Mid-Coast Bacteria TWG 17th Meeting (final) - oregon.gov · Mid-Coast Bacteria TMDL: Seventeenth...

Mid-Coast Bacteria TMDL: Seventeenth TWG MeetingMid Coast Bacteria TMDL: Seventeenth TWG Meeting

Kevin BrannanOregon DEQA il 13 2016April 13, 2016

Meeting Objectives

• Updates and Announcements from TWG members• Load Duration Curves StatusLoad Duration Curves Status• Beaches Status• Big Elk Creek Modelingg g• Other modeling

Slide 2

Load Duration Curve based TMDLs - Status

• Compiling reports• Conducting additional analysisConducting additional analysis

– Source characterization– Estuary

Slide 3

Beach TMDLs

• Compiling reports• Additional source characterizationAdditional source characterization• Additional monitoring

Slide 4

Watershed Modeling: Big Elk Creek

• Extended the simulation period– Updated simulation period 1995-10-01 to 2014-05-31Updated simulation period 1995 10 01 to 2014 05 31– Previous simulation period 1995-10-01 to 2010-12-31

• Model calibration re-done• Model parameter uncertainty analysis re-done• Contractor reviewing overall model

F d d f USEPA– Funded form USEPA– Methods for review developed– Completion Nov 2015

Slide 5

Completion Nov 2015


Slide 6

Big Elk Creek: Extended the simulation periodDaily Total Precip

Previous

Slide 7

Updated

Big Elk Creek: Extended the simulation periodAverage Annual Precip

Previous

Slide 8

Updated

Big Elk Creek: Extended the simulation periodAnnual Total Precip

Previous

Slide 9

Updated

Big Elk Creek: Extended the simulation periodMonthly Average Precip

Previous

Slide 10

Updated

Model Calibration

• Calibrated 24 parameters• Varied by land useVaried by land use

(forest, pasture or residential-transportation• Varied by soil permeability

(high and low permeable areas related to elevation)• Calibrated parameters related to:

Soil infiltration along with surface and sub surface storage– Soil, infiltration along with surface and sub-surface storage– Sub-surface flow, interflow and groundwater flow– Plants, ET and interception

Slide 11

Parameter Names and DescriptionsParameter Name Parameter Description

LZSN Lower zone nominal soil moisture storage

INFILT Index to mean soil infiltration rate

AGWRC Groundwater recession rate

DEEPFR Fraction of infiltrating water lost to deep aquifers

BASETP ET by riparian vegetation as active groundwater enters streambedBASETP ET by riparian vegetation as active groundwater enters streambed

AGWET Fraction of watershed subject to direct evaporation from groundwater storage

UZSN Nominal upper zone soil moisture storagepp g

INTFW Interflow coefficient

IRC Interflow recession coefficient

CEPSC M i i it ti i t ti

Slide 12

CEPSC Maximum precipitation interception

LZETP Index to lower zone evapotranspiration

Calibration Performance MeasuresCalibration Performance MeasuresDefinitions

• Log-transformed daily mean flows (qlog)• Total flow volumes (QT t l)Total flow volumes (QTotal)• Summer (June - August) flow volumes (Qsmr)• Winter (December - March) flow volumes (Qwtr)( ) ( wtr)• Storm volumes and peak flows for selected storm periods

(Qstm-i and qpeak-i for the i-th storm)Fl d i i i ( d i f• Flow duration statistics (exceedance times for q5%, q10%, q25%, q50% and q95% flows)

• Baseflow index (the ratio of groundwater outflow to total

Slide 13

Baseflow index (the ratio of groundwater outflow to total outflow from pervious land segments, b-index)


Period Start End Years StormsOriginal 1995-10-01 2010-12-31 15 18Updated 1995-10-01 2014-05-31 19 23

Slide 14

Calibration Performance Measures:Calibration Performance Measures: Original Period

