SPARROW Watershed Modeling of the Entire Great Lakes Basin

IAGLR Conference, McMaster UniversityMay 29, 2014dzrobert@usgs.gov (608) 821-3867

By Dale M. Robertson1,

D.A. Saad1, G.A. Benoy2, W. Jenkinson3, and C.M. Johnston4

1 U.S. Geological Survey, Wis. WSC2 International Joint Commission

3 National Research Council of Canada4 U.S. Geological Survey, NH/VT WSC

Pelee Island, Lake Erie Manitowoc, Lake Michigan

Eutrophication Issues in the Great Lakes

Lake Erie

Cladophora on Beaches

Delavan Lake, Wisconsin

The Importance of Nutrient Loading to Lakes is Well KnownAnd several models available for smaller systems

Goals of SPARROW Modeling for the Great Lakes Basin:

1. Determine P and N loading over the entire basin (large spatial scale).

2. Determine where are the main contributing basins(Rank all contributing basins based on loads and yields).

3. Determine what are the main causes of the high loads (Describe the relative importance of nutrient sources).

4. Provide information to various regional organizations to support regional interpretation and guide local, more in-depth studies.

Pastureland runoff

Pastureland runoff

Cultivated land runoff

Cultivated land runoff

Atmospheric deposition

Atmospheric deposition

Urban runoffUrban runoff

Wastewater discharge

Wastewater discharge

Land to water Land to water delivery

delivery

MonitoringMonitoringstationstation

Grabhorn StudiosGrabhorn Studios

Forested land runoff

Forested land runoff

Flux

Flux

out

Flux

out

Flux inFlux in

Flux out = – Instream Decay Flux in + ( Sources x Delivery)Target =

Instream transport Instream transport and decayand decay

s D I

SPARROW: SPAtially Referenced Regression on Watershed Attributes Watershed Model

ss

DD

II

Mass Balance Modeling

Original Upper Midwest (U.S.) SPARROW ModelOriginal Upper Midwest (U.S.) SPARROW Model

One Source: 2002 Farm Fertilizer TP inputs, kgOne Source: 2002 Farm Fertilizer TP inputs, kg One Land-to-Water Delivery: Soil PermeabilityOne Land-to-Water Delivery: Soil Permeability

River Network – RF1River Network – RF1 Long-term detrended Loads for 810 sites Long-term detrended Loads for 810 sites

Calibration

Distribution in Incremental Phosphorus Yields

Total Phosphorus Yields(kg km-2)0 – 1213 - 1718 - 2525 - 3334 - 4142 - 5152 - 6465 - 8384 - 114115 - 10001001 – 2,980

Superior

Huron

Mic

higa

n

Erie

Ontario

Total Phosphorus Yields(kg km-2)0 – 1213 - 1718 - 2525 - 3334 - 4142 - 5152 - 6465 - 8384 - 114115 - 10001001 – 2,980

Superior

Huron

Mic

higa

n

Erie

Ontario

Distribution in Incremental Phosphorus Yields

Robertson and Saad, 2011

Phosphorus Yields from the U.S. Portion of the Basins by SourceRobertson and Saad, 2011

0

20

40

60

80

Superior Michigan Huron Erie Ontario

TO

TA

L D

EL

IVE

RE

D Y

IEL

D,

KG

/KM

2Forested areas

Manure (unconfined)

Manure (confined)

Fertilizers (farm)

Urban areas

Point Sources

St. LouisSt. Louis

FoxFox

OntonagonOntonagon

MaumeeMaumee

SaginawSaginaw

GrandGrand

St. JosephSt. Joseph

OswegoOswego

GeneseeGenesee

Sagina

w Bay

Superior

Huron

Mic

higa

n

Erie

Ontario

Prioritizing/Ranking TributariesBy Yield

Major Criticisms of Past SPARROW Modeling

1.Only part (U.S. side) of the watersheds have been modeled.

