SCENARIO ANALYSIS OF FREMONT WEIR NOTCH · PDF fileerdc tr l-17-t scenario analysis of fremont weir notch – integration of engineering designs, telemetry, and flow fields d r david

ER

DC

TR

/E

L-1

7-D

RA

FT

SCENARIO ANALYSIS OF FREMONT WEIR NOTCH – INTEGRATION OF

ENGINEERING DESIGNS, TELEMETRY, AND FLOW FIELDS

En

gin

ee

r R

es

ea

rch

an

d

De

ve

lop

me

nt

Ce

nte

r

David L. Smith, Tammy Threadgill, Yong Lai, Christa Woodley,

R. Andrew Goodwin, Josh Israel

February 2017

Approved for public release; distribution is unlimited.

1

The US Army Engineer Research and Development Center (ERDC) solves the

nation’s toughest engineering and environmental challenges. ERDC develops innovative

solutions in civil and military engineering, geospatial sciences, water resources, and

environmental sciences for the Army, the Department of Defense, civilian agencies, and

our nation’s public good. Find out more at www.erdc.usace.army.mil.

To search for other technical reports published by ERDC, visit the ERDC online library

at http://acwc.sdp.sirsi.net/client/default.

http://www.erdc.usace.army.mil/

http://acwc.sdp.sirsi.net/client/default

Program Title ERDC EL/TR-17-DRAFT

February 2017

SCENARIO ANALYSIS OF FREMONT WEIR NOTCH – INTEGRATION OF

ENGINEERING DESIGNS, TELEMETRY, AND FLOW FIELDS

David L. Smith

Environmental Laboratory

US Army Engineer Research and Development Center

3909 Halls Ferry Road

Vicksburg, MS 39180-6199

others

Final report

Approved for public release; distribution is unlimited.

Prepared for US Army Corps of Engineers

Washington, DC 20314-1000

Under Project <####>, “Project Title”

Monitored by Lead ERDC Laboratory [FOR CRs ONLY]

US Army Engineer Research and Development Center

Street, City, ST ZIP-FOUR

ERDC/EL TR-17-DRAFT ii

Abstract

[Insert a 200-word summary here and use exactly the same one in Block 2

14 of the SF 298.] 3

The United States Bureau of Reclamation and the California Department 4

of Water Resources are planning a notch in the Fremont Weir on the Sac-5

ramento River. The notch is intended to provide access to the Yolo Bypass 6

floodplain for juvenile salmon across a range of flows. This study estimat-7

ed the entrainment rate of 12 separate notch scenarios. Entrainment es-8

timates vary from approximately 1 to 25%. Across all scenarios larger 9

notch flows entrain greater fish numbers, although not proportionally to 10

the volume through the notch. West located notches entrain more fish 11

than central and east and intake perform better than shelfs. However, in-12

takes and shelfs both performed poorly, regardless of notch flows, when 13

intake channels were angled from the mainstem. Entrainment estimates 14

are comparable to measured entrainment rates elsewhere in the Sacra-15

mento River suggesting that the modeled estimates are reasonable. The 16

results further suggest that the approach used is valuable for incorporating 17

structural modifications and evaluating expected outcomes. 18

DISCLAIMER: The contents of this report are not to be used for advertising, publication, or promotional purposes.

Citation of trade names does not constitute an official endorsement or approval of the use of such commercial products.

All product names and trademarks cited are the property of their respective owners. The findings of this report are not to

be construed as an official Department of the Army position unless so designated by other authorized documents.

DESTROY THIS REPORT WHEN NO LONGER NEEDED. DO NOT RETURN IT TO THE ORIGINATOR.

ERDC/EL TR-17-DRAFT iii

Contents

Abstract .................................................................................................................................... ii 19

Figures and Tables .................................................................................................................. iv 20

Unit Conversion Factors .......................................................................................................... v 21

1 Introduction ...................................................................................................................... 1 22

1.1 Fremont Weir...................................................................................................... 3 23

2 Goals and Objectives ....................................................................................................... 5 24

3 Scenario Descriptions and Domain Development ....................................................... 6 25

3.1 Scenarios ........................................................................................................... 6 26 3.2 Domain development ........................................................................................ 7 27

4 Study Design and Model Application............................................................................. 9 28

4.1 Fish telemetry .................................................................................................. 10 29 4.2 2D hydraulic models ........................................................................................11 30 4.3 ELAM description ............................................................................................ 23 31

4.3.1 Movement .................................................................................................................. 23 32 4.4 Fish movement modeling procedure .............................................................. 24 33

5 Results ............................................................................................................................. 26 34

5.1 Spatial distribution ......................................................................................... 26 35 5.2 Kernel density estimates ............................................................................... 26 36 5.3 Speed Estimates ............................................................................................. 28 37 5.4 Entrainment across all scenarios .................................................................. 28 38 5.5 Flow and entrainment relationships for Scenario 1 and 2 .......................... 30 39