Flow Statistic Observed Modeled Percent ErrorQTotal (in/yr) 53.12 52.87 -0.5%Qsmr (in/yr) 2.11 2.09 -0.8%Qwtr (in/yr) 28.31 27.15 -4.1%Q t (in/yr) 4 39 3 89 -11 4%Qstm (in/yr) 4.39 3.89 11.4%qmean (cfs) 2,350 2,176 -7.4%Baseflow Index 0.44 0.41 -6.0%

Slide 15

Calibration Performance Measures:Calibration Performance Measures: Updated Period

Flow Statistic Observed Modeled Percent ErrorQTotal (acre-ft) 52.44 51.51 -1.8%Qsmr (acre-ft) 2.08 2.05 -1.1%Qwtr (acre-ft) 29.76 28.25 -5.1%Q t (acre-ft) 3 70 3 29 -11 0%Qstm (acre ft) 3.70 3.29 11.0%qmean (cfs) 1,872 1,786 -4.6Baseflow Index 0.44 0.435 -1.0%

Slide 16

Calibration Performance Measures:Calibration Performance Measures: Updated Period

Flow Statistic Original UpdatedQTotal (acre-ft) -0.5% -1.8%Qsmr (acre-ft) -0.8% -1.1%Qwtr (acre-ft) -4.1% -5.1%Q t (acre-ft) -11 4% -11 0%Qstm (acre ft) 11.4% 11.0%qmean (cfs) -7.4% -4.6Baseflow Index -6.0% -1.0%

Slide 17

Flow Duration Curves

Original Period Updated Period

Slide 18


Slide 19

Uncertainty Analysis Re-Do

• Original– Generate random parameter values without considering

parameter covariance– Run model for each parameter set– sub-set based on USGS FDCsub set based on USGS FDC

• Updated– use Covariance Matrix to generate random parameter

values– Constrain random parameters by calibration – sub-set based on USGS FDC

Slide 20

sub set based on USGS FDC

Uncertainty Analysis

• Used criteria that FDC within error bar bounds of USGS eq

• Allowed 2 iterations• Set Φlimit to within 1%

of calibration Φ

Slide 21


Slide 22

Select Uncertainty Analysis Runs

• None of the runs within UGS Eq error bars• Vary number of points within error barsVary number of points within error bars

– 5 out of 5 points = no runs out of 1,000– 4 out of 5 points = 47 runs out of 1,000– 3 out of 5 points = 636 runs out of 1,000– 2 out of 5 points = 984 runs out of 1,000

1 out of 5 points = 995 runs out of 1 000– 1 out of 5 points = 995 runs out of 1,000• None of the runs with value within error bar for lowest flow• Redo with tighter range and more calibration constraints

Slide 23

Redo with tighter range and more calibration constraints

Additional calibration????

• Do additional hydrologic calibration?• Include hydrologic calibration in water quality calibration?Include hydrologic calibration in water quality calibration?• Use current hydrologic calibration and move on to water

quality calibration

Slide 24


Slide 25

Watershed Modeling: Big Elk Creek (continued)

• Current ODEQ work– Conduct hydrologic parameter estimation/uncertaintyConduct hydrologic parameter estimation/uncertainty

analysis for new simulation period– Setup bacteria model for new simulation period

C ti t i t TWG f db k i t b t i b– Continue to incorporate TWG feedback into bacteria sub-model input

• Compile reportp p

Slide 26

Additional Watershed Modeling:Additional Watershed Modeling:Upper Yaquina River

• Develop watershed model like Big Elk CreekS t i l t ti• Support implementation

• Begin model development for Yaquinadevelopment for Yaquina River

• Contractor began work J 2015June-2015

Slide 27


• Utilize information from Big Elk Creek modeling

WDM file– WDM file– Hydro parameters– Bacteria input– …

Slide 28

Watershed Modeling: Upper Yaquina River

• Utilize information from Big Elk Creek modeling

WDM file– WDM file– Hydro parameters– Bacteria input– Model review– …

• Completion estimatedCompletion estimated toward end of 2016

Slide 29


• Use for implementation and possibly for DO TMDL• Hydrologic calibration – used Big Elk parameter valuesHydrologic calibration used Big Elk parameter values

without modification • Pause on bacteria modeling

Slide 30

Mean Daily FlowAvg Monthly Rainfall (in)Avg Observed Flow (10/1/2003 to 9/30/2014 )Avg Modeled Flow (Same Period)

