structure describing relationships
in Table 2 and Fig. 4. Sub-watersheds with high urban area ra- among those indices and their indicators are shown in
tio (DJ, SB, and HU) show more direct runoff and smaller Fig. 6. PFD, PSD and PWQD consist of pressure, state, and
baseflow than the others (WG, OJ, SS). Sub-watersheds with response which are based on the basic framework of sustain-
steep slopes (>1/100; SB1, SA) show high direct runoff and able development. Weights of indicators were established us-
low baseflow ratios even though they have small urban area ra- ing Analytic Hierarchy Process (AHP; Satty, 1977) Survey
tios (<10%). From Fig. 4, DJ shows the highest BOD concen- was conducted to 31 local government officials and re-
tration of streamflow due to the high urban area ratio and searchers working on river management. The results of
intrusion of sewage. Sub-watersheds OJ, SB, and SB1 show weighting values are shown in Table 3. Based on the sugges-
average concentrations from 10 to 15 mg/L and WG, SS, tion of Hartmann et al. (1987), all sub-watersheds can be re-
and SA under 1.5 mg/L. The average concentration at the spectively divided into three groups as to each index: sound
sub-watershed outlet is 9.0 mg/L. On the other hand, the total state (0e0.3), acceptable state (0.3e0.6), poor state (0.6e1).
annual loads per unit area (unit load) of all sub-watersheds are
not broad, and range from 30.3 kg/ha (SS) to 63.1 kg/ha (DJ). 4.2.1. PFD
The entire study watershed emits large unit loads due to PFD measures the vulnerability of a watershed to flooding
a wastewater treatment plant. These results indicate that sub- with respect to hydrology, society and economy. PFD of this
watersheds DJ, OJ, and SB1 need management action to en- study was revised and calculated by CP using the same indica-
hance water quality and lessen total BOC loads. tors as in previous research (Kim, 2004). The equation to cal-
PLOAD was used to identify the yearly change of pollutant culate PFD of the sub-watershed, n, is shown in Eq. (3).
(BOD, COD, SS, TN, TP, .) loads from land-use transforma- n 
tion. PLOAD requires two kinds of input data, such as land PFDn ¼ w1;3 b1;1 sPD;n þ b1;2 sPV;n þ b1;3 sSOC;n
use maps and loading values of land use (urban, agricultural, for- b 
þ b1;4 sNCR;n þw1;2 b1;5 sRI;n þ b1;6 sUR;n þ b1;7 sSW;n
est, and pasture). Since non-point source pollution from urban b 
areas occupies a major part of the total, its loading value plays þ b1;8 sFD;n þw1;3 b1;9 sLO;n þ b1;10 sNR;n
a critical role in this process. Therefore, actual pollutant loads b o1=b
of urban areas (Chung et al., 2006) were carefully estimated, þ b1;11 sSO;n ð3Þ
with many data measured 4 times at the outlet of the study wa-
tershed. Fig. 5 shows the calculated loads of BOD, chemical where weighting values (b) and names of parameters (s) are
oxygen demand (COD), suspended solids (SS), total nitrogen shown in Table 3.
K.S. Lee, E.-S. Chung / Journal of Hydro-environment Research 1 (2007) 95e109 101

(1.3%)
40.0 350

(47.8%)
(15.0%)
(44.0%)
(100%)

(7.0%)
(2.4%)
Concentration
35.0 Unit loads 300

Unit load (kg/ha/year)


Concentration (mg/L)
17.1
94.5
32.7
1355.6
648.6
203.5
596.9
30.0

Entire
250
25.0
200
20.0

(7.3%)
(50.4%)
(19.4%)
(45.8%)
(100%)
150

(8.1%)
(3.0%)
15.0
100
10.0

99.1
40.4
1331.7
660.1
253.5
621.2
109.1
SB1

5.0 50

0.0 0
WG OJ DJ SB HU SS SA SB1 Entire
(2.3%)
(46.7%)
(20.5%)
(43.5%)
(100%)

(9.4%)
(2.6%)

Fig. 4. BOD concentrations and unit loads of all sub-watersheds and the entire
watershed.
35.5
29.7
1331.7
612.3
267.2
587.1
126.6
SA

GIS software Arcview 3.2 was used to analyze geographic


(1.4%)
(46.1%)
(20.4%)
(43.1%)
(100%)

(9.4%)
(2.7%)

characteristics and the urban area ratio; HECeRAS and


HECegeoRAS were used to find the defense vulnerability of
41.1
20.8
1528.9
705.5
312.4
659.6
143.5

levee overflow and extended surcharge. The values of other in-


dicators were obtained by the Korea National Statistical Office
SS

(KNSO).
The average values and grades of PFD are reported in Table 4.
(2.5%)
(47.0%)
(19.8%)
(44.0%)
(100%)

(8.8%)
(2.6%)