6 Discussion ....................................................................................................................... 38 40

6.1 Distribution and speed comparisons ................. Error! Bookmark not defined. 41 6.2 Notch style ........................................................... Error! Bookmark not defined. 42 6.3 Notch location ...................................................... Error! Bookmark not defined. 43 6.4 Notch flow ........................................................................................................ 41 44 6.5 Entrainment trends across flows ........................ Error! Bookmark not defined. 45 6.6 Unknown factors that influence entrainment ............................................... 42 46 6.7 Background .......................................................... Error! Bookmark not defined. 47 6.8 Fremont Weir........................................................ Error! Bookmark not defined. 48

7 Bibliography .................................................................................................................... 43 49

Report Documentation Page 50

51

ERDC/EL TR-17-DRAFT iv

Figures and Tables

Figures 52

Figure 1. Map of project site......................................................................................................... 4 53

Figure 2. Scenario notch locations .............................................................................................. 7 54

Figure 3. Workflow for development of fish movement model ................................................ 9 55

Figure 4. Detection array at Fremont Weir ............................................................................... 10 56

Figure 5. Measured and Modeled Fish Locations ................................................................... 26 57

Figure 6. Contour lines showing the density Speed estimates ............................................. 27 58

Figure 7. Box plot of fish speed .................................................................................................. 28 59

Figure 8. Scenario vs Entrainment ............................................................................................ 30 60

Figure 9. Scenarios 1 and 2 ....................................................................................................... 31 61

Tables 62

Table 1 Physical properties of modeled scenarios.................................................................... 8 63

Preface 64

This study was conducted for the United States Bureau of Reclamation us-65

ing Interagency Agreement xxxx. The technical monitor was Dr. Patrick 66

Deliman. 67

The work was performed by the Water Quality and Contaminant Modeling 68

Branch (EPW) of the Environmental Processes and Effects Division (EP), 69

US Army Engineer Research and Development Center, Environmental La-70

boratory. At the time of publication, Mark Noel was Acting Chief, CEERD-71

EPW; Warren Lorenz was Chief, CEERD-EP; and Pat Deliman, CEERD-72

XX-X was the Technical Director. The Deputy Director of ERDC-EL was 73

Dr. Jack Davis and the Director was Dr. Beth Fleming. 74

COL Bryan Green was the Commander of ERDC, and Dr. David Pittman 75

was the Director. 76

ERDC/EL TR-17-DRAFT v

Unit Conversion Factors

Multiply By To Obtain

acres 4,046.873 square meters

acre-feet 1,233.5 cubic meters

angstroms 0.1 nanometers

atmosphere (standard) 101.325 kilopascals

bars 100 kilopascals

British thermal units (International Table) 1,055.056 joules

centipoises 0.001 pascal seconds

centistokes 1.0 E-06 square meters per second

cubic feet 0.02831685 cubic meters

cubic inches 1.6387064 E-05 cubic meters

cubic yards 0.7645549 cubic meters

degrees (angle) 0.01745329 radians

degrees Fahrenheit (F-32)/1.8 degrees Celsius

fathoms 1.8288 meters

feet 0.3048 meters

foot-pounds force 1.355818 joules

gallons (US liquid) 3.785412 E-03 cubic meters

hectares 1.0 E+04 square meters

horsepower (550 foot-pounds force per second) 745.6999 watts

inches 0.0254 meters

inch-pounds (force) 0.1129848 newton meters

kilotons (nuclear equivalent of TNT) 4.184 terajoules

knots 0.5144444 meters per second

microinches 0.0254 micrometers

microns 1.0 E-06 meters

miles (nautical) 1,852 meters

miles (US statute) 1,609.347 meters

miles per hour 0.44704 meters per second

mils 0.0254 millimeters

ounces (mass) 0.02834952 kilograms

ounces (US fluid) 2.957353 E-05 cubic meters

pints (US liquid) 4.73176 E-04 cubic meters

ERDC/EL TR-17-DRAFT vi

Multiply By To Obtain

pints (US liquid) 0.473176 liters

pounds (force) 4.448222 newtons

pounds (force) per foot 14.59390 newtons per meter

pounds (force) per inch 175.1268 newtons per meter

pounds (force) per square foot 47.88026 pascals

pounds (force) per square inch 6.894757 kilopascals

pounds (mass) 0.45359237 kilograms

pounds (mass) per cubic foot 16.01846 kilograms per cubic meter

pounds (mass) per cubic inch 2.757990 E+04 kilograms per cubic meter

pounds (mass) per square foot 4.882428 kilograms per square meter

pounds (mass) per square yard 0.542492 kilograms per square meter

quarts (US liquid) 9.463529 E-04 cubic meters

slugs 14.59390 kilograms

square feet 0.09290304 square meters

square inches 6.4516 E-04 square meters

square miles 2.589998 E+06 square meters

square yards 0.8361274 square meters

tons (force) 8,896.443 newtons

tons (force) per square foot 95.76052 kilopascals

tons (long) per cubic yard 1,328.939 kilograms per cubic meter

tons (nuclear equivalent of TNT) 4.184 E+09 joules

tons (2,000 pounds, mass) 907.1847 kilograms

tons (2,000 pounds, mass) per square foot 9,764.856 kilograms per square meter

yards 0.9144 meters

[AUTHOR: delete all units that are not used in your report.]