0

2

46000

7000

8000

n)

Avg Modeled Flow (Same Period)

4

6

8

103000

4000

5000

ily R

ain

fall

(in

Flo

w (

cfs

)

10

12

140

1000

2000

Oct-03 Apr-05 Oct-06 Apr-08 Oct-09 Apr-11 Oct-12 Apr-14

Da

Oct 03 Apr 05 Oct 06 Apr 08 Oct 09 Apr 11 Oct 12 Apr 14

Date

Slide 31

Model DSN 262 vs. S 14306030 Yaquina River Near Chitwood, Oregon

Mean Monthly FlowAvg Monthly Rainfall (in)Avg Observed Flow (10/1/2003 to 9/30/2014 )Avg Modeled Flow (Same Period)

20

25

800

1000

1200

(in)

g ( )

10

15

400

600

800

nth

ly R

ain

fall

Flo

w (

cfs

)

0

5

0

200

O-03 A-05 O-06 A-08 O-09 A-11 O-12 A-14

Mon

O-03 A-05 O-06 A-08 O-09 A-11 O-12 A-14

Month

Slide 32


Monthly Regression and Temporal VariationAvg Flow (10/1/2003 to 9/30/2014 )Line of Equal Value

Avg Observed Flow (10/1/2003 to 9/30/2014 )Avg Modeled Flow (10/1/2003 to 9/30/2014 )

y = 0.9386x + 24.609R² = 0.94941000

1200

w (

cfs

)

Best-Fit Line

80%

90%

100%

+ M

od

)

g ( )Line of Equal Value

600

800

Mo

de

led

Flo

w

40%

50%

60%

70%

ance

(O

bs +

0

200

400

Ave

rag

e M

0%

10%

20%

30%

Wa

ter

Ba

l

00 200 400 600 800 1000 1200

Average Observed Flow (cfs)

0%O-03 A-05 O-06 A-08 O-09 A-11 O-12 A-14

Month

Slide 33


Seasonal Regression and Temporal AggregateAvg Flow (10/1/2003 to 9/30/2014)Line of Equal ValueBest-Fit Line

Avg Monthly Rainfall (in)Avg Observed Flow (10/1/2003 to 9/30/2014)Avg Modeled Flow (Same Period)

y = 0.9551x + 20.773R² = 0.9862

600

800

w (

cfs)

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep0

2600

800

(in)

g ( )

400

Mo

de

led

Flo

w

4

6

8

400

hly

R

ain

fall

Flo

w (

cfs

)

0

200

Ave

rag

e 8

10

120

200

Mo

nth

0 200 400 600 800

Average Observed Flow (cfs)

10 11 12 1 2 3 4 5 6 7 8 9

Month

Slide 34


Seasonal Medians and Ranges

0900

Observed (25th, 75th) Average Monthly Rainfall (in)Median Observed Flow (10/1/2003 to 9/30/2014) Modeled (Median, 25th, 75th)

Oct Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep0

2

4600

700

800

900

(in)

4

6

8300

400

500

600

thly

R

ain

fall

Flo

w (

cfs

)

8

10

120

100

200

300

Mo

nt

10 11 12 1 2 3 4 5 6 7 8 9

Month

Slide 35


Observed Flow Duration (10/1/2003 to 9/30/2014 )

Modeled Flow Duration (10/1/2003 to 9/30/2014 )


1000

10000

s)( )

