DESIGNING MONITORING PROGRAMS TO
EVALUATE BMP EFFECTIVENESS
Nancy Mesner - Utah State University, Dept of Watershed Sciences
Ginger Paige - University of Wyoming
David Stevens, Jeff Horsburgh, Doug Jackson-Smith, Darwin Sorensen, Ron
Ryel – USU
Funded by grants from USDACSREES, EPA 319, NSF
Examples from the
Little Bear River CEAP Project
Pre-treatment problems:
Bank erosion, manure management, flood irrigation problems
Treatments:
 bank stabilization,
 river reach restoration,
 off-stream watering,
 improved manure and
water management
Common problems in BMP monitoring
programs:
•
Failure to design monitoring plan around BMP objectives
• A failure to understand pollutant pathways and
transformations and sources of variaiblity in these dynamic
system.
• Tend to draw on a limited set or inappropriate approaches
• Failure to design monitoring plan
around BMP objectives
• A failure to understand pollutant pathways and
transformations and sources of variaiblity in these
dynamic system.
• Tend to draw on a limited set or inappropriate
approaches Designing monitoring to address
specific objectives
v
Little Bear River Watershed, Utah
Total Observations at Watershed Outlet site
1976 - 2004:
1994 - 2004:
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
Discharge
Total phosphorus
162
72
241
99
11
10
10
11
6
7
6
4
2
4
1
13
13
13
4
10
10
5
7
8
8
8
Number of
observations
each year
• Failure to design monitoring plan around BMP objectives
• A failure to understand pollutant pathways and
transformations and sources of variability in these dynamic
system.
• Tend to draw on a limited set or inappropriate approaches
Understanding natural variability –
annual variation
10105900 - LITTLE BEAR RIVER AT PARADISE, UT
900
800
700
FLOW, STREAM, MEAN DAILY CFS
600
500
400
300
200
100
0
1995
2000
Date
2005
Since 2005, measure flow and turbidity at 30 minute
intervals
Stage recording
devices to estimate
discharge
http://www.campbellsci.com
Turbidity sensors
http://www.ftsinc.com/
Dataloggers and
telemetry
equipment
http://www.campbellsci.com
Capturing pollutant movement from source
to waterbody.
Little Bear River Near Paradise
450
Storm
Event
400
350
300
250
200
150
100
Date
Streamflow (cfs)
Turbidity (NTU)
3/21/2006
3/19/2006
3/17/2006
3/15/2006
3/13/2006
3/11/2006
3/9/2006
3/7/2006
3/5/2006
3/3/2006
0
3/1/2006
50
The relative importance of different sources of variability
Coefficient of variation of estimates
100
90
Sampling
frequency
80
70
60
50
40
Regressions of
TP and turbidity
30
20
10
0
30
20
10
5
1
Grab samples -- sampling frequency (d)
0.45 0.55 0.64 0.72 0.78 0.84 0.89 0.94 0.98
Continuous monitoring -- R2 between TP and turbidity
• Failure to design monitoring plan around BMP objectives
• A failure to understand pollutant pathways and transformations and
sources of variability in these dynamic system.
• Tend to draw on a limited set or inappropriate approaches
Problems with “one-size-fits-all” monitoring design
Rees Creek TSS load
50000
45000
40000
kg / day
35000
30000
25000
20000
Above
15000
Problem:
excess sediment
10000
Below
Average
flow
=
20
cfs
5000
BMP = series
of in-stream sediment basins
0
1
2
3
4
5
weeks
6
7
8
9
Bear River phosphorus load
400
350
load (kg/day)
300
250
200
150
100
Problem:
excess phosphorus
50
Average
flow = 1000 cfs
0
BMP = fence
cattle
OUT
of 4riparian
1
2
3
5 area
6 and 7revegetate
8
weeks
9
Focuses on the
considerations and
decisions necessary as a
project is first being
considered.
NOT a “how-to” manual of
protocols
Document in review
Training workshops underway
What is your objective?
 Long term trends?
 UPDES compliance?
 Educational?
 Assessment for impairment?
 Track response from an implementation?
How do pollutants “behave” within your watershed.
 How does the pollutant move from the source to the waterbody?
 How is the pollutant processed or transformed within a waterbody?
 What is the natural variability of the pollutant? Will concentrations
change throughout a season? Throughout a day?
 What long term changes within your watershed may also affect this
pollutant?
 What else must be monitored to help interpret your data?
What to monitor?
 Monitor the pollutant(s) of concern?
 Monitor a “surrogate” variable?
 Monitor a response variables?
 Monitor the impacted beneficial use?
 Monitor the BMP itself?
 Monitor human behavior?
 Model the response to a BMP implementation.
 Collect other data necessary to interpret monitoring results OR
calibrate and validate the model?
Where and when to monitor?
Choose appropriate monitoring or modeling
Control
Treatment “A”
Sampling
points
BACI Design
Abovetreatment
monitoring
stations
Below-treatment
monitoring
stations
Above and below
treatment design
How to monitor?
 points in time versus continuous
 integrated versus grab samples
 consider:
cost
skill and training required
accessibility of sites
The road to more effective monitoring….
 Monitoring plans require careful thought before
anything is implemented.
 Consider how the data will be used to demonstrate
change.
 Use your understanding of your watershed and how
the pollutants of concern behave to target monitoring
most effectively
 Use different approaches for different BMPs
different sources of
variability in
estimates of loads