ERDC/EL TR-17-DRAFT 1

1 Introduction

As California’s largest river, the Sacramento is an important economic, 77

recreational, and ecological resource. The river has an extensive flood 78

control infrastructure that includes a system of dams, levees, and flood-79

ways intended to protect agricultural and urban regions. In particular, the 80

metropolitan area of Sacramento with some 2 million residents is protect-81

ed from flooding from this system. Protection is due to levees but flood 82

events are conveyed out of the river channels and onto floodways such as 83

the Yolo Bypass. The floodways are historic floodplains that when inun-84

dated protect the city. The floodways also receive sediment, nutrients, and 85

fish. [1]. The flood protection system has also impacted ecosystem pro-86

cess including those associated with floodplain access by fish. 87

The Yolo Bypass is a 24,000 ha basin protected by levees and inundated 88

during high flow on the Sacramento River. The floodway is 61 km long 89

and is flooded approximately 7 out of 10 years with a peak flow of 14,000 90

m3/s. Water is conveyed over the Fremont Weir downstream from Knights 91

Landing and just upstream from the confluence with the Feather River. 92

[2]. 93

The Fremont Weir was constructed in 1924 by the USACE. It is the first 94

overflow structure on the river's right bank and its two-mile overall length 95

marks the beginning of the Yolo Bypass. It is located about 15 miles 96

northwest of Sacramento and eight miles northeast of Woodland. South of 97

this latitude the Yolo Bypass conveys 80% of the system’s floodwaters 98

through Yolo and Solano Counties until it connects to the Sacramento Riv-99

er a few miles upstream of Rio Vista. The weir’s primary purpose is to re-100

lease overflow waters of the Sacramento River, Sutter Bypass, and the 101

Feather River into the Yolo Bypass. The crest elevation is approximately 102

32.0feet (NAVD88) and the project design capacity of the weir is 343,000 103

cfs. Adding the notch to this weir will change the amount of time that wa-104

ter flows over it and increase access to the floodplain for juvenile salmon. 105

On June 4, 2009, the National Marine Fisheries Service (NMFS) issued its 106

Biological Opinion and Conference Opinion on the Long-term Operation 107

of the Central Valley Project (CVP) and State Water Project (SWP) (NMFS 108

Operation BO). The NMFS Operation BO concluded that, if left un-109


changed, CVP and SWP operations were likely to jeopardize the continued 110

existence of four federally-listed anadromous fish species: Sacramento 111

River winter-run Chinook salmon, Central Valley spring-run Chinook 112

salmon, Central Valley steelhead (O. mykiss), and Southern Distinct Popu-113

lation Segment (DPS) North American green sturgeon (Acipenser 114

medirostris). The NMFS Operation BO sets forth Reasonable and Prudent 115

Alternative (RPA) actions that would allow continuing SWP and CVP op-116

erations to remain in compliance with the federal Endangered Species Act 117

(ESA). These include restoration of floodplain rearing habitat, through a 118

“notched” channel that increases seasonal inundation within the lower 119

Sacramento River basin. A significant component of these risk reduction 120

actions is lowering a section of the Fremont Weir (Figure 1) to allow juve-121

nile fish to enter the bypass and adult fish to more easily ascend this haz-122

ard. Questions remain on the details of notch implementation (size, 123

location), fish entrainment efficiency, and species-specific and ontology-124

based behaviors. 125

Among actions being considered are alternatives to “increase inundation 126

of publicly and privately owned suitable acreage within the Yolo Bypass.” 127

During inundation, the Yolo Bypass has been shown to have beneficial ef-128

fects on growth of juvenile salmonids (Sommer et al. 2001) due to the fa-129

vorable rearing conditions (e.g., increased primary productivity, relatively 130

slow water velocities, abundant invertebrates). In order to increase inun-131

dation within the Yolo Bypass, a notch of some configuration will need to 132

be constructed in the vicinity of Fremont Weir. Entrainment of juvenile 133

salmonids into the bypass routes them around the Delta, thereby minimiz-134

ing the potential for entrainment by the pumps at the State Water Project 135

and Central Valley Project. Therefore, maximizing entrainment into the 136

bypass, particularly at lower stages, is of particular interest. Uncertainty 137

exists about how the location, approach channel, and notch design and 138

setting influence the effectiveness for entraining juvenile salmonids from 139

the Sacramento River onto the Yolo Bypass. 140

It is generalized recognized that fish are unevenly distributed across a 141

channel cross section and that the position of the fish influences the prob-142

ability that entrainment occurs [3]. The distribution of fish is in part re-143

lated to secondary circulations which tend to concentrate passive particles 144

such as sediment away from the channel margins and towards the bank of 145