100

1000

rag

e F

low

(cfs

10Da

ily A

ve

10% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Percent of Time that Flow is Equaled or Exceeded

Slide 36

Percent of Time that Flow is Equaled or Exceeded

Observed Flow Volume (10/1/2003 to 9/30/2014 )

Normalized Cumulative Flow

100%

120%

10

0%

)Modeled Flow Volume (10/1/2003 to 9/30/2014 )

80%

100%

(Ob

se

rve

d a

s 1

40%

60%

Flo

w V

olu

me

(

0%

20%

Oct-03 Apr-05 Oct-06 Apr-08 Oct-09 Apr-11 Oct-12 Apr-14

No

rma

lize

d

Slide 37

HSPF Simulated Flow Observed Flow Gage

REACH OUTFLOW FROM DSN 262 S 14306030 Yaquina River near Chitwood, Oregon

11-Year Analysis Period: 10/1/2003 - 9/30/2014 Flow volumes are (inches/year) for upstream drainage area Manually Entered Data

Drainage Area (sq-mi): 71

Total Simulated In stream Flow: 45 92 Total Observed In stream Flow: 43 95Total Simulated In-stream Flow: 45.92 Total Observed In-stream Flow: 43.95

Total of simulated highest 10% flows: 22.53 Total of Observed highest 10% flows: 20.43Total of Simulated lowest 50% flows: 4.07 Total of Observed Lowest 50% flows: 3.41

Simulated Summer Flow Volume (months 7-9): 1.36 Observed Summer Flow Volume (7-9): 0.96( ) ( )Simulated Fall Flow Volume (months 10-12): 13.74 Observed Fall Flow Volume (10-12): 13.09Simulated Winter Flow Volume (months 1-3): 21.95 Observed Winter Flow Volume (1-3): 21.89Simulated Spring Flow Volume (months 4-6): 8.88 Observed Spring Flow Volume (4-6): 8.02

Total Simulated Storm Volume: 20.69 Total Observed Storm Volume: 14.11Si l t d S St V l (7 9) 0 30 Ob d S St V l (7 9) 0 15Simulated Summer Storm Volume (7-9): 0.30 Observed Summer Storm Volume (7-9): 0.15

Errors (Simulated-Observed) Error Statistics Recommended Criteria

Error in total volume: 4.48 10Error in 50% lowest flows: 19.36 10Error in 10% highest flows: 10.27 15Seasonal volume error - Summer: 41.02 30Seasonal volume error - Fall: 4.96 30Seasonal volume error - Winter: 0.27 30Seasonal volume error - Spring: 10.78 30Error in storm volumes: 46.55 20Error in summer storm volumes: 105 65 50

>> Clear

Slide 38

Error in summer storm volumes: 105.65 50Nash-Sutcliffe Coefficient of Efficiency, E: 0.643Baseline adjusted coefficient (Garrick), E': 0.611 Monthly NSE 0.933

Model accuracy increases

Calibration Performance Measures

Flow Statistic Observed Modeled Percent ErrorQTotal (in) 44.0 45.92 4.8%Qsmr (in) 1.0 1.36 41.0%Qwtr (in) 21.9 22.0 0.27%Q t (in) 14 1 20 69 46 6%Qstm (in) 14.1 20.69 46.6%

Slide 39

Next Steps

• Big Elk Creek– Finish uncertainty analysis for hydrology simulationFinish uncertainty analysis for hydrology simulation– Conduct bacteria modeling– Continue compiling report

• Upper Yaquina River on hold• Compile LDC and Beach reports

Slide 40

Discussion

Slide 41Slide 41

Mid-Coast Bacteria TWG 17th Meeting (final) - oregon.gov · Mid-Coast Bacteria TMDL: Seventeenth...

Documents

Transcript of Mid-Coast Bacteria TWG 17th Meeting (final) - oregon.gov · Mid-Coast Bacteria TMDL: Seventeenth...