All sub-watersheds were divided into three groups: sound state


(WG, OJ, SS), acceptable state (HU), poor state (DJ, SB, SA,
35.1
33.9
1340.8
630.4
265.3
595.5
118.6
Annual average values of hydrological cycle components from HSPF simulation results (1996w2005) (unit: mm)

SB1). Since the average PFD value of the study sub-watersheds


HU

is lower than that of entire AYC watershed, it can be concluded


that the watershed is usually safe.
(4.6%)
(48.2%)
(20.3%)
(44.5%)
(100%)

(9.1%)
(2.8%)

4.2.2. PSD
37.4
61.5
1331.7
632.3
264.5
601.3
122.3

PSD quantifies streamflow depletion using factors such as


SB

streamflow seepage, pumping from streams and groundwater,


channel slope, and other factors. PSD of this study was mod-
(5.4%)

ified to be calculated by CP. The equation to calculate PSD of


(49.0%)
(19.6%)
(45.2%)
(100%)

(8.5%)
(2.8%)

the sub-watershed, n, is shown in Eq. (4).


36.9
73.0
1331.7
642.2
255.7
611.2
114.4

n
DJ

 b 
PSDn ¼ w2;1 b2;1 sPD;n 2 þw2;2 b2;2 sSS;n þ b2;3 sUR;n þ b2;4 sGW;n
b 
(4.3%)
(48.3%)
(22.6%)
(40.5%)
(11.3%)
(100%)

þ b2;5 sSW;n 2 þw2;3 b2;6 sRT;n þ b2;7 sNR;n þ b2;8 sIT;n


(4.2%)

b o1=b2
þ b2;9 sUG;n 2 ð4Þ
1331.7
632.3
294.4
547.8
151.3
55.7
55.4
OJ

(3.1%)

where weighting values (b) and names of parameters (s) are


(48.1%)
(22.7%)
(39.4%)
(11.4%)
(100%)

(4.2%)

shown in Table 3.
Values of indicators were obtained by many field surveys,
1331.7
629.6
296.2
532.9
152.7
56.2
39.7

KNSO, and Arcview software. The values and grades of


WG

PSD are reported in Table 4. All sub-watersheds were divided


into three groups: sound state (HU), acceptable state (OJ, DJ,
Change in storage
Deep percolation

SB, SS, SB1), poor state (WG, SA). Since the average PSD
value of study sub-watersheds is larger than that of entire
Direct runoff
Precipitation

Base runoff
Component

Infiltration

AYC watershed and is close to 0.60, which is the lower limit


Total ET
Table 2

of poor condition, it can be concluded that the study sub-


watersheds have high probabilities of streamflow depletion.
102 K.S. Lee, E.-S. Chung / Journal of Hydro-environment Research 1 (2007) 95e109

600

500
WG

Pollutant loads (kg/ha/year)


OJ
400
DJ

SB
300
HU

SA
200
SS
SB1
100 Entire

0
1975 2000 1975 2000 1975 2000

BOD COD SS

40

35
WG
Pollutant loads (kg/ha/hear)

30 OJ

DJ
25
SB
20 HU

SA
15
SS

10 SB1

Entire
5

0
1975 2000 1975 2000

TN TP

Fig. 5. Pollutant (BOD, COD, SS, TN, and TP) annual loads per unit area for the 8 AY sub-watersheds and the entire watershed (Outlet).

4.2.3. PWQD All loads were obtained from the STEP 1 (PLOAD simula-
PWQD quantifies the possibility of water quality deteriora- tion results) and the remaining indicators were examined by
tion. The equation to calculate PWQD of the sub-watershed, n, many field surveys and KNSO. The values and grades of
is shown in Eq. (5). PWQD are reported in Table 4. All sub-watersheds were di-
vided into three groups: sound state (WG, OJ, HU, SA, SS,
n  b  SB1), acceptable state (DJ, SB), poor state (None). Since the
PWQDn ¼ w3;1 b3;1 sP;n 3 þw3;2 b3;2 sLB;n þ b3;3 sLC;n
average PWQD value of the study sub-watersheds is lower
þ b3;4 sLS;n þ b3;5 sLP;n þ b3;6 sLN;n þ b3;7 sWI;n than that of entire AYC watershed and is included in the sound
b  state, the study watershed is relatively safe as a whole.
þ b3;8 sPD;n þ b3;9 sCS;n 3 þw3;3 b3;10 sST;n
b o1=b3 4.3. Setting clear and specific goals (STEP 3)
þ b3;11 sRS;n 3 ð5Þ
The previous step identified that streamflow depletion is the
where weighting values (b) and names of parameters (s) are most serious problem in the study watershed. Therefore, the
shown in Table 3. ultimate goal of this study is to devise feasible management
K.S. Lee, E.-S. Chung / Journal of Hydro-environment Research 1 (2007) 95e109 103

