Integrated Modeling of Surface Water and …...1 Integrated Modeling of Surface Water and Groundwate r by Using Combined SWAT-MODFLOW N. W. Kim, I. M. Chung, Y. S. Won**, J. Lee*,

1

Integrated Modeling of Surface Water and Groundwater by Using Combined SWAT-MODFLOW

N. W. Kim*, I. M. Chung*, Y. S. Won**, J. Lee*, J. G. Arnold***

* Korea Institute of Construction Technology**River Information Center of Han River Flood Control Office

*** USDA-Agricultural Research Service

SWAT-K

2

� Background and Purpose

� Overview of SWAT-K and

SWAT-MODFLOW

� Application to test watersheds

� Results of simulation

� Conclusion

Content

3

BACKGROUNDBACKGROUND

� Current issues on water management in Korea :

� Water rights (and regulations) on groundwater use near streams

� Accurate estimation of groundwater recharge / discharge

� Potential amount of groundwater development based on groundwater recharge

� Groundwater is an alternative water resources in Korea

4

BACKGROUNDBACKGROUND

� Problems

� Hydrologic component analysis in Korea has concentrated on surface water management, so problems related to groundwater were not dealt with rigorously

� In groundwater modelling, groundwater recharge could not be considered in terms of hydrological processes

� The role of groundwater model in surface water management was less significant

5

PURPOSEPURPOSE

� Construction of a long term rainfall runoff model, producing integrated analysis for both the groundwater and surface water

� To explain the effects of spatial-temporal distributions of recharge and groundwater head

� To examine the hydrological effects allowing hydraulic interaction between surface water and groundwater

� Application to Underground dam / River intake / Reservoir capacity/ Potential amount of groundwater development / Stream flow decrease in urbanized watershed

6

SWATSWAT--KK

� What is SWAT-K(Korea)?

SWAT+Variation of water cycle structure (Natural+Artificial)SWAT+Variation of water cycle structure (Natural+Artificial)

SWATSWATSWATSWAT----AGRIMANAGRIMANAGRIMANAGRIMANSWATSWATSWATSWAT----MODFLOWMODFLOWMODFLOWMODFLOW

SWATSWATSWATSWAT----SWMMSWMMSWMMSWMMSWATSWATSWATSWAT----ROMROMROMROM

7

▶▶▶▶ Technological Roadmap

SWATSWAT--KK

8

SWATSWAT--MODFLOWMODFLOW

� Fully combined SWAT-MODFLOW

� Development of tool for exchanging characteristics between HRU(Hydrologic Response Unit) of SWAT and CELL of MODFLOW

� Groundwater module in SWAT is replaced by MODFLOW

� Spatially distributed GW recharge rates and heads can be simulated by SWAT-MODFLOW

� Stream-aquifer interactions by RIVER package in MODFLOW

9


* * * * This approach is quite different to the method used by Sophocleous et al.(1997) which involves separate modelling by sequential linkages.

� Schematic diagram

10


■ Fully coupled work by Serial linkage

simulate

Std3..

command-1

subbasin-1

Varinit..

gwmod-1

command-2..

tilmix

subbasin-2..

saveconc

gwmod-2..

virtual

MODFLOW

SWAT

SWAT

- - - - -

- - - - - - -

- Groundwater recharge- River stage- Water transfer- Drainage- Aquifer EVT

- Cell based recharge- Aquifer-river exchange rate- Variation of GW head- Drainage- Aquifer EVT

11


� Major input data for SWAT-MODFLOW

� HRU distribution which links HRU with MODFLOW grids

� Cell based recharge rate which is made by HRU based in SWAT

� Stream data(stage, width) is common to both model

� Data for water transfer of shallow aquifer in SWAT is commonly used in WELL package in MODFLOW

12

SWATSWAT--MODFLOWMODFLOW� Linking RECHARGE in MODFLOW with SWAT

7 7 7 7

7 7 7 7

7 7 7 7

7 5 5

5 55 556 6 6

6 6 6

6 6

64

4 4 4 4

4 4

4 4

4 4

44 913554

SWAT

MODFLOW

HRU = 54+44+35+91

HRU

HRU

MODFLOW cell

HRU based recharge

HRU in small watershed

HRU in small watershed

Cell in MODFLOW

13


Stream inflow in SWAT =discharge in small watershed +SW & GW exchange(23)