long radius of a river bend. This conceptual model is often applied to 146

downstream movement of fish such as juvenile salmon in the Sacramento 147


River. Notch entrainment efficiency is potentially improved by placing the 148

notch where fish density is maximized along the outside bend. Of course, 149

the specifics of the fish distribution are related to the unique attributes of 150

each cross section, notch design and the behavior of fish therein. The effi-151

ciency of an entrainment channel is most important factor impacting fish 152

benefits based on the Fishery Benefit Model (Cramer Fish Science, in 153

prep). 154

In 2015, two-dimensional (2-D) positions were measured for hatchery late 155

fall and winter run Chinook along a portion of the Fremont Weir. These 156

tracks provided the basis for this study. The objective of this study was to 157

validate an existing fish behavior model for use on this project, simulate a 158

range of alternate notch designs, and evaluate the sensitivity on entrain-159

ment to different locations and designs. Additionally, this modeling ap-160

proach allowed for exploration of different hypotheses regarding fish 161

behavior and the influence they could have on movement and entrainment 162

through the simulated notches. These results are useful for evaluating and 163

comparing probable outcomes of different notches. 164

1.1 Fremont Weir

Fremont Weir is a 1.8-mile long flood control structure designed with a 165

concrete, energy-dissipating splash basin, which minimizes scouring dur-166

ing overtopping events at the weir. The splash basin lies just downstream 167

of the crest of the weir and spans the full length of the weir. 168

When the river stage is sufficiently higher than the weir, all juvenile salm-169

onids are hypothesized to enter the bypass due to the overwhelming extent 170

of Sacramento River flows being pushed out of the channel and onto the 171

bypass. It is also hypothesized that during lower-stage overtopping events, 172

when the Sacramento River is just barely above the crest of Fremont Weir, 173

this effect is also the predominant cause of entrainment of Sacramento 174

River fish onto the bypass. Overtopping events can vary in duration from 175

just a few hours to several weeks, but are relatively short-lived compared 176

with the resulting flooded footprint of the Yolo Bypass, which persists fol-177

lowing the overtopping events. This footprint is a result not just of over-178

topping at the Fremont Weir, but substantial out-of-channel flows from 179

four westside tributaries: Knights Landing Ridge Cut, Cache Creek, Willow 180

Slough, and Putah Creek. 181


As part of Action I.6.1, inundation flows from the Sacramento River onto 182

the Yolo Bypass will occur at river flows lower than when the weir is over-183

topped, while species of interest are migrating past the Fremont Weir 184

reach towards the Delta. It is during this period that the action aims to in-185

crease entrainment of salmonids. Acierto et al. (2015) evaluated the po-186

tential for entrainment based on proportion of flow entering the bypass 187

and identified that it was potentially limited. Uncertainty exists about how 188

fish utilize the channel for migration and rearing and their relationship to 189

cross-channel flow patterns and secondary circulations and this study 190

evaluates how these bathymetric and hydraulic structures may influence 191

fish entrainment and flow relationships. 192

Fremont Weir modifications. As part of Action I.6.1, Fremont Weir will be 193

modified to allow seasonal, partial floodplain inundation in order to pro-194

vide increased habitat for salmonid rearing and to improve fish passage. 195

The same physical feature used for floodplain inundation flows will be 196

used for juvenile fish entrainment. The primary modification of Fremont 197

Weir will add a notch with one or more bays. 198

Figure 1. Map of project site. 199


2 Goals and Objectives

This study analyzes 12 notch scenario in the Fremont weir in terms of entrainment 200

of juvenile salmon. The goal is to quantify the relative entrainment rates (be-201

tween 0 and 1) across the suite of scenarios and to identify possible strategies for 202

enhancing entrainment outcomes. This study does not predict future entrainment 203

as models generally don’t predict future outcomes so much as highlight trends. 204

As there is no notch yet built, predictions of absolute entrainment rates risk miss-205

ing any number of unforeseen variables driving the movement of complex ani-206

mals like salmon in riverine systems. In a planning context, relative changes 207

across scenarios are an accepted standard practice. The outcomes of this study 208

will be one factor of the overall decision on which alternative is most suited for 209

meeting the larger project objectives. Once the notch is constructed, follow on 210

evaluation studies will provide the opportunity for additional calibration and veri-211

fication of model output. 212

The objectives of this study necessary to meet the goals include the following: 213