Basic Indicators
2nd level 3rd level 4th level
Population density
Property value
Damage Object
Number of infrastructure
Number of natural and cultural resources
Rainfall intensity Potential
Urban area ratio Damage Possibility Flood
Geographic Characteristics Damage (PFD)
Number of levees
Number of reservoir Indefensibility
Surcharged Overflow

Population Damage Object


Streamflow Seepage Potential Watershed
Urban Area Ratio Streamflow Evaluation
Damage Possibility Depletion Index
Groundwater Withdrawal
(PSD) (WEI)
Slope of Watershed
Reuse of Treated Wastewater
Reservoir
Defense Vulnerability
Interbasin Transfer
Use of Groundwater Collected
by Subway Stations
Population Damage Object
Loading of BOD
Loading of COD
Loading of TSS Potential
Water Quality
Loading of TP Damage Possibility Deterioration
Loading of TN (PWQD)
Untreated wastewater intrusion
Population density
Covered stream
Streamflow treatment facility
Riverside and street sweeping Defense Vulnerability

Fig. 6. The indicator structure of indices.

alternatives that can recover the distorted hydrological cycle method (Vogel and Kroll, 1992), and the baseflow correlation
and consequently, the depleted streams of the watershed to method (Stedinger and Thomas, 1985). After applying all
a certain target, which is often called the instreamflow require- three methods to the study watershed, this study concluded
ments. But since water quality is closely related to water quan- that the regional regression method, which estimates average
tity, the target water quality was also set. This section low flow as a function of the basin characteristics, is the
addresses how the instreamflow requirements are calculated most appropriate. This study used natural inflow data recorded
and what water quantity and quality was determined. in six dams and nine gauging to identify a regression equation
and calibrate parameters. After testing the candidate models, it
was found that the watershed area-only model was the best,
4.3.1. Target water quantity which is written as:

4.3.1.1. Low flow. The numerical definition of hydrological


low flow varies by country. In Korea, the term, ‘‘the average Q ¼ 0:0357A0:55 ð6Þ
low’’, is often used as the low flow index. The average low
flow is defined as the mean value of annual daily flows that ex-
ceed 355 days of a year. However, historical flow records that where Q is the average low flow (m3/sec), and A is the water-
are sufficient to establish a statistically reliable value are sel- shed area (km2). The second column of Table 5 presents the
dom available in Korea; the study watershed is no exception. average low flow for the 8 study sub-watersheds from regional
This study therefore reviewed low flow estimation methods regression equation.
for ungauged basins.
As reviewed by Smakhtin (2001), many low flow estima-
tion methods for ungauged basins are available in the hydrol- 4.3.1.2. Fish flow. Stream hydraulic modeling simulates water
ogy literature, but the most popular choices include the depths and velocities as a function of discharge. PHABSIM
drainage-area method (Riggs, 1972), the regional regression uses the HEC-RAS and HECegeoRAS model for the water
104 K.S. Lee, E.-S. Chung / Journal of Hydro-environment Research 1 (2007) 95e109

Table 3
Weighting values of all indicators using AHP
Name of index Sustainability component Weighting value Components of indicators Weighting value
PFD (21/31) Pressure (19/31) 0.372 Property value ðsPV;n Þ 0.208
Population density ðsPD;n Þ 0.350
Infrastructure ðsSOC;n Þ 0.275
Natural & cultural resources ðsNCR;n Þ 0.166
State (22/31) 0.293 Rainfall intensity ðsRI;n Þ 0.282
Urban area ratio ðsUR;n Þ 0.256
Watershed slope ðsSW;n Þ 0.221
Amount of flood damage ðsFD;n Þ 0.241
Response (23/31) 0.335 Stability of levee inundation ðsLO;n Þ 0.392
Number of pumping station ðsSO;n Þ 0.268
Number of reservoirs ðsNR;n Þ 0.341

PSD (22/31) Pressure 0.371 Population density ðsPD;n Þ 1.000


State (24/31) 0.375 Streamflow depletion/diversion ðsSS;n Þ 0.219
Urban area ratio ðsUR;n Þ 0.373
Groundwater withdrawal ðsGW;n Þ 0.274
Watershed slope ðsSW;n Þ 0.134
Response (20/31) 0.254 Reuse of treated wastewater ðsRT;n Þ 0.270
Reservoir capacity ðsNR;n Þ 0.342
Use of groundwater collected by subway stations ðsUG;n Þ 0.196
Diversion from other watershed ðsIT;n Þ 0.192