Stream inflow in SWAT

Small watershed

Cell in MODFLOW

Stream in MODFLOW

� Linking RIVER in MODFLOW with SWAT

14

Runoff ProcessRunoff Process

- Processes -

1. Qsur, Qinf

2. Qlat, Qperc

3. ET

4. Qrchge

5. Qgw

6. Qrevap

7. Qda

• Behavior of hydrological components in single HRU

15

� Enhancement of recharge parameter estimation for vadosezone modeling

: recharge from soil zone to aquifer (mm)

1,, ]/1exp[)]/1exp[1( −−+−−= irchrggwseepgwirchrg www δδ

irchrgw ,

gwδ

seepw

: delay parameter (day)

: amount of water exiting the bottom of soil profile (mm)

Single linear reservoir

Multiple linear reservoir

OIdt

dS −=

OKS =

Continuity Eq.

Storage Eq.

n system of Eq.

16


� Enhancement of water transfer in SWAT-MODFLOW

� In SWAT’s water transfer option, when the source type is aquifer, then this case is related with Well package in MODFLOW

� Water transfer rate in HRU is correspondent to pumping rate Q (discharge or recharge)

17

GYUNGANCHEON WATERSHEDGYUNGANCHEON WATERSHED

A=260 km2

18

APPLICATIONAPPLICATION

� The characteristics of subbasins

19


� Input data for SWAT

�AVS2000 (DiLuzio et al., 2001) was used to automate the construction of model input parameters�Daily precipitation of Suwon gauging station (MOCT)�Daily values of temperatures, solar radiation, wind speed etc.(KMA)�Land use digital data(1:25,000) (MOCT)�The detailed soil association map(1:25,000) was used for selection of soil attributes. 38 hydrologic soil groups within basin were used. �Related soil physical properties (texture, bulk density, available water capacity, saturated hydraulic conductivity, soil albedo and some additional factors) were obtained from the Agricultural Soil Information System (http://asis.rda.go.kr)

20


N

EW

S

DEMDEMDEMDEM landuselanduselanduselanduse soilsoilsoilsoil

HRU distributionHRU distributionHRU distributionHRU distribution

The spatially distributed HRUs are used to match discretised basin with MODFLOW grids.

21


� Input data for MODFLOW

� GW information from GIMS( http://www.gims.go.kr) was used

- Hydraulic conductivity in alluvial aquifer : 2.1 ××××10-3cm/sec.

- Specific yield ranges from 0.1 to 0.3.

�The boundaries of basin were designated as no-flow cells.

� Recharge was distributed according to SWAT simulation outputs for

each day

� River stage of MODFLOW is imported from SWAT's daily

simulation outputs

22

RESULTSRESULTS

� Model calibration (R2=0.79)

Measured and simulated runoff for studied basin during 1990

23

RESULTSRESULTS

� Model validation (R2=0.66/ME=0.65)

Measured and simulated runoff for studied basin during 1991

24

RESULTSRESULTS

� Spatial-temporal variation of groundwater recharge

T=500days T=1000days

25

RESULTSRESULTS

� Spatial-temporal variation of groundwater heads

T=500days T=800days T=1000days

26

RESULTSRESULTS

� Groundwater level variation near stream

27

RESULTSRESULTS

� River aquifer interaction

28

RESULTSRESULTS

� Application of SWAT-MODFLOW pumping moduleA

A’

B B’

Q=1000m3/d

29

RESULTSRESULTS

� Application of SWAT-MODFLOW pumping module

Drawdown of groundwater level at the cross section of south to north direction

-2.5

-2

-1.5

-1

-0.5

0

55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80

Row

Dra

wdo

wn

(m )

pumping rate= 1000cmd


zone of influence=7.5km

30

RESULTSRESULTS

� Application of SWAT-MODFLOW pumping module

Drawdown of groundwater level at the cross section of east to west direction

-2.5

-2

-1.5

-1

-0.5

0

30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55

Column

Dra

wdo

wn

(m )



zone of influence=7.5km

31

MUSIM WATERSHEDMUSIM WATERSHED

• Watershed information

-Area : 198 km2

-No. of subbasins : 34-Simulation period: 2001~2005

DEMDEMDEMDEM

Naeduk

GadukGroundwater stations

Stream gauging station

32

• Spatial information

(((( ))))

Land use Soil association map

MUSIM WATERSHEDMUSIM WATERSHED

33

• Recharge according to time delay (Naeduk station)

