SWAT Model Overview and Application in

the Upper Mississippi

Jeff Arnold USDA-ARS - Temple, Texas

y Readily available input – Physically based

y Computer efficient y y

Model Philosophy

Comprehensive – Process Interactions Simulate Management

1

Stream Delineation

y Continuous Time Daily Time Step (Sub-hourly) One Day Hundreds of Years

y

y y

General Description

y Subwatersheds y Hydrologic Response Units y y

Possible Configurations

Distributed Parameter Unlimited Number of Subwatersheds

Comprehensive – Process Interactions Simulate Management

Output from other Models - APEX, SWAT Point Sources - Treatment Plants

2

3

Non-CultivatedLands

CEAP Watershed System

Channel/Flood PlainProcesses

Point Sources

APEX CultivatedFields

Root Zone

Shallow (unconfined)

Aquifer

Vadose(unsaturated)

Zone

Confining Layer

Deep (confined) Aquifer

Precipitation

Evaporation and Transpiration

Infiltration/plant uptake/ Soil moisture redistribution

Surface Runoff

Lateral Flow

Return FlowRevap from

shallow aquiferPercolation to

shallow aquifer

Recharge to deep aquifer

Flow out of watershed

Hydrologic Balance

(Zone

( ) ⎟⎟ ⎠

⎞ ⎜⎜ ⎝

⎛ ⎥ ⎦

⎤ ⎢ ⎣

⎡ − −⋅−=

drain tilewtwtr t

hhtile 24 exp1 lyly FCSW >if

⎟⎟ ⎠

⎞ ⎜⎜ ⎝

⎛ ⋅

⋅ ⋅ ⋅ ⋅ =

hilld

sat lat L

slpKSWQ

φ , 2

Tile Flow

Root Zone

Shallow (unconfined)

Aquifer

Vadose unsaturated)

Confining Layer

Deep (confined) Aquifer

excess ly 024.0

NO3 - NH4

+

Soil Organic Matter

NO2 -

clay

mineralization

nitrification

Symbiotic

NO3 -

anaerobic conditions

N2 N2O

NH3

)

fertilizer fertilizer

denitrification

n

runoff manures, wastes

and sludge

ammonium fixation

immobilization

immobilization

fixation

Atmospheric N fixation (lightning arc discharge

leaching

Harvest

Nitrogen Cycle

ammonia volatilizatio

4

Pesticide Dynamics

Runoff

Foliar Application

Degradation

Washoff

Infiltration

Leaching

Surface Application

Degradation

• • Removal of Biomass as Harvest/

• • • • Pesticide Applications • • Water Impoundment (Wetlands, Rice) • Subsurface (Tile) Drainage

Management Crop Rotations

Conversion of Biomass to Residue Tillage / Biomixing of Soil Fertilizer and Manure Applications Edge-of-Field Buffers

Irrigation

5

Topographic InputsSubbasin DelineationRouting StructureChannel Slopes & LengthsOverland Slopes & Lengths

Land Use Inputs• Type – Ag, Forests,

Range, Urban• Management – Rotations

Tillage, Irrigation, Fertilizer

Soil Physical PropertiesTextureBulk DensityWater CapacityConductivityOrganic Carbon

Channel Processes

• • •

•

• Pesticide Toxic balance

Channel Processes

Flood Routing Transmission Losses, Evaporation Sediment Routing

Degradation and deposition Nutrients

modified QUAL2E

Weather

Max/Min Temp

ArcView GIS Interface

• Daily Precip and

• Monthly Radiation, Windspeed, Humidity

6

Calibration and Uncertainty

•

• Auto Calibration Shuffled Complex Evolution Algorithm Multi-Objective Optimization

• Uncertainty Analysis

Uncertainty Analysis

• Parameter Uncertainty Separate SCE optimization into “Good” and “Not Good” Sets χ2 and Bayesian Methods

•Predictive Uncertainty

Assessment using Split SamplES (SUNGLASSES)

Sensitivity Analysis Latin Hypercube – One-factor-at-a-time

Sources of UNcertainty GLobal

7

Conservation Effects Assessment Project (CEAP)