PWQD (24/31) Pressure 0.302 Population ðsP;n Þ 1.000


State (24/31) 0.388 BOD loads ðsLB;n Þ 0.094
COD loads ðsLC;n Þ 0.093
SS loads ðsLS;n Þ 0.093
TN & TP loads ðsLN;n Þ 0.093
Intrusion of wastewater ðsWI;n Þ 0.234
Population density ðsPD;n Þ 0.219
Ratio of covered length ðsCS;n Þ 0.274
Response 0.310 Streamflow treatment facility ðsST;n Þ 1.000
Riverside and street sweeping ðsRS;n Þ 0.000

depth simulation and the VELSIM model, which uses Man- Gyeonggi Research Institute (2003), who investigated the
ning’s equation to simulate velocities. dominant species for the study watershed, and thus selected
This habitat modeling derives a relationship between Carassius auratus (Goldfish) for the Gia bridge, Zacco platy-
weighted usable area and discharge by combining the hydrau- pus for the HU sub-watershed, Rhynchocypris oxycephalus for
lic modeling result with the habitat suitability criteria of target the DJ, SB, SS, and SB1 sub-watersheds as the dominant spe-
species. To determine the target species, this study cited cies. This study also used information from former studies,

Table 4
Values of PFD, PSD, and PWQD
Subject PFD PSD PWQD
Value* Grade Value* Grade Value* Grade
WG 0.380 Sound 0.490 Poor 0.345 Sound
OJ 0.435 Sound 0.474 Acceptable 0.368 Sound
DJ 0.517 Poor 0.508 Acceptable 0.565 Acceptable
SB 0.519 Poor 0.457 Acceptable 0.570 Acceptable
HU 0.365 Acceptable 0.236 Sound 0.153 Sound
SA 0.596 Poor 0.466 Poor 0.364 Sound
SS 0.398 Sound 0.337 Acceptable 0.365 Sound
SB1 0.562 Poor 0.446 Acceptable 0.342 Sound
Entire study watershed 0.442 Acceptable 0.427 Acceptable 0.384 Sound
AYC** 0.491 Acceptable 0.490 Acceptable 0.476 Acceptable
* Average of balancing factors 1 and 2.
** It was obtained from Lee et al. (2006).
K.S. Lee, E.-S. Chung / Journal of Hydro-environment Research 1 (2007) 95e109 105

Table 5
Instreamflow requirements and maximum total daily loads (unit: cms, kg/day)
Name of sub-watershed Objective Low flow (Nov.wMar.) Representative fish Fish flow
Apr. May Jun. Jul. Aug. Sep. Oct.
WG Quantity 0.074 Rhynchocypris Oxycephalus 0.6 0.6 0.9 0.9 0.9 0.9 0.9
Quality 19.2 155.5 155.5 233.3 233.3 233.3 233.3 233.3
OJ Quantity 0.079 Rhynchocypris Oxycephalus 0.5 0.5 0.9 0.9 0.9 1.2 1.2
Quality 20.5 129.6 129.6 233.3 233.3 233.3 311.0 311.0
DJ Quantity 0.090 Rhynchocypris Oxycephalus 0.5 0.5 0.9 0.9 0.9 1.2 1.2
Quality 23.3 129.6 129.6 233.3 233.3 233.3 311.0 311.0
SB Quantity 0.128 Rhynchocypris Oxycephalus 1.6 1.6 2.2 2.2 2.2 2.2 2.2
Quality 33.2 414.7 414.7 570.2 570.2 570.2 570.2 570.2
HU Quantity 0.288 Zacco Platypus 1.2 1.2 1.8 1.8 1.8 1.8 1.8
Quality 74.6 311.0 311.0 466.6 466.6 466.6 466.6 466.6
SS Quantity 0.147 Rhynchocypris Oxycephalus 1.6 1.6 2.2 2.2 2.2 2.2 2.2
Quality 38.1 414.7 414.7 570.2 570.2 570.2 570.2 570.2
SA Quantity 0.112 Rhynchocypris Oxycephalus 1.6 1.6 2.2 2.2 2.2 2.2 2.2
Quality 29.0 414.7 414.7 570.2 570.2 570.2 570.2 570.2
SB1 Quantity 0.082 Rhynchocypris Oxycephalus 0.6 0.6 0.9 0.9 0.9 0.9 0.9
Quality 155.5 155.5 155.5 233.3 233.3 233.3 233.3 233.3
Entire Quantity 0.512 Carassius Auratus 3.0 0.6 0.6 1.2 1.2 3.0 3.0
Quality 132.7 777.6 155.5 155.5 311.0 311.0 777.6 777.6