RECHARGE SIMULATIONRECHARGE SIMULATION

1/1/01 1/1/02 1/1/03 1/1/04 12/31/04Date (month/day/year)

0.0

4.0

8.0

12.0

16.0

20.0

Rec

harg

e R

ate

(mm

/d)

1/1/01 1/1/02 1/1/03 1/1/04 12/31/04

150

100

50

0

Dai

ly R

ain

fall

(mm

/d)

Naeduk gauging stationdelta=1.04delta=4.48delta=19.49

34


1/1/01 1/1/02 1/1/03 1/1/04 12/31/04Date (month/day/year)

0.0

4.0

8.0

12.0

16.0

20.0

Rec

harg

e R

ate

(m

m/d

)

1/1/01 1/1/02 1/1/03 1/1/04 12/31/04

150

100

50

0

Daily

Rain

fall

(mm

/d)

Kaduk gauging stationdelta=1.04delta=4.48delta=19.49

• Recharge according to time delay(Gaduk obsevatory)

35


0.0 0.2 0.4 0.6 0.8 1.0EXP(-1/delta)

-1.0

-0.5

0.0

0.5

1.0C

OR

RE

LA

TIO

N

NaedukKaduk

• Correlation with GW level according to exp(-1/δδδδgw)

36

• Simulated recharge and measured groundwater level(Naeduk)


7/1/03 7/10/03 7/19/03 7/28/03 8/6/03 8/15/03Date (month/day/year)

0.0

2.0

4.0

6.0

8.0

10.0

Rech

arg

e R

ate

(m

m/d

ay)

34.0

34.2

34.4

34.6

34.8

35.0

Gro

undw

ate

r Leve

l (E

l.m)

Groundwater Level (Obs.)Recharge Rate (Sim.)

37

• Simulated recharge and measured groundwater level(Gaduk)


7/1/03 7/31/03 8/30/03 9/29/03 10/29/03 11/28/03 12/28/03Date (month/day/year)

0.0

0.5

1.0

1.5

2.0

2.5

Rech

arg

e R

ate

(m

m/d

ay)

74.0

74.5

75.0

75.5

76.0

76.5

77.0

Gro

undw

ate

r Le

vel (

El.m

)

Groundwater Level (Obs.)Recharge Rate (Sim.)

38

• Monthly recharge


2001 2002 2003 2004 2005Date (month-year)

0

50

100

150

200

Rec

harg

e R

ate

(m

m/m

)

2001 2002 2003 2004 2005

500400300200100

0

Mon

thly

Rai

nfal

l (m

m/m

)

Gauging StationNaedukKaduk

39

• Simulated stream flow (2001~2005)– Overall correlation (R2=0.75)

RUNOFF SIMULATIONRUNOFF SIMULATIONSt

ream

flow

Stre

am fl

owSt

ream

flow

Stre

am fl

ow

Obs.

Sim.

40

• Groundwater head simulation

GROUNDWATER SIMULATIONGROUNDWATER SIMULATION

� R2=0.95

ObservedObservedObservedObserved SimulatedSimulatedSimulatedSimulated

41

• Verification of groundwater level


33.033.033.033.0

33.533.533.533.5

34.034.034.034.0

34.534.534.534.5

35.035.035.035.0

35.535.535.535.5

36.036.036.036.0

36.536.536.536.5

37.037.037.037.0

2001/ 012001/ 012001/ 012001/ 01 2001/ 042001/ 042001/ 042001/ 04 2001/ 072001/ 072001/ 072001/ 07 2001/ 102001/ 102001/ 102001/ 10

Time (day)Time (day)Time (day)Time (day)

(EL

.m)

(EL

.m)

(EL

.m)

(EL

.m)

ObsSim.

Gro

undw

ater

leve

lG

roun

dwat

er le

vel

Gro

undw

ater

leve

lG

roun

dwat

er le

vel

42



33.033.033.033.0

33.533.533.533.5

34.034.034.034.0

34.534.534.534.5

35.035.035.035.0

35.535.535.535.5

36.036.036.036.0

36.536.536.536.5

37.037.037.037.0

2002/ 012002/ 012002/ 012002/ 01 2002/ 042002/ 042002/ 042002/ 04 2002/ 072002/ 072002/ 072002/ 07 2002/ 102002/ 102002/ 102002/ 10


(EL

.m)

(EL

.m)

(EL

.m)

(EL

.m)

Gro

undw

ater

leve

lG

roun

dwat

er le

vel

Gro

undw

ater

leve

lG

roun

dwat

er le

vel

ObsSim.

