Estimating Bank Erosion in the Estimating Bank Erosion ...
Estimating Bank Erosion in the Wissahickon Creek Watershed: A Bank Pin Monitoring Approach
Erik Haniman Philadelphia Water Department Office of Watersheds
Introduction & Background
Wissahickon Sediment TMDL Monitoring Accumulation of sediment can be harmful to aquatic habitat PADEP has identified “siltation” to be an impairment in Wissahickon Creek – 6 mainstem segments g – 15 tributary segments
Where is the sediment originating from? – Runoff, StreamStream-bank erosion
What is the sediment load from each source?
Philadelphia’s 20042004-2009 S Stormwater tormwater Permit Includes a commitment by the City of Philadelphia to initialize efforts toward compliance p with the Wissahickon Creek Sediment TMDL – (1) Establish baseline data on the City’s S Sediment C Contribution – (2) Evaluate and implement BMPs St Stream R Restoration, t ti Stormwater St t Basin, B i Green G Infrastructure
– ((3)) Monitor effectiveness of BMPs
Channel Evolution Model
Schumm, 1984; Simon & Hupp, 1986
Rosgen, 2006
Methods
Rosgen’s Rosgen s BANCS Model BANCS = Bank Assessment for Non on--Point source Consequences of Sediment Model was developed to allow field practitioners to quantitatively assess and estimate bank erosion Based on streambank erodibility variables and d energy di distribution t ib ti within ithi th the stream t – BEHI Rating – Near Near--Bank Stress
BEHI = Bank Erosion Hazard Index
Rosgen, 2006
NBS – Near Bank Stress
Rosgen, 2006
Rosgen, 2006
Estimate Verification – Bank Pin Installation and Monitoring
Rosgen, 2006
Schedule BANCS Model – 12 Tributaries, 130,000 feet of stream--bank stream – October, 2005 – August, 2006
Bank Profile Monitoring (Bank Pins) – 82 sites w/ at least 1 full year of data – August, 2006 – August, 2008
Data Processing & Analysis – August, 2008 – December, 2008
Results
BANCS Model Prediction
4.2 Million Pounds Per Year
Bank Pin Measurement Verification
2.3 Million Pounds Per Year 95% CI: +/+/- 2.5 Million Pounds Per Year
0 Tributary I
Wises Mill
Va alley Green
Thomas Mill T
Monoshone M
Kitcchens Lane
Hillcrest
Hartwell
Go orgas Lane
Creshiem
Cathedral
Bells Mill
(lb/ac/y yr)
Erosion Per Acre of Drainage Area
1800
1600
1400
1200
1000
800
600
400
200
0 Tributary I
Wises Mill
Valley Green
Thomas Miill
Monoshone e
Kitchens Lane K e
Hillcresst
Hartwe ell
Gorgas Lane e
Creshiem m
Cathedra al
Bells Mill
(lb/ft/ yr)
Erosion Per Foot of Stream Length
80
70
60
50
40
30
20
10
Bank Erosion Rates Measured at 81 Bank Pin Locations Classed by BEHI Category 0.2
0.1
Erosion Rate (ftt/yr)
0.0
-0.1
-0.2
-0.3
-0.4 Mean
Mean±SE
Non-Outlier Range
-0.5 Low
Mod Erosion Category
High g
Hypothesis of Individual p p populations based upon BEHI Class is not justified
Wissahickon Phila Tributaries Bank Erosion Rate Estimate
Wissahickon Phila Tributaries Bank Erosion Estimate
-0.01
4 5E6 4.5E6 -0.02
4E6 -0.03
-0.04
-0.05 0 05
-0.06
3E6 Bank Erosion (lb/yr)
Bank Erosion (ft/yr)
3 5E6 3.5E6