such as those by Korea Institute of Construction Technology The BOD standard in this study, which is under 3 mg/L, was pro-
(1995) and Kim (1999), to derive the habitat suitability criteria posed to all local governments. Using the monthly instreamflow
of the target species. Table 6 reports the habitat suitability cri- requirement, total maximum daily loads can be determined.
teria used in this study.
By increasing discharge gradually, PHABSIM searches for 4.4. Developing list of management options (STEP 4)
the discharge value that maximizes the weighted usable area
and hence determines the optimal fish flow value. Table 5 pres- Once the problems and specific goals have been identified,
ents the fish flow calculated with PHABSIM for the 8 study managers should create a long list of all possible alternatives.
sub-watersheds. Note that the fish flows are available only The challenge for managers is to overcome preconceptions
for the spawning season from April to October. about workable options and create a broad and imaginative
range of solutions for further investigation. Broad creativity
4.3.1.3. Instreamflow requirement. The calculated average is necessary in this step.
low flow and the fish flow for each study sub-watershed are Management options may include measures that use tech-
compared and their maximum values are determined as the in- nology or structures to change existing conditions and those
streamflow requirements. As shown in Table 5, the fish flows that rely on changes in human behavior or management prac-
for the spawning season are always greater than the average tices (nonstructural). In virtually every planning scenario, one
low flow in all the study sub-watersheds, therefore the fish simple management strategy is to keep doing what is currently
flow from April to October and the average low flow from being done; in other words, to maintain the status quo. Various
November to March were selected as the instreamflow management options were developed that appeared appropri-
requirements. ate for the study watershed as follows: do nothing, regulation
of groundwater withdrawal (nonstructural), restoration of cov-
4.3.2. Target water quality ered stream, retention pond, reservoir redevelopment, use of
Some local governments within the study watershed therefore groundwater collected by subway stations, infiltration facility,
set their own higher standards and are trying to achieve them. interbasin transfer: reuse of wastewater treatment plant

Table 6
Habitat suitability criteria of target species
Species Water depth (cm) Velocity (cm/sec)
Spawning Fry Adult Spawning Fry Adult
Rhynchocypris oxycephalus 10w20 (Apr., May) 20w30 (summerwautumn) 30w50 (springwautumn) 10w30 20w40 30w120
Zacco platypus 10w30 (Apr., May) 10w40 (summerwautumn) 10w70 (springwautumn) 10w30 10w40 20w60
Carassius auratus 20w50 (May, Jun.) 10w40 (summerwautumn) 30w200 (springwautumn) 5w10 10w20 20w30
106 K.S. Lee, E.-S. Chung / Journal of Hydro-environment Research 1 (2007) 95e109

effluent, separated sewers, local wastewater treatment plant From the screening approach (according to the first goal),
(structural). To constrain these options it was established eight alternatives are eliminated: N1, N2, N4, N6 (technical
that at a minimum, no option should increase the possibility infeasibility), I1, I2, I3, I4 (economic infeasibility). Since de-
of flood damage. All feasible alternatives and their specific de- tailed assessment using a computer simulation model is per-
scriptions are shown in Table 7. formed in the next step, the process for the second goal was
not done in this step. As a result, a feasible master plan was
derived (Fig. 7).
4.5. Eliminating infeasible options (STEP 5)

Several approaches for the elimination of infeasible options 4.6. Testing the effectiveness of remaining feasible
are possible. A preliminary screening can inexpensively pro- options (STEP 6)
vide rapid insight into the probable effectiveness of different
management strategies. It may not, however provide the Since detailed analyses to test the effectiveness of options
kind of detailed, quantitative results necessary to justify fiscal consist of quantitative simulations, a hydrological simulation
commitment or definitely sway public or political opinion. model was used. Once an appropriate model was selected,
Any screening program should have the following goals: calibrated, and validated, it was available for evaluating alter-
native management approaches in the watershed. Scenario-
1. To determine which of the available alternatives is feasible testing is usually the foundation for plan development and
e that is, meets the constraints imposed by technical, eco- resource expenditure. Generally speaking, individual manage-
nomic, and environmental aspects (Walesh, 1989). ment options are tested first, and their performance evaluated
2. To determine which of the remaining alternatives performs against the decision criteria. Those that do not perform at an
best in terms of specified evaluation criteria. adequate level are rejected. When all options have been tested

Table 7
Specific descriptions of feasible alternatives
Alternatives Sub-watershed Description Name
Reservoir redevelopment OJ - Construction of sluice gate R1
- Proper operation (release discharge: 0.01 CMS from
Oct. to May)
HU - Proper operation (release discharge: 0.1 CMS from R2
Oct. to May)
SS - Proper operation (release discharge: 0.01 CMS from R3
Oct. to May)
New retention pond DJ - Capacity: 60,000 m3 N1
SB - Release discharge: 0.01 CMS from Oct. to May N2
WG N3
SS N4
SA N5
SB1 N6
Restoration of covered stream DJ - To remove roads and restore the stream S1
- Covered length: 1.59 km
- Construction of sewers
SB - To remove roads and restore the stream S2
- Covered length: 2.74 km
- Construction of sewers
SA - To remove roads and restore the stream S3
- Covered length: 0.645 km
- Construction of sewers
Interbasin transfer (reuse of WG - To transfer highly-treated wastewater of WWTP I1
WWTP effluent) OJ - Maximum quantity is 21,000 m3/day but used quan- I2
DJ tity is dependent upon the actual operation result of I3
SB WWTP I4
HU I5
SS(1) I6
SS(2) I7
SA I8
SB1 I9
Use of groundwater collected HU - To transfer groundwater collected by subway station U1
by subway stations into the stream
- Average quantity: 3720 m3/day
K.S. Lee, E.-S. Chung / Journal of Hydro-environment Research 1 (2007) 95e109 107

BOD concentration and total daily loads. Though these criteria


can reflect the hydrologic cycle during the dry period, future
research supporting these values is necessary.