43



33.033.033.033.0

33.533.533.533.5

34.034.034.034.0

34.534.534.534.5

35.035.035.035.0

35.535.535.535.5

36.036.036.036.0

36.536.536.536.5

37.037.037.037.0

2003/ 012003/ 012003/ 012003/ 01 2003/ 042003/ 042003/ 042003/ 04 2003/ 072003/ 072003/ 072003/ 07 2003/ 102003/ 102003/ 102003/ 10


(EL

.m)

(EL

.m)

(EL

.m)

(EL

.m)

ObsSim.

44

RECHARGE DISTRIBUTIONRECHARGE DISTRIBUTION

� Distributed recharge (mm)

Recharge(AugRecharge(AugRecharge(AugRecharge(Aug. 2002). 2002). 2002). 2002) Recharge(SepRecharge(SepRecharge(SepRecharge(Sep. 2002). 2002). 2002). 2002)

45

RECHARGE DISTRIBUTIONRECHARGE DISTRIBUTION

� Distributed recharge (mm)

Recharge(OctRecharge(OctRecharge(OctRecharge(Oct. 2002). 2002). 2002). 2002) Recharge(NovRecharge(NovRecharge(NovRecharge(Nov. 2002). 2002). 2002). 2002)

46

� Distribution of wells (N=2,176)

EFFECTS OF WELL PUMPINGEFFECTS OF WELL PUMPING

47

� Distributed drawdown by pumping (Unit: m)


48

� Stream flow variation due to pumping

BeforeBeforeBeforeBefore

AfterAfterAfterAfter

Stre

am fl

owEFFECTS OF WELL PUMPINGEFFECTS OF WELL PUMPING

49

� Flow pattern variation by pumping

0.1

1

10

100

1000

0 20 40 60 80 100

Exceedance Probability (%)

Dis

char

ge (

cms)

BeforeBeforeBeforeBeforeAfterAfterAfterAfter


50

CONCLUSIONCONCLUSION

� Potential application of combined SWAT-MODFLOW model in simulating groundwater recharge and level is demonstrated

� The combined models are also capable of reflecting river-aquifer exchange rate properly

� The reliability of groundwater discharge and total runoff of watershed is enhanced

� Application to planning of groundwater development / underground dam operation / river intake etc.

� For larger watershed, application of SWAT-MODFLOW is limited due to the limitation of MODFLOW

51

11111111stststststststst SWATSWATSWATSWATSWATSWATSWATSWAT--------KOREA ConferenceKOREA ConferenceKOREA ConferenceKOREA ConferenceKOREA ConferenceKOREA ConferenceKOREA ConferenceKOREA Conference

7 Nov. 2006 / Venue : Main hall of KICT, Ilsan

Proceeding Program

52

1st SWAT1st SWAT1st SWAT1st SWAT1st SWAT1st SWAT1st SWAT1st SWAT--------KOREA ConferenceKOREA ConferenceKOREA ConferenceKOREA ConferenceKOREA ConferenceKOREA ConferenceKOREA ConferenceKOREA Conference

Staffs/Presenters

53

1st SWAT1st SWAT1st SWAT1st SWAT1st SWAT1st SWAT1st SWAT1st SWAT--------KOREA ConferenceKOREA ConferenceKOREA ConferenceKOREA ConferenceKOREA ConferenceKOREA ConferenceKOREA ConferenceKOREA Conference

Participants

54

INTERNATIONAL ACTIVITIESINTERNATIONAL ACTIVITIES

55

Thank you

The authors express thanks for a grant (code: 2-2-2) from the Sustainable Water Resources Research Center of

the 21st Century Frontier Research Program.

Documents

Integrated Modeling of Surface Water and …...1 Integrated Modeling of Surface Water and Groundwate r by Using Combined SWAT-MODFLOW N. W. Kim*, I. M. Chung*, Y. S. Won**, J. Lee*,

Integrated Modeling of Surface Water and …...1 Integrated Modeling of Surface Water and Groundwate r by Using Combined SWAT-MODFLOW N. W. Kim, I. M. Chung, Y. S. Won**, J. Lee*,