2 5E6 2.5E6 2E6 1 5E6 1.5E6
-0.07 Mean = -0.043 Mean±0.95 Conf. Interval = (-0.0719, -0.014)
1E6 -0.08 Erosion Rate (less outliers)
5E5
Mean = 2.3933E6 Min-Max = (7.8E5, 4E6)
0 Var1
Tributary
BEHI/NBS Predictions (lb/yr)
Total Erosion {3 Bank Pins Populations Classed by BEHI Rating} (lb/yr)
Bells Mill
310,000
160,000
310,000
Cathedral
300,000
160,000
100,000
Creshiem
740,000
570,000
650,000
Gorgas g Lane
310,000 ,
160,000 ,
120,000 ,
Hillcrest
160,000
120,000
30,000
Hartwell
62,000
34,000
80,000
Kitchens Lane
260,000
180,000
150,000
Monoshone
140,000
62,000
130,000
Thomas Mill
300,000
180,000
250,000
Valley Green
210,000
110,000
110,000
Wises Mill
310,000
410,000
440,000
Tributary I
210,000
110,000
70,000
Total
4,200,000
2,300,000 (+/- 2,500,000)
2,400,000 (+/- 1,600,000)
Total Erosion {Bank Pins One Population} (lb/yr)
Discussion & Conclusions
Rosgen, 2006
Stream Bank Erodibility 10 High Very High Moderate Low Expon. (Low) Expon. (Moderate)
Bank Erosion Rate (ft/yr)
1
Expon. (High) Expon. (Very High) y = 0.1175e
0.1746x
2
R = 0.0076
y = 0.1036e
0.1
0.0062x
2
R = 2E-05
y = 0.0418e
0.2355x
2
R = 0.0069
0.01
y = 1.6328e
-1.8189x
2
R = 0.9259
0.001 Very Low
Low
Moderate
High
Near Bank Stress
Very High
Extreme
What Did We Learn Learn. . . Philadelphia’s Philadelphia s Wissahickon Tributary StreamStreambanks erode roughly 2,000,000 lbs/yr of sediment The BANCS model is an excellent model for predicting stream stream--bank erosion – Order of magnitude or better
Bank Erosion Curves are difficult to develop Understanding Channel Evolution is key
Erik Haniman Philadelphia Water Department Office of Watersheds
Introduction & Background
Wissahickon Sediment TMDL Monitoring Accumulation of sediment can be harmful to aquatic habitat PADEP has identified “siltation” to be an impairment in Wissahickon Creek – 6 mainstem segments g – 15 tributary segments
Where is the sediment originating from? – Runoff, StreamStream-bank erosion
What is the sediment load from each source?
Philadelphia’s 20042004-2009 S Stormwater tormwater Permit Includes a commitment by the City of Philadelphia to initialize efforts toward compliance p with the Wissahickon Creek Sediment TMDL – (1) Establish baseline data on the City’s S Sediment C Contribution – (2) Evaluate and implement BMPs St Stream R Restoration, t ti Stormwater St t Basin, B i Green G Infrastructure
– ((3)) Monitor effectiveness of BMPs
Channel Evolution Model
Schumm, 1984; Simon & Hupp, 1986
Rosgen, 2006
Methods
Rosgen’s Rosgen s BANCS Model BANCS = Bank Assessment for Non on--Point source Consequences of Sediment Model was developed to allow field practitioners to quantitatively assess and estimate bank erosion Based on streambank erodibility variables and d energy di distribution t ib ti within ithi th the stream t – BEHI Rating – Near Near--Bank Stress
BEHI = Bank Erosion Hazard Index
Rosgen, 2006
NBS – Near Bank Stress
Rosgen, 2006
Rosgen, 2006
Estimate Verification – Bank Pin Installation and Monitoring