4.7. Developing the final options (STEP 7)

In many cases, budget and resources are generally limited


and thus all feasible alternatives are seldom accepted simulta-
neously. Managers should therefore find a set of alternatives
that maximizes the desired objective (i.e. maintenance of
the minimum instreamflow requirement and water quality
enhancement).
However, especially when the constraints are uncertain,
ranking feasible alternatives could be preferred to finding an
optimal solution, then decision makers can execute an IWM
project according to the rankings whenever budget and re-
sources are available. The evaluation index (EI) equation,
used to quantify the effect of alternatives, has the following
form:
Fig. 7. A master plan from the pre-feasibility analysis.
f ðai Þ ¼ w1 f1 ðai Þ þ w2 f2 ðai Þ ð7Þ

where ai is the ith alternative, f1 , and f2 are the effects to pre-


vent the streamflow depletion and water quality deterioration
in this way, individual options are combined into management respectively, and w (sum of w is one) are weighting factors
strategies. That is, one strategy may, for example, encompass (w1 ¼ w2 ¼ 0:5 in this study).
urban management options, while another group includes ag- The evaluation equation can consider the sustainability,
ricultural management options. Mixed strategies may also be PSR concept as follows.
developed.
The HSPF model was selected from many hydrological fj ðai Þ ¼ aPRj;i þ bSTj;i þ gREj;i ; j ¼ 1; 2 ð8Þ
simulation models available for use in this study. It has been
both widely and exactly applied to the watershed as well as where j is the effectiveness, PR, ST, and RE mean pressure,
having undergone calibration and validation (Lee, 2007). All state, and response respectively, and a, b, and g are the
alternatives were tested by HSPF; results are shown in Table weighting factors (a þ b þ g ¼ 1). It is a role of manager to
8. The evaluation criteria for water quantity and quality select the indicators of pressure, state and response.
were set at drought and low flow values, which are 355th In this study, f1 and f2 are calculated as two following
and 275th daily flow in the flow-duration curve, and average equations.

Table 8
Effectiveness of alternatives using HSPF
Name of alternative Quantity Quality
Drought flow (Q355) (cms) Number of days to satisfy Average BOD BOD total
instreamflow requirement (days) concentration (mg/L) daily loads (kg/day)
Before After Before After Before After Before After
R1 0.003 0.047 17 18 15.2 2.6 59.3 60.9
R2 0.06 0.157 17 35 1.1 0.9 86.8 87.1
R3 0.006 0.006 53 56 1.1 0.9 86.8 87.1
N3 0.001 0.011 8 18 1.1 0.9 86.8 87.1
N5 0.000 0.006 0 25 1.1 0.9 86.8 87.1
S1 0.000 0.000 0 18 30 3.9 300 50.8
S2 0.000 0.001 0 55 30 3.9 300 50.8
S3 0.000 0.002 0 48 20 1.2 300 54.9
I1 0.06 0.139 150 203 1.1 3.3 86.8 143.8
I2 0.006 0.068 53 72 1.1 3.3 86.8 143.8
I3 0.003 0.107 46 203 1.4 4.1 45.3 135.9
I4 0.004 0.119 48 209 1.2 3.6 54.9 143.5
I5 0.001 0.064 48 105 15.3 4.6 44.4 100.5
U1 0.06 0.123 150 175 2.2 1.0 290.9 314.2
108 K.S. Lee, E.-S. Chung / Journal of Hydro-environment Research 1 (2007) 95e109


f1 ðai Þ ¼ a1 POðai Þ þ b1 PSDðai Þ þ g1 f0:5Dðai Þ þ 0:5Hðai Þg Dlðai Þ Dlðai Þ
Lðai Þ ¼ max ð14Þ
ð9Þ t3 ðai Þ i t3 ðai Þ

where POðai Þ is the population of a sub-watershed where the


alternative ai is applied, weighting values (a, b and g) are where Dlðai Þ is the decreased total daily loads of BOD and
shown in Table 3, and Dðai Þ and Hðai Þ are efficiencies against t3 ðai Þ is the target total maximum daily load of the sub-water-
the drought flow and satisfying days per year for instreamflow shed which alternative ai will be planned.
requirement, respectively. dðai Þ and hðai Þ can be obtained as This equation is based on the concept of pressure-state-re-
following Eqs. (10) and (11). sponse, but various formats are now being tested in ongoing
research to find the most appropriate index. Currently, the