Develop a base fish movement data set under existing conditions (no 214

notch). This work was completed as part of Steel et al (2016). 215

Develop a calibrated three dimensional and two dimensional time varying 216

hydrodynamic model of the project reach. This work was completed as 217

part of Lai (2016). 218

Integrate engineering designs of proposed notches into existing bathyme-219

try and landscape (LiDAR) data capturing important differences in loca-220

tion, widths, invert elevations, and construction techniques. 221

Develop two dimensional flow fields for each of the scenarios that capture 222

the hydraulic impacts of each unique notch. 223

Calibrate a fish movement model using data from Steele et al (2016) and 224

Lai (2016). 225

Apply the calibrated fish movement model to the flow fields produced by 226

each scenario and summarize relative entrainment rates between 0 and 1. 227

Make recommendations on next steps and possible improvements. 228


3 Scenario Descriptions and Domain

Development

3.1 Scenarios

A suite of twelve notch scenarios was developed by the California Department of 229

Water Resources (DWR) and the United States Bureau of Reclamation (USBR). 230

The scenarios fall into two broad categories: 1) those with an extensive shelf ad-231

jacent to the notch and 2) those with a narrow channel or intake leading to the 232

notch headworks. The headworks are where fish will exit the Sacramento River 233

and enter the Yolo Bypass. The shelf based scenarios have a larger project foot-234

print than does the intake based scenarios. The primary purpose of the headworks 235

for the shelf and intake configurations is to create a hydraulic connection between 236

the Sacramento River and the Yolo Bypass during lower flows in the Sacramento 237

River than currently exists. The headworks will consist of the inlet transition, the 238

control structure, and the outlet transition, and will control the diversion of flow 239

(up to about 12,000 cfs) from the River into the Yolo Bypass. 240

Scenario notch locations are concentrated in the west, central, and east portion of 241

the Fremont Weir (Figure 2). Table 1 highlights the dimensions captured in the 242

landscape model of each scenario. Each scenario is different in terms of size, lo-243

cation, notch invert elevations, and width. These differences are translated into 244

the 2D simulation of the flow field which, in turn, translates into simulated fish 245

movement. 246


Figure 2. Scenario notch locations 247

3.2 Domain development

An IGES (initial graphic exchange specification) file was received from the 248

USBR for each of the scenarios. Upon receipt of these files, each file was loaded 249

into Capstone and an STL (stereolithography) file was created of the intake area. 250

Once the intake area had a mesh associated with it, the original STL file of the 251

river and intake STL file were then merged to create one mesh that represented 252

the mesh used for the scenario. The STL was exported as a 2dm file using Para-253

view and extraneous faces were removed from the dataset or modified to best 254

work with SRH-2D. 255

Fremont weir

1, 2, 3, 4, 9

5, 10, 10B

6, 7, 8, 9

11, 12


Table 1 Physical properties of modeled scenarios. Notch/River is the ratio of notch 256 flow to river flow 257

Scenario Lower Intake Upper Intake # of

Points # of El-ements

Notch Flow (cfs)

River Flow (cfs)

Notch/ River

Elevation Width Elevation Width

Original NA NA NA NA

Scenario 1

14 ft 31 ft 20 ft 44 ft 31200 33924 6000.22 42202.51 0.14

Scenario 2

14 ft 32 ft 20 ft 44 ft 33427 36126 6000.22 42202.51 0.14

Scenario 3

17 ft 21 ft 23 ft 24 ft 32858 35596 3000.11 42202.51 0.07

Scenario 4

22 ft 14 ft NA NA 32913 35782 1105.75 48289.31 0.02

Scenario 5

14 ft 31 ft 20 ft 41 ft 31308 33702 5981.18 42202.51 0.14

Scenario 6

14 ft 32 ft 20 ft 43 ft 29238 32313 5952.99 44843.49 0.13

Scenario 7

14 ft 33 ft 20 ft 44 ft 37538 40628 6000.22 47957.43 0.13

Scenario 8

17 ft 21 ft 23 ft 25 ft 31115 33941 3000.11 47029.93 0.06

Scenario 9 – West

17 ft 21 ft 23 ft 37 ft

38372 41453

3000.11 47029.93 0.06

Scenario 9 – East

17 ft 21 ft 23 ft 25 ft 3000.11 47029.93 0.06

Scenario 10 – West (A/B)

14 ft 33 ft 17 ft 35 ft 42119 45016 480.91 30809.31 0.02

Scenario 10 – Central (C)

18 ft 142 ft - - 42119 45016 2379.52 30809.31 0.07

Scenario 10 – East (D)

