The Wisconsin River TMDL:Linking Monitoring and Modeling
Ann Hirekatur, Pat Oldenburg, & Adam FreihoeferMarch 7, 2013
Wisconsin River TMDL Project TeamWDNR Bureau of Water Quality
The Wisconsin River Basin (WRB)
The WRB – A managed system
http://www.lakenokomiscc.org
Water Quality Concerns within the WRB
Petenwell and Castle Rock FlowagesFish Kills in Big Eau Pleine ReservoirDells of the Eau Claire
Runoff in the Baraboo
Pollutant Sources within the WRBN
on
po
int
So
urc
es
Non-MS4 Stormwater
Agricultural Runoff
Rill, Gully, & Bank Erosion
Construction
Barnyards
Po
int
So
urc
es
Municipal WasteIndustrial Waste
WRB Total Maximum Daily Load (TMDL)
Waste Load Allocation• Municipal Wastewater• Industrial Wastewater• Stormwater (MS4s)
Load Allocation• Runoff from the landscape
Background Load• Naturally occurring from
wetlands, forests
Load Allocation
Waste LoadAllocation
Margin of Safety
TMDL
+ +
Each subwatershed is assessed for:
Components of the WRB TMDL
Monitoring• Flow• Chemistry• In-lake
Conceptualization• Data Analysis• Windshield
Surveys
Modeling• Watershed• Point Sources• Fate and Transport• Reservoirs
Allocations• Load• Wasteload• Margin of Safety
Draft TMDLInternal Review Public Comment
Final TMDL EPA Approval
Implementation
• Civic Engagement • Public Outreach/Communications
Past Activities in the WRB
1995 2000 2005 2010 2015
1996Petenwell
Castle RockComprehensive
Management Plan
2001WDNR/USACE develop joint TMDL
monitoring proposal for $1.5 million. Proposal not funded
2004WDNR proposal to USEPA for river
water quality monitoring. Proposal not funded
2008First Pontoons and
Politics at Petenwell Castle Rock
2010WNDR receives Section 22
Planning Assistance from USACE
2009 – 2013Wisconsin River
TMDL Monitoring
2009State legislature appropriates $750,000
($150,000/yr. for 5 years) for monitoring upstream of Castle Rock
Wisconsin RiverTMDL
Monitoring
The WRB Monitoring - Modeling Relationship
Conceptualization
WRB Monitoring Approach
Critical IdeaDesign a monitoring plan that provides an understanding of the system and supports the modeling of the watershed & reservoirs
WRB Monitoring Approach
= ungauged tributary area
Tributary Watershed Loads19 stations with daily discharge & bi-monthly water quality
WRB Monitoring ApproachTributary Watershed Loads19 stations with daily discharge & bi-monthly water quality
Phosphorus Evaluation Sites98 stations with monthly P samples between May – October 2012 with 31 sites proposed for re-sample in 2013
WRB Monitoring ApproachTributary Watershed Loads19 stations with daily discharge & bi-monthly water quality
Phosphorus Evaluation Sites98 stations with monthly P samples between May – October 2012 with 31 sites proposed for re-sample in 2013
Main stem Loads13 stations with daily discharge & bi-monthly water quality
WRB Monitoring ApproachTributary Watershed Loads19 stations with daily discharge & bi-monthly water quality
Phosphorus Evaluation Sites98 stations with monthly P samples between May – October 2012 with 31 sites proposed for re-sample in 2013
Reservoirs5 major reservoirsBig Eau Pleine, Lake Dubay, Petenwell, Castle Rock , & Lake Wisconsin
Main stem Loads13 stations with daily discharge & bi-monthly water quality
• Simulates conditions on landscape each day based on climate data
• Input data intensive
• Output information is provided for each subwatershed defined (TMDL reach)
• Outputs include crop yields, discharge, sediment, & water chemistry
WRB Model Approach – Watershed
WRB Model Approach – Watershed
WATERSHED MODEL
Watershed Model Calibrated
Outputs
HydrographyFlow accumulation, Internal drainage, groundwater
200
9 –
201
3 C
on
dit
ion
s
TopographySlope
SoilsType and Attributes
Climate Precipitation, Temp, etc.