Rosgen, 2006
Schedule BANCS Model – 12 Tributaries, 130,000 feet of stream--bank stream – October, 2005 – August, 2006
Bank Profile Monitoring (Bank Pins) – 82 sites w/ at least 1 full year of data – August, 2006 – August, 2008
Data Processing & Analysis – August, 2008 – December, 2008
Results
BANCS Model Prediction
4.2 Million Pounds Per Year
Bank Pin Measurement Verification
2.3 Million Pounds Per Year 95% CI: +/+/- 2.5 Million Pounds Per Year
0 Tributary I
Wises Mill
Va alley Green
Thomas Mill T
Monoshone M
Kitcchens Lane
Hillcrest
Hartwell
Go orgas Lane
Creshiem
Cathedral
Bells Mill
(lb/ac/y yr)
Erosion Per Acre of Drainage Area
1800
1600
1400
1200
1000
800
600
400
200
0 Tributary I
Wises Mill
Valley Green
Thomas Miill
Monoshone e
Kitchens Lane K e
Hillcresst
Hartwe ell
Gorgas Lane e
Creshiem m
Cathedra al
Bells Mill
(lb/ft/ yr)
Erosion Per Foot of Stream Length
80
70
60
50
40
30
20
10
Bank Erosion Rates Measured at 81 Bank Pin Locations Classed by BEHI Category 0.2
0.1
Erosion Rate (ftt/yr)
0.0
-0.1
-0.2
-0.3
-0.4 Mean
Mean±SE
Non-Outlier Range
-0.5 Low
Mod Erosion Category
High g
Hypothesis of Individual p p populations based upon BEHI Class is not justified
Wissahickon Phila Tributaries Bank Erosion Rate Estimate
Wissahickon Phila Tributaries Bank Erosion Estimate
-0.01
4 5E6 4.5E6 -0.02
4E6 -0.03
-0.04
-0.05 0 05
-0.06
3E6 Bank Erosion (lb/yr)
Bank Erosion (ft/yr)
3 5E6 3.5E6
2 5E6 2.5E6 2E6 1 5E6 1.5E6
-0.07 Mean = -0.043 Mean±0.95 Conf. Interval = (-0.0719, -0.014)
1E6 -0.08 Erosion Rate (less outliers)
5E5
Mean = 2.3933E6 Min-Max = (7.8E5, 4E6)
0 Var1
Tributary
BEHI/NBS Predictions (lb/yr)
Total Erosion {3 Bank Pins Populations Classed by BEHI Rating} (lb/yr)
Bells Mill
310,000
160,000
310,000
Cathedral
300,000
160,000
100,000
Creshiem
740,000
570,000
650,000
Gorgas g Lane
310,000 ,
160,000 ,
120,000 ,
Hillcrest
160,000
120,000
30,000
Hartwell
62,000
34,000
80,000
Kitchens Lane
260,000
180,000
150,000
Monoshone
140,000
62,000
130,000
Thomas Mill
300,000
180,000
250,000
Valley Green
210,000
110,000
110,000
Wises Mill
310,000
410,000
440,000
Tributary I
210,000
110,000
70,000
Total
4,200,000
2,300,000 (+/- 2,500,000)
2,400,000 (+/- 1,600,000)
Total Erosion {Bank Pins One Population} (lb/yr)
Discussion & Conclusions
Rosgen, 2006
Stream Bank Erodibility 10 High Very High Moderate Low Expon. (Low) Expon. (Moderate)
Bank Erosion Rate (ft/yr)
1
Expon. (High) Expon. (Very High) y = 0.1175e
0.1746x
2
R = 0.0076
y = 0.1036e
0.1
0.0062x
2
R = 2E-05
y = 0.0418e
0.2355x
2
R = 0.0069
0.01
y = 1.6328e
-1.8189x
2
R = 0.9259
0.001 Very Low
Low
Moderate
High
Near Bank Stress
Very High
Extreme
What Did We Learn Learn. . . Philadelphia’s Philadelphia s Wissahickon Tributary StreamStreambanks erode roughly 2,000,000 lbs/yr of sediment The BANCS model is an excellent model for predicting stream stream--bank erosion – Order of magnitude or better
Bank Erosion Curves are difficult to develop Understanding Channel Evolution is key