Ddðai Þ Ddðai Þ weighting factors a, b, and g of pressure, state, and response
Dðai Þ ¼ max ð10Þ are estimated as the equi-weighted value, 1/3, which is also
t1 ðai Þ i t1 ðai Þ
a topic of a future research.
where Ddðai Þ is the increased drought flow due to alternative Results are shown in Table 9. According to EI values, all
ai and t1 ðai Þ is the target low flow of the sub-watershed which alternatives can be divided into three groups, as in STEP 2:
alternative ai will be planned. Poor (‘P’, 0e0.3), Acceptable (‘A’, 0.3e0.6), and Good
(‘G’, 0.6e1). Classification of all alternatives is as follows:
Dn1 ðai Þ G (S1, S2), A (R1, R2, I4, I5, U1), P (R3, N3, N5, I1, I2,
Hðai Þ ¼ ð11Þ I3, U1). Table 9 shows EI constituents (water quantity and
max Dn1 ðai Þ
i
quality). Since EI results change with manager’s preference
(which are weighting values (w)), other decision-makers in-
where Dn1 ðai Þ is the increased number of satisfying days for
cluding governmental organizations (GOs) and non-govern-
instreamflow requirement due to alternative ai .
mental organizations (NGOs) can get involved in this step.
f2 ðai Þ ¼ a2 POðai Þ þ b2 PWQDðai Þ þ g3 f0:5Qðai Þ þ 0:5Lðai Þg
ð12Þ 5. Conclusions

where weighting values (a, b, and g) are shown in Table 3 and This study provides a systematic seven step procedure and
Qðai Þ and Lðai Þ are effectiveness against average BOD con- application of integrated watershed planning and management
centration and average BOD total daily loads, respectively. that allows sustainable development in a Korean watershed.
Qðai Þ and Lðai Þ can be obtained as following Eqs. (13) and This procedure contains various techniques as follows:
(14). MCDM methods (CP and AHP) in STEP 2 and 7, continuous
 simulation model (HSPF) in STEP 1 and 6, sustainability con-
Dqðai Þ Dqðai Þ cept (PSR model) in STEP 2 and 7, pollutant load estimation
Qðai Þ ¼ max ð13Þ
t2 ðai Þ i t2 ðai Þ based on unit load (PLOAD) in STEP 1 and 2, and instream-
flow requirement and TMDL in STEP 3, screening approach in
where Dqðai Þ is the decreased average BOD concentration, STEP 5. Additionally, hydrologic hazard ranking procedure
t2 ðai Þ is the target quality of the sub-watershed which alterna- and evaluation index to quantify effectiveness of alternative
tive ai will be planned. are proposed. This study serves as a guide to construct

Table 9
Calculation results of evaluation index
Name of alternative Quantity Quality Total
Pressure State Response EI Pressure State Response EI EI Rank Grade
Weights 0.371 0.375 0.254 1 0.302 0.388 0.31 1
R1 0.315 0.875 0.250 0.509 0.315 0.516 0.397 0.418 0.463 6 A
R2 0.315 0.875 0.605 0.599 0.315 0.516 0.183 0.352 0.475 5 A
R3 0.117 0.371 0.010 0.185 0.117 0.508 0.183 0.289 0.237 12 P
N3 0.012 0.934 0.094 0.378 0.012 0.460 0.183 0.239 0.309 9 A
N5 0.315 0.846 0.108 0.461 0.315 0.506 0.183 0.348 0.405 8 A
S1 1.012 1.000 0.061 0.766 1.012 0.988 1.000 0.999 0.882 1 G
S2 0.792 0.813 0.190 0.647 0.792 1.000 1.000 0.937 0.792 2 G
S3 0.315 0.846 0.170 0.477 0.315 0.506 0.867 0.560 0.519 3 A
I1 0.372 0.000 0.301 0.215 0.372 0.000 0.058 0.130 0.173 14 P
I2 0.117 0.371 0.251 0.246 0.117 0.508 0.058 0.251 0.249 11 P
I3 0.000 0.371 0.845 0.354 0.000 0.508 0.000 0.197 0.276 10 P
I4 0.315 0.846 1.000 0.688 0.315 0.506 0.008 0.294 0.491 4 A
I5 0.241 0.772 0.534 0.514 0.241 0.453 0.283 0.337 0.425 7 A
U1 0.372 0.000 0.182 0.184 0.372 0.000 0.167 0.164 0.174 13 P
K.S. Lee, E.-S. Chung / Journal of Hydro-environment Research 1 (2007) 95e109 109