Land ManagementAgriculture, Urban
Land Cover
Model Input
Model Output
Point Sources
Model Calibration
Required
Petenwell Bathymetry(Developed by USACE)
WRB Model Approach – Reservoir
WRB Model Approach – Reservoir
Model Inputs
Reservoir Model
Outputs
RESERVOIRMODEL
Model Calibration
Required
Watershed Model Calibrated Outputs
Model Input
Model Output
WRB Model Approach – Model Calibration
• Calibration is process of matching the simulation to measured data
• Modeler adjusts model parameters to allow for better fit
Input Parameter Controls
Integrating Monitoring Data into Models
1. Assess the daily flow and TP concentrations
2. Use regression model to estimate monthly TP load, referred to as “measured load”
3. Compare the watershed model “simulated load” to the “measured load”. Adjust the model to reflect measured conditions, a process referred to as model calibration
Tributary Loads P Evaluation Sites River Loads Reservoirs
Dill Creek - Big Eau Pleine River
Integrating Monitoring Data into Models
Tributary Loads P Evaluation Sites River Loads Reservoirs
Jan-10 Jun-10 Dec-10 Jun-11 Dec-11 Jun-12 Dec-120
2,000
4,000
6,000
8,000
10,000
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1Big Eau Pleine River at State Hwy 97 (2010 - 2012)
Ave
rage
Dai
ly D
isch
arge
(cf
s)T
otal Ph
osph
orus C
oncen
tration (m
g/L)
Integrating Monitoring Data into Models
Tributary Loads P Evaluation Sites River Loads Reservoirs
Jan-10 Apr-10 Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct-11 Jan-12 Apr-12 Jul-12 Oct-120
20,000
40,000
60,000
80,000
100,000
120,000
140,000
160,000
180,000Big Eau Pleine River at State Hwy 97 (2010 - 2012)
Measured Loads
Tot
al P
hos
ph
oru
s L
oad
(lb
s. p
er m
onth
)
Integrating Monitoring Data into Models
Tributary Loads P Evaluation Sites River Loads Reservoirs
Jan-10 Apr-10 Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct-11 Jan-12 Apr-12 Jul-12 Oct-120
20,000
40,000
60,000
80,000
100,000
120,000
140,000
160,000
180,000Big Eau Pleine River at State Hwy 97 (2010 - 2012)
Measured Loads Simulated Loads
Tot
al P
hos
ph
oru
s L
oad
(lb
s. p
er m
onth
)
EXAMPLE
Integrating Monitoring Data into Models
Tributary Loads P Evaluation Sites River Loads Reservoirs
05/01/2012 05/31/2012 06/30/2012 07/30/2012 08/29/2012 09/28/2012 10/28/20120.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60Randall Creek Phosphorus Concentrations (2012)
Model Simulated TP ConcentrationMeasured Phosphorus
Dai
ly P
hos
ph
oru
s C
once
ntr
atio
n (
mg/
L)
EXAMPLE
Integrating Monitoring Data into Models
Tributary Loads P Evaluation Sites River Loads Reservoirs
• Same approach as the tributary loads (daily discharge, bi-monthly water quality)
Jan-10 Apr-10 Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct-11 Jan-12 Apr-12 Jul-12 Oct-120
50,000
100,000
150,000
200,000
250,000
300,000
350,000
400,000Wisconsin River at Nekoosa (entering Petenwell) (2010 - 2012)
Measured Loads
Tot
al P
hos
ph
oru
s L
oad
(lb
s. p
er m
onth
)
Integrating Monitoring Data into Models
Tributary Loads P Evaluation Sites River Loads Reservoirs
• Same approach as the tributary loads (daily discharge, bi-monthly water quality)
Jan-10 Apr-10 Jul-10 Oct-10 Jan-11 Apr-11 Jul-11 Oct-11 Jan-12 Apr-12 Jul-12 Oct-120
50,000
100,000
150,000
200,000
250,000
300,000
350,000
400,000Wisconsin River at Nekoosa (entering Petenwell) (2010 - 2012)
Measured Loads Simulated Loads
Tot
al P
hos
ph
oru
s L
oad
(lb
s. p
er m
onth
)
EXAMPLE
Integrating Monitoring Data into Models
Tributary Loads P Evaluation Sites River Loads Reservoirs
6 8 10 12 14 16 18 20 22 24 260
2
4
6
8
10
12
14
16
Petenwell DO / Temp Profile (Measured vs. Model Simulated)
Measured DO Simulated DO Measured Temp Simulated Temp
Dep
th
EXAMPLE
Wrap-Up• The scale of the Wisconsin River TMDL required more monitoring sites, frequency
• Defining the question and the methods used at the start improved the project’s monitoring design.
• Each part of the TMDL (monitoring, load calculations, watershed modeling) introduces uncertainty. In many cases, a robust monitoring dataset improves our understanding and limits the uncertainty with respect to the model results.
Questions?
Contact the Wisconsin River TMDL Project [email protected]