2. Spatial extent rather crude – ~HUC 11, ~100 km2

3. Interbasin transfers and direct deposition were not included

0

20

40

60

80

Superior Michigan Huron Erie Ontario

TO

TA

L D

EL

IVE

RE

D Y

IEL

D,

KG

/KM

2

Forested areas

Manure (unconfined)

Manure (confined)

Fertilizers (farm)

Urban areas

Point Sources

65% 67%

30%

44%

0%

Percent of the Watersheds that were not Modeled

Feb 24-25, 2014Windsor, ON

Mid-Continent SPARROW Models for P and NMid-Continent SPARROW Models for P and N

SPARROW Input Information

NUTRIENT SOURCES (Base Year -2002)

•Point sources and other urban contributions

•Atmos. N deposition – (CMAQ)•Farm fertilizer use allocated to

major crops

•N2 fixation – cultivated lands

•Animal manure•Natural and residual sources

(lands in forest, barren, shrub)

AQUATIC ATTENUATION•Streams

– First-order decay ~ f(water travel time)

•Reservoirs– First-order decay ~ f(areal

hydraulic load)

LAND-TO-WATER DELIVERY•Climate (precipitation,

temperature)•Soils (permeability)•Topography/subsurface (slope,

specific catchment area)•Artificial drainage (tiles,

ditches, new ARSC coverage)

NUTRIENT Load Sites •U.S. - NAWQA•Canada

Streamflow Network •U.S. – NHD NAWQA•Canada - NHN

International QW modeling team:Project lead and principal investigators

• Glenn Benoy – IJC - Ottawa

• Dale Robertson – USGS - Wisconsin

• Wayne Jenkinson – National Research Council – Ottawa

Other PIs:

• David Saad – USGS

• Craig Johnston – USGS

• Mike Laitta – IJC – Washington D.C.

• Erika Klyszejko – Environment Canada

• Martin Serrer – National Research Council – Ottawa

• Richard Burcher – National Research Council – Ottawa

• John Gaiot – Ontario Ministry of Natural Resources

• Jason Vanrobaeys – AAFC

• Rich Moore – USGS

• Pete Steeves – USGS

Data Providers:

• Aaron Todd & Georgina Kaltenecker – Ontario Ministry of the Environment

• Conrad Wyrzykowski – AAFC

• Pam Minifie – Saskatchewan Environment

• Mark Henry – Statistics Canada

• Susan Collins – National Research Council – Ottawa

• Elaine Page – Manitoba Water Stewardship

• O. S. (Arasu) Thirunavukkarasu –Saskatchewan Environment

• Antonette Arvai – IJC – Windsor

• Craig McCrimmon & Tim Pascoe – Environment Canada

Mid-Continental SPARROW Models

Delivered Incremental Phosphorus Yield (kg/ha)

Total Phosphorus

Binational Red/Assiniboine

SPARROW Models

Potential Load Sites for Model Development

U.S. SiteCanadian Site

Streamflow – Environment CanadaWater Quality – PWQMN and MCWS

Ontario NHN Stream Network for the Great Lakes in Canada

1.1 million reaches reduced to 188,535

Harmonizing data in Canada with NHDPlus with border rivers

Pigeon River

St. Marys River

St Clair RiverLake St Clair Detroit River

Niagara RiverSt Lawrence River

2020

Final Catchment Grid for the Great Lakes Stream Network

Nutrient Inputs and Environmental Characteristics of ALL basins

Describing Nutrient Inputs (fertilizers, manure, fixation)

Fertilizer Sales, Animals, Crop types

Wastewater Treatment Plant Effluent

Ontario Ministry of the EnvironmentOntario Clean Water AgencyMissing Areas based on Population

Mid-Continental SPARROW ModelMid-Continental SPARROW ModelOne Source: 2002 Farm Fertilizer TP inputs, kgOne Source: 2002 Farm Fertilizer TP inputs, kg One Land-to-Water Delivery: Soil PermeabilityOne Land-to-Water Delivery: Soil Permeability