consistent decision support system for integrated watershed Heathcote, I.W., 1998. Integrated Watershed Management. John Wiley &
management. Sons.
Kim, J.H., 2004. Estimation of the Basin Flood Control Safety. Master’s thesis,
Future studies are needed. The ability to lessen flood dam- Seoul National University, Seoul, Korea (in Korean).
age should be considered since N3 and N5 have proven even Kim, K.H., 1999. Evaluation of Habitat Conditions and Estimation of Opti-
more effective. Additional creative feasible options should mum Flow for the Freshwater Fish. Ph.D. thesis, Yonsei University, Seoul,
be considered. Finally weighting values should be carefully Korea (in Korean).
assessed since they have a great influence on the vulnerability Korea Institute of Construction Technology, 1995. Development and Applica-
tion of Determination Method of Instreamflow Requirement. Korea Water
of sub-watersheds and evaluation index. Resources Corporation.
Lee, J.S., 2007. Effect Analysis of Measures to Prevent Streamflow Depletion
Acknowledgements in the Anyangcheon Watershed Using HSPF. Master’s thesis, Seoul
National University, Seoul, Korea (in Korean).
This research was supported by a grant (1-7-3, 80%) from Lee, K.S., Chung, E.S., Kim, Y.O., 2006. Integrated approach for watershed
management in an urban area. Journal of Korea Water Resources Associ-
Sustainable Water Resources Research Center of 21st Century ation 39 (2), 151e167 (in Korean).
Frontier Research Program of Ministry of Science and Tech- Lee, Y.W., Bogardi, I., Stansbury, J., 1991. Fuzzy decision making in dredged-
nology through Engineering Research Institute of Seoul material management. Journal of Environmental Engineering 122 (4),
National University and BK 21 Safe and sustainable Infra- 248e258.
structure Research (20%). Lee, Y.W., Dahab, M.F., Borgadi, I., 1992. Nitrate risk management under un-
certainty. Journal of Water Resources Planning and Management 118 (2),
151e165.
References OECD, 1993. OECD core set of indicators for environmental perfor-
mance reviews. In: OECD Environment Monographs, vol. 83.
Bardossy, A., Bogardi, I., 1983. Network design for the spatial estimation of en- OECD, Paris.
vironmental variables. Applied Mathematics and Computation 12, 339e369. Riggs, H.C., 1972. Low Flow Investigation. Book 4. In: U.S. Geological
Bicknell, B.R., Imhoff, J.C., Kittle Jr., J.L., Jobes, T.H., Donigian Jr., A.S., Survey Surface Water Techniques Series. U.S. Geological Survey,
2001. Hydrologic Simulation Program e FORTRAN (HSPF) User’s Washington, DC.
Manual for Version 12. U.S. Environmental Protection Agency, National Satty, T.L., 1977. A scaling method for priorities in hierarchical structures.
Exposure Research Laboratory, Athens, GA. Journal of Mathematical Psychology 15, 234e281.
Chung, E.S., Lee, K.S., Shin, M.J., 2006. Characteristics of water quantity and Shim, M.P., 2003. Technology of Sustainable Surface water Development.
quality of the anyangcheon using SWAT model and calculation result of The 21st Century Frontier R&D Program. Sustainable Water Re-
EMC. Journal of Korean Society on Water Quality 22 (4), 648e657. sources Research Center, Inha University, Gyeonggi-do, Korea (in
Connell, E.O., Bathurst, J., Kilsby, C., Parkin, G., Quinn, P., Younger, P., Korean).
Anderton, S., Riely, M., 2000. Integrating mesoscale catchment experi- Smakhtin, V.U., 2001. Low flow hydrology: a review. Journal of Hydrology
ences with modeling: the potential for sustainable water resources 240, 147e186.
management. In: Fifth IHP/IAHS George Kovacs Colloquium, HELP, Stedinger, J.R., Thomas Jr., W.O., 1985. Low flow frequency estimation
International Hydrological Program, UNESCO, Paris. using base flow measurement. U.S. Geological Survey Open File
Edwards, C., Miller, M., 2001. PLOAD Version 3.0 User’s Manual. USEPA. Report, 85e95.
Goicoechea, A., Hansen, D.R., Duckstein, L., 1982. Multiobjective Decision US Geological Survey, 2001. PHABSIM for Windows, User Manual and Exer-
Analysis with Engineering and Business Applications. Wiley, New York. cises. Midcontinent Ecological Science Center, Fort Collins, Colorado, USA.
Gyeonggi Research Institute, 2003. A Counterplan of Revival for the Anyang- Vogel, R.M., Kroll, C.N., 1992. Regional geohydrologicegeomorphic
cheon Watershed. Gyeonggi-do (in Korean). relationships for the stimation of low-flow statistics. Water Resources
Hagemeister, M.E., Jones, D.D., Woldt, W.E., 1995. Hazard ranking of land- Research 28 (9), 2451e2458.
fills using fuzzy composite programming. Journal of Environmental Engi- Walesh, S.G., 1989. Urban Surface Water Management. John Wiley & Sons,
neering 122 (4), 248e258. New York.
Hartmann, L., Torno, C., Bogardi, I., Higler, L., 1987. Methodological Guide- Yurdusev, M.A., O’Connell, P.E., 2005. Environmentally-sensitive water re-
lines for the Integrated Environmental Evaluation of Water Resources sources planning: 1. Methodology. Water Resources Management 19,
Development. UNESCO, Paris. 375e397.

Das könnte Ihnen auch gefallen