Adapting SWAT for the Modeling of Pesticide Transport for ...

Adapting SWAT for the Modeling of Pesticide

Transport for a Tile-drained River Basin

Presented by Ting Tang

19-Jul-13 @Toulouse, France – 2013 SWAT conference

by

Ting Tang

Ann van Griensven

Linh Hoang

Surface runoff

Drainage flow (if present) 2013 International SWAT conference 2

Background

Surface runoff Soil profile

Shallow aquifer

Leaching

Pesticide application

Lateral flow

GW discharge

Wind drift

Volatilization Transport

Point source

Deposition

Drainage flow

Source: Adapted from Neitsch et al. (2005)

Agricultural

pesticide loss in

surface water:

<1% (Carter 2000)

Most important pathways

Realistic quantification of pesticide fluxes in

subsurface drainage flow at the basin scale

remains a challenge.

SWAT is believed to be a promising tool for

pesticide transport and flux modeling.

SWAT does not simulate pesticide transport

in tile flow.

2013 International SWAT conference 3

Background

Adaptations to SWAT model

Ad.1 – For detailed mass balance analysis

I. Print out intermediate outputs

II. Outputs treated with MATLAB to generate MB


Methodology: SWAT adaptations

Outputs Source file

modified Outputs

Source file

modified

Total application apply.f Storage in soil pesty.f

Effective application apply.f Leaching pestlch.f

Decay on foliage decay.f Channel load (SR_soluble) sumv.f

Decay in soil decay.f Channel load (SR_sorbed) sumv.f

Storage on plant decay.f Channel load (lateral flow) sumv.f

Ad.2 – To incorporate pesticide transport in tiles

I. Introduce a new variable 'fqtile (ly, j)‘ in percmain.f

II. Compute pesticide loss through tile flow in pestlch.f

with

Where: psttile,ly: amount of pesticide removed in tile flow from the layer

tile located (ly) in kg pst/ha;

concpst,flow: concentration of pesticide in the mobile water for the given

layer, in kg pst/mm H2O;

Qtile,ly: equals to 'fqtile (ly, j)', amount of tile flow loss in given soil layer in

for HRU ‘j’, in mm H2O.


Methodology: SWAT adaptations

The Odense River Basin

Island of Funen, Denmark

Area: 612 km2

Agriculture dominated

Heavily tile-drained

Calibrated SWAT hydrological model

• 30 subbasins • 808 HRUs


Methodology: case study

Source: Hoang et al. (2012)


Selected pesticides: Bentazone, MCPA, Pendimethalin

Parameters unit Bent. MCPA Pend. IPNUM - 31 234 167

SKOC (mg/kg)/(mg/L) 34 35 5000

WOF - 0.6 0.6 0.4 HLIFE_F days 2 8 30 HLIFE_S days 20 25 90 AP_EF - 0.75 0.75 0.75 WSOL mg/L 2300000 825 0.275

Ann. Aver. App. rate kg/ha 0.106 0.511 0.067

Mobile Non-persistent

Immobile Persistent

Methodology : Model setup

Simulation duration:

8 years (1994-2001) Warming-up: 1994


Inc. of tile pst. transport Simulation code No RunOri.L Yes RunTile.L

Methodology : Model setup

Not fully closed, but acceptable in general;

Better for Bent./MCPA;

Influence on pesticide load prediction.


Results: Mass balance

Mass mismatch in % of the effective application (‘+’: loss; ‘-’: creation)

Total load to streams: < 0.25%!!

Pesticide Bentazone MCPA Pendimethalin

RunOri.L -0.002% 0.000% -0.388%

RunTile.L -0.002% 0.000% -0.388%

0.0E+00

5.0E-06

1.0E-05

1.5E-05

2.0E-05

2.5E-05

3.0E-05

01-01-95 15-05-96 27-09-97 09-02-99 23-06-00 05-11-01

RunTile.L

RunOri.L

Modified SWAT2005 enables flux prediction during

the wet season (winter). More realistic


Results: Model improvement

Total MCPA load into streams, kg/ha

Not predicted by the original model


Results: Pathways

Average annual total channel load by pathway, kg/yr

0

2

4

6

8

10

12

14

16

RunOri.L RunTile.L RunOri.L RunTile.L RunOri.L RunTile.L

Bentazone MCPA Pendimethalin

Surface runoff (soluble)

Surface runoff (sorbed)

Lateral flow

Tile flow

01-02-96 21-02-96 12-03-96

0.0E+00

1.0E-04

2.0E-04

3.0E-04

4.0E-04

5.0E-04

6.0E-04

01-01-95 15-05-96 27-09-97 09-02-99 23-06-00 05-11-01

RunTile.L

RunOri.L

MCPA Maxobs=1.6E-03 mg/L


Results: Instream concentration

Pesticide conc. at the watershed outlet, mg/L

Maxobs: Observed maximum conc. in

headwaters (Kronvang et al. 2003; Styczen et al. 2004)

The pattern for bentazone is similar

to MCPA

0.0E+00

5.0E-04

1.0E-03

1.5E-03

2.0E-03

01-01-95 15-05-96 27-09-97 09-02-99 23-06-00 05-11-01

Pendimethalin Maxobs=1.7E-04~5.2E-04 mg/L

Modified SWAT2005: more realistic for Odense,

maybe used for modeling tile-drained basins with

additional validation;

Bentazone/MCPA more reliably simulated than

pendimethalin (mass balance, load, instream

concentration);

Preferred pathways not identical among pesticides.


To conclude…

Carter (2000), Herbicide movement in soils: principles, pathways and processes.

Weed Research 40: 113-122 DOI 10.1046/j.1365-3180.2000.00157.x

Neitsch, S. L., J. Arnold, J. R. Kiniry and J. R. Williams (2005), Soil and Water

Assessment Tool Theoretical Documentation, Version 2005 USDA Agricultural

Research Service & Blackland Research Centre, Texas A&M University, USA.

Kronvang, B., H. L. Iversen, K. Vejrup, B. B. Mogensen, A. M. Hansen and L. B.

Hansen (2003), Pesticides in streams and subsurface drainage water within two

arable catchments in Denmark: Pesticide application, concentration, transport

and fate. Ministry of Environment, Danish Environmental Protection Agency.

Styczen, M., S. Petersen, M. Jessen, O. Z. Rasmussen, D. Andersen, M. Buck and

P. B. Sørensen (2004), Calibration of Models Describing Pesticide Fate and

Transport in Lillebæk and Odder Bæk CatchmentPesticides Research, pp. 218.


References

Supplementary info.


Period Criteria/

Station

Daily Monthly

45_26 45_21 45_01 45_26 45_21 45_01

Calibration NSEQ 0.81 0.76 0.79 0.85 0.80 0.84

RQ 0.92 0.91 0.90 0.96 0.95 0.95

Validation NSEQ 0.79 0.76 0.80 0.84 0.81 0.85

RQ 0.90 0.90 0.91 0.93 0.93 0.93

From Hoang et al, 2012

Supplementary info.


0

10

20

30

40

50

01/01/1999 11/04/1999 20/07/1999 28/10/1999 05/02/2000 15/05/2000 23/08/2000 01/12/2000

Date

Dis

ch

arg

e (

m3/s

)

SWAT

measured

From Hoang et al, 2012

Adapting SWAT for the Modeling of Pesticide Transport for ...

Documents

Transcript of Adapting SWAT for the Modeling of Pesticide Transport for ...