21 ft 146 ft - - 42119 45016 542.32 30809.31 0.02

Scenario 11

16 ft 220 ft - - 34037 36504 12077.32 44843.49 0.27

Scenario 12

16 ft 40 ft 20 ft 60 ft 33288 35711 6105.22 47029.93 0.13


4 Study Design and Model Application

Developing a fish movement model to assist with scenario evaluation for the 258

Fremont weir notch requires integration of data and information from several 259

sources and professional disciplines (Figure 3). The report used biological data 260

from a telemetry study, hydrodynamic data and models, and landscape modeling 261

techniques. 262

Figure 3. Workflow for development of fish movement model. SOG is speed over 263 ground. 264

265

266

Telemetry

SOG, distributions

Measure WRC LFC movement at

project site

CFD Demonstrate model can

simulate Fremont weir

flow fields

Scenario development

12 notch scenarios

Integrate with CFD domain

Fish movement model of Fremont

weir site, calibrate to measured fish

movement data

Fish movement model for scenario

notches – relative entrainment es-

timates

CFD model of all twelve

notches/scenarios

Biological

data Hydraulic/bathymetric

data

Landscape

data modifi-

cation


4.1 Fish telemetry

In 2015, 250 winter run Chinook (mean fork length of 103 mm) from Livingston 267

Stone Hatchery and 250 late fall run Chinook (mean fork length of 145 mm) from 268

Coleman National fish hatchery were tagged with acoustic tags and released 269

through a detection area at Fremont weir. The array was in a long sweeping bend 270

located at the head of the upstream end of the Fremont weir. This location was 271

thought to have the best conditions for redistributing fish to the outside bend 272

where susceptibility to entrainment by a future notch would be higher. All fish 273

were released over 24 hour periods at Knights Landing. River discharge was low 274

and stable with gage readings at Fremont weir of approximately 14 ft and flows of 275

approximately 5700 cfs. Analysis suggested little difference in movement be-276

tween winter run Chinook and late fall run Chinook at Fremont weir. Speeds over 277

grounds and size were not statistically different for winter and late fall run Chi-278

nook. The combined mean speed over ground was 0.67 m/s. 279

Cross-channel spatial distributions were also similar for winter and late fall run 280

Chinook. There was a moderate shift in the spatial distribution to the outside 281

bend of approximately 5 to 8 m away from the channel center. Chanel width is 282

approximately 70 m with the centerline, therefore 35 m away from either bank. 283

Figure 4. Detection array at Fremont Weir 284

For more detail please refer to Steele et al (2015) describes in detail the telemetry 285

study that was completed to support work described in this report. 286


4.2 2D hydraulic models and landscape modeling

SRH-2D is a 2D depth-averaged hydraulic and sediment transport model for river 287

systems. It was developed at the Technical Service Center, Bureau of Reclama-288

tion. The hydraulic flow modeling theory and user manual were documented by 289

Lai (2008; 2010). 290

SRH-2D adopts the arbitrarily shaped element method of Lai et al. (2003a, b), the 291

finite-volume discretization method, and an implicit integration scheme. The nu-292

merical procedure is very robust so SRH-2D can simulate simultaneously all flow 293

regimes (sub-, super-, and trans-critical flows) and both steady and unsteady 294

flows. A special wetting-drying algorithm makes the model very stable in han-295

dling flows over dry surfaces. The mobile-bed sediment transport theory has been 296

documented by Greimann et al. (2008), Lai and Greimann (2010), and Lai et al. 297

(2011). The mobile-bed module predicts vertical stream bed changes by tracking 298

multi-size, non-equilibrium sediment transport for suspended, mixed, and bed 299

loads, and for cohesive and non-cohesive sediments, and on granular, erodible 300

rock, or non-erodible beds. The effects of gravity and secondary flows on the sed-301

iment transport are accounted for by displacing the direction of the sediment 302

transport vector from that of the local depth-averaged flow vector. 303

Major capabilities of SRH-2D are listed below: 304

• 2D depth-averaged solution of the St. Venant equations (dynamic wave 305

equations) for flow hydraulics; 306

• An implicit solution scheme for solution robustness and efficiency; 307

• Hybrid mesh methodology which uses arbitrary mesh cell shapes. In 308

most applications, a combination of quadrilateral and triangular meshes 309

works the best; 310

• Steady or unsteady flows; 311

• All flow regimes simulated simultaneously: subcritical, supercritical, or 312

transcritical flows; 313

• Mobile bed modeling of alluvial rivers with a steady, quasi-unsteady, or 314

unsteady hydrograph. 315

• Non-cohesive or cohesive sediment transport; 316


• Non-equilibrium sediment transport; 317

• Multi-size sediment transport with bed sorting and armoring; 318

• A single sediment transport governing equation for both bed load, sus-319

pended load, and mixed load; 320

• Effects of gravity and secondary flows at curved bends; and 321

• Granular bed, erodible rock bed, or non-erodible bed. 322

SRH-2D is a 2D model, and it is particularly useful for problems where 2D ef-323

fects are important. Examples include flows with in-stream structures such as 324