• Current Analysis Grass Waterways, Contours, Terracing, Filter Strips, Buffers

• Future Analysis Tillage Systems, Manure and Fertilizer Management

CEAP Baseline

and NRI data at CEAP sample points

modeling )

CEAP Baseline, onsite estimates

modeling (HUMUS/ SWAT)

CEAP Baseline, off-site water quality estimates

Schematic for Construction of

Farm survey

Field-level

(APEX

Watershed

8

Flow Calibration Procedure

Total Average Annual Water Yield (USGS Streamflow)

• Calibrate APEX and SWAT within 15%

• Water Balance will Differ Cultivated – Higher Surface Runoff Non Cultivated – Higher Base Flow

R 2

0

50

0 100 200

( )

(mm

)

= 0.89 12

y = 0.897x + 5.6698

100

150

200

250

300

350

400

300 4 00

USGS Flow m m

SW

AT

Sim

ula

ted

Flo

w

9

Flow Validation of SWAT

0

low

(

l

l l i i

0

25

50

75

100

l(m

m)

2000

4000

6000

8000

10000

12000

Year

Stre

am F

cubi

c m

eter

s pe

r sec

ond)

USGS Measured Streamf ow

SWAT Simulated Streamflow

Average month y stream f ow for Illinois r ver at Valley c ty, IL

1961

1962

1963

1964

1965

1966

1967

1968

1969

1970

1971

1972

1973

1974

1975

1976

1977

1978

1979

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

Mon

thy

mea

n st

ream

flow

Observed Predicted

•

•

Sediment and nutrient calibration/validation Assembled USGS-SPARROW annual nutrient loads

Sediment and nutrient data at USGS-Gages

10

Nitrogen Validation of SWATAlton, IL

012345678

0.05 0.15 0.25 0.35 0.45 0.55 0.65 0.75 0.85 0.95

Exceedance Probability

SWAT NO3 (mg/l) USGS NO3 (mg/l)

Nitr

ate

Con

c. (m

g/l)

Alton, IL

012345678

0.05 0.15 0.25 0.35 0.45 0.55 0.65 0.75 0.85 0.95

Exceedance Probability

SWAT NO3 (mg/l) USGS NO3 (mg/l)

Nitr

ate

Con

c. (m

g/l)

SWAT Tot N (mg/l)

Alton, IL

0123456789

10

0.05 0.15 0.25 0.35 0.45 0.55 0.65 0.75 0.85 0.95

Exceedance Probability

SWAT Tot N (mg/l) USGS Tot N (mg/l)

Tota

l N (m

g/l)

SWAT Tot N (mg/l)

0123456789

10 /l)

(/l)

Alton, IL

0.05 0.15 0.25 0.35 0.45 0.55 0.65 0.75 0.85 0.95

Exceedance Probability

SWAT Tot N (mg USGS Tot N (mg/l)

Tota

l Nm

g

Percent Reduction in Sediment Loading by HUC

All Sources Agland (APEX) Only (Ag, Non-Ag, Point Sources)

11

Percent Reduction in Total NitrogenLoading by HUC

All Sources (Ag, Non-Ag, Point Sources) Agland (APEX) Only

Percent Reduction in Total PhosphorusLoading by HUC

All SourcesAgland (APEX) Only (Ag, Non-Ag, Point Sources)

12

Preliminary Reduction ResultsPreliminary Results Grafton, IL

• Sediment Reduction 820,000 tons/yr

• Total Nitrogen 11,036 tons/yr

• 2,351 tons/yr

4.2%

5.1% Total Phosphorus

4.8%

Development (CARD) Iowa State University

)

Center for Agricultural and Rural

14 4-digit watersheds & 131 8-digit watersheds

(111 above Grafton

13

Key Model Components •

•

practices, etc.

•

SWAT model – configured with 8-digit subwatersheds & HRUs

1997 USDA NRI database – 114,000 UMRB points representing baseline land use, conservation

Other tillage, fertilizer, climate, & soil data

Conservation Practices

• Retire Land • Terracing and Contouring • Grassed Waterways • Conservation Tillage

14

15

Future Development

• •

Controlled Drainage • Landscape Processes • • • Databases

Potholes – Surface Tile Inlets Tile Flow – Drainmod Equations and

Pathogens – Bacteria Source Tracking Sediment and In-Stream Routing

16