River Network – NHD/NHNRiver Network – NHD/NHN Long-term detrended Loads for ~1200 sites Long-term detrended Loads for ~1200 sites

Calibration

0

20

40

60

80

Superior Michigan Huron Erie Ontario

TO

TA

L D

EL

IVE

RE

D Y

IEL

D,

KG

/KM

2

Forested areas

Manure (unconfined)

Manure (confined)

Fertilizers (farm)

Urban areas

Point Sources

+65%

+0%

+67%

+30%

+44%

Complete Basins to be modeled

St. LouisSt. Louis

FoxFox

OntonagonOntonagon

MaumeeMaumee

SaginawSaginaw

GrandGrand

St. JosephSt. Joseph

OswegoOswego

GeneseeGenesee

Sagina

w Bay

Superior

Huron

Mic

higa

n

Erie

Ontario

Prioritizing/Ranking TributariesBy Yield

All Tributaries to be ranked

Online Mapping and Scenario Testing Tools will be Created

1. SPARROW Mapper – Easy and simple way to get SPARROW results, especially by hydrologic and political boundaries.

2. Decision Support System Scientists/Managers – Capable of using to visualize SPARROW output and run various scenarios.

NAWQA SPARROW Modeling

2012 SPARROW Models

Robertson & Saad, WI

RF1 (HUC11) > NHD Plus Scale(1:100,000) (HUC14)2002 Inputs > 2012 Inputs

Predicting Future Phosphorus Loading to the Great Lakes

HydroSPARROW: A GIS Regional Modeling Approach for Estimating Nutrient Loading to the Great Lakes Under Current and Future Climate and Land-Use Conditions

HydroSPARROW

Total P and Total NConcentrations by Reach Streamflow by reach

SPARROW

Total P and Total NLoads

Loads/Yields

Forested land (103 km2)

Agricultural land (103 km2)

Urban land area (103 km2)

Population (106)

Predicted change in population and land use in basin

Projected Changes in Population and Land Use

Predicting Future Nutrient Loading with Changing Land Use

Used to estimate changes in Nutrient Inputs

LaBeau et al., In Review

Change in P loads (% Change) by watershed in 2040 for BioFuels Future Scenario

Predicting the Effects of Climate Change

Streamflow by RF1 reach

PRMS(Christiansen and others)

Total P and Total NConcentrations by Reach

SPARROW

Forecasting Future Nutrient Loading with Changing Climate

Air Temperature

0

1

2

3

4

5

6

7

1981-2000 2046-2065 2081-2100

Ch

ang

e in

Air

Tem

per

atu

re,

C

CGM3-A1BCNRM-A1BCSIRO-A1BGFDL-A1BGISS-A1BMIUB-A1BMPI-A1BMRI-A1BCGM3-A2CNRM-A2CSIRO-A2GFDL-A2GISS-A2MIUB-A2MPI-A2MRI-A2CGM3-B1CNRM-B1CSIRO-B1GFDL-B1GISS-B1MIUB-B1MPI-B1MRI-B1Average

Precipitation

-35%

-25%

-15%

-5%

5%

15%

25%

35%

1981-2000 2046-2065 2081-2100

Per

cen

tag

e C

han

ge

in P

reci

pit

atio

n

Streamflow

-35%

-25%

-15%

-5%

5%

15%

25%

1981-2000 2046-2065 2081-2100

Per

cen

tag

e C

han

ge

in S

trea

mfl

ow

Predicting Future Changes in Nutrient Loading

PRMSPhosphorus Load

-35%

-25%

-15%

-5%

5%

15%

25%

1981-2000 2046-2065 2081-2100

Per

cen

tag

e C

han

ge

in P

ho

sph

oru

s L

oad

HydroSPARROW

Air Temperature

Precipitation

StreamflowPhosphorus Loading

Projected Changes (average) by 2046-2065

Average of 8 models and 3 carbon emission scenarios

Dale M. Robertsondzrobert@usgs.gov (608) 821-3867

Questions??

SPARROW Watershed Modeling of the Entire Great Lakes Basin