weirs, diversion dams, release gates, coffer dams, etc.; bends and point bars; 325

perched rivers; and multi-channel systems. 2D models may also be needed if cer-326

tain hydraulic characteristics are important such as flow recirculation and eddy 327

patterns; lateral variations; flow overtopping banks and levees; differential flow 328

shears on river banks; and interaction between the main channel, vegetated areas 329

and floodplains. Some of the scenarios listed above may be modeled in 1D, but 330

additional empirical models and input parameters are needed and extra calibration 331

must be carried out with unknown accuracy. 332

The 2D model was built and calibrated for the same conditions under which fish 333

were released and their locations measured at Fremont Weir in 2015. This served 334

as the base case. Refer to Lai (2016) for model specifics. 335

We represented each of the twelve scenario notch designs by integrating basic 336

CAD designs into topography and bathymetry data. We used the Capstone soft-337

ware which is part of the DOD CREATE software suite. Capstone is a feature-338

rich application designed to produce analyzable representations of geometry for 339

use with physics based solvers. In particular the geometry, mesh and associative 340

attribution required for a computational simulation can be produced. 341

Geometry-related capabilities include: 342

• Geometry import and export for the IGES and STEP file formats 343

• Low-level geometry creation 344

• Edge and face splitting and merging 345


• Boolean operations 346

• Lofting, sweeping and extrusion 347

• Fillet and chamfer 348

• Various healing and stitching operations 349

Capstone excels at generating unstructured meshes for complex geometries. Due 350

to the robust topology model, high-quality meshes can be generated for the mani-351

fold and non-manifold geometries often required in aerospace applications. 352

Meshing-related capabilities include: 353

• Mesh import and export for common formats including STL, CGNS, SURF and 354

UGRID 355

• Mesh import and export for Create file formats including Kestrel (avm) 356

and Sentri (Exodus) 357

• Robust and flexible sizing field 358

• Robust unstructured surface mesh generation 359

• Unstructured tet-dominant volume mesh generation 360

• Extruded boundary layer generation via the third-party AFLR volume 361

mesher 362

• Sliding interfaces 363

• Mesh manipulation and repair operations 364

• Mesh export with associated attribution 365

One of the most important capabilities that Capstone provides is a framework for 366

attributing a mesh based on the underlying geometry. For supported output for-367

mats the mesh is exported with associated attributes to be used in a physics-based 368

analysis 369


By integrating the CAD designs with existing landscape data and then modeling 370

the 2D flow fields we captured the influence of notch details such as size, angle, 371

step heights and the subsequent influence the local flow field and thus fish distri-372

bution and potential for entrainment. 373

Each of the notch designs are represented in Figure 6. Flows through the notch 374

were represented using rating curves developed by the CA DWR. See Lia (2017, 375

under development) 376

4.3 Scenario descriptions

4.3.1 Scenario 1 West 377

This scenario is located past the west end of the Fremont Weir. It has a 378

minimum invert of 14 feet and a maximum flow of 6000 cfs. A broad shelf 379

starts from the river and tapers toward the notch structure. The location is 380

coincident with the Steele (2016) fish movement study location. 381

382


This scenario is located past the west end of Fremont Weir. It has a mini-384

mum invert of 14 feet and a maximum flow of 6000 cfs. A narrow intake 385

channel starts from the river and leads toward the notch structure. Com-386

paring Scenario 1 and 2 allows for direct evaluation of the shelf versus in-387

take approach. The location is coincident with the Steele (2016) fish 388

movement study location. 389



minimum invert of 17 feet and a maximum flow of 3000 cfs. A broad shelf 392

starts from the river and tapers toward the notch structure. Scenario 3 is 393

most comparable to Scenario 1 with the exception of the minimum invert 394

height. In addition, Scenario 3 and Scenario 1 have different rating curves 395

leading to different notch flows at similar stages (Figure 5). The location is 396

coincident with the Steele (2016) fish movement study location. 397



minimum invert of 22 feet and a maximum flow of 1,106. A broad shelf 400



placed in a similar location to Scenario 1, 2 and 3. It is distinct because of 402

the high minimum invert elevation and low maximum flow. Scenario 4 403

represents the smallest scenario in terms of concrete. 404

Figure 5. rating curves for notches 405

(1) For Scenario 1, 2, 5, 6, 7 (2) For Scenario 3, 8, 9

(3) Scenario 4 (4) Scenario 10

(5) Scenario 11 (6) Scenario 12


4.3.5 Scenario 5 Central 406

This scenario is in the central portion of the Fremont Weir located past the 407

west end of the Fremont Weir. It has a minimum invert of 14? feet and a 408

maximum flow of 6000 cfs. A broad shelf starts from the river and tapers 409

toward the notch structure. Scenario 5 and Scenario 1 are similar in terms 410

of size, have the same rating curve (Figure 5) and therefore allow compari-411

son of the entrainment rate between the west and central positions. How-412

ever, fish movement data was not collected in the Scenario 5 location in 413

2015. This reach has some remnant pilings, revetment and may require 414

bank modification if constructed. 415

4.3.6 Scenario 6 East 416

This scenario is at the east portion of the Fremont Weir. It has a mini-417

mum invert of 14 feet and a maximum flow of 6000 cfs. A broad shelf 418


comparable to Scenario 1 in terms of terms of size, have the same rating 420

curve (Figure 5) and therefore allow comparison of the entrainment rate 421

between the west and east positions. 422


This scenario is in the east portion of the Fremont Weir. It has a mini-424

mum invert of 14 feet and a maximum flow of 6,000 cfs. A narrow intake 425

channel broad shelf starts from the river and leads toward the notch struc-426

ture. Scenario 7 is comparable to Scenario 6 and allows entrainment esti-427

mates between a shelf and intake style notch at the east location. In 428

addition, Scenario 7 is comparable to Scenario 2 in terms of terms of size, 429

have the same rating curve (Figure 5) and therefore allow comparison of 430

the entrainment rate between the west and central positions. However, 431

fish movement data was not collected in the Scenario 7 location. 432


This scenario is in the east portion of the Fremont Weir. It has a mini-434

mum invert of 17 feet and a maximum flow of 3000 cfs. A broad shelf 435

starts from the river and tapers toward the notch structure. Scenario 8 436

and Scenario 3 are comparable in terms of size and rating curves. 437


4.3.9 Scenario 9 East and West 438

This scenario has a structure located off of the west end of the Fremont 439

Weir and in the east portion of the Fremont Weir. The east and the west 440

structures are identical with minimum inverts of 17 feet and maximum 441

flows of 3000 cfs each for a total of 6000 cfs. Both structures have a broad 442

shelf that tapers to the notch. Scenario 9 has the same rating curves as 443

Scenario 3 and 8. 444

4.3.10 Scenario 10 and 10B Central 445

This scenario has a three structure cluster in the central portion of the 446

Fremont Weir. The structures combine to have a maximum flow of ap-447

proximately 3600 cfs. The structures have a range of minimum inverts of 448

14, 18 and 21 feet. The structures are connected to the river with a narrow 449

intake channel. Scenario 10B is structurally the same as 10 with some 450

modifications to the underlying bathymetry and landscape model. Scenar-451

io 10 and 10B are not readily comparable to other scenarios in terms of 452

size, invert elevations and rating curves. Scenario 10 is most comparable 453

to 10B and allows estimating entrainment as a function of terrain modifi-454

cation. 455

4.4 Scenario 11 West

Scenario 11 is located at the west end of Fremont Weir. Unlike Scenarios 1 456

through 4, which are set off the end of the Fremont weir, Scenario 11 457

placement is further downstream and intersects the Fremont weir struc-458

ture. An intake channel leads from the river to the structure. Scenario 11 459

has a minimum invert of 16 feet and a maximum flow of 12,000 cfs. It is 460

the largest structure in the study. 461

4.5 Scenario 12 West

Scenario 12 is located at the west end of Fremont Weir and like Scenario 11 462

intersects the Fremont weir structure. An intake channel leads from the 463

river to the structure. Scenario 12 has a minimum invert of 16 feet and a 464

maximum flow of 6,000 cfs. It is comparable to Scenario 1 in terms of size 465

but has a different rating curve. 466



Figure 6. Images of notches as modeled. 467

Scenario 1 – West - 6K – Shelf

Scenario 2 – West - 6K - Intake


Scenario 4 – West - 1K - Shelf


Scenario 5 – Central - 6K – Shelf

Scenario 6 – East - 6K - Shelf

Scenario 7 – East - 6K – Intake





Scenario 10 – Inundation - Central - 3K

Scenario 11 – Inundation - West - 12K


Scenario 12 – Inundation - West - 6K – Intake

468


4.6 ELAM description

The ELAM is a mechanistic representation of individual fish movement 469

which accounts for local hydraulic patterns represented in computational 470

fluid dynamic models (CFD) such as the 2D models developed for this pro-471

ject. Rule-based behaviors can be implemented within the model to drive 472

fish movement. The model is agent based providing a mathematical 473

means of representing the environment from the perspective of animal 474

perception. The approach is informed by observations of fish movement 475

such as was collected at Fremont weir (Steel et al. 2016) but individual 476

tracks are not directly modeled. Rather, statistical properties of the meas-477

ured tracks are used to guide model coefficient development. The ap-478

proach supports extension of empirical observations toward unmeasured 479

environmental conditions such as the wide scenario range evaluated as 480

part of this project. The ELAM is documented in a number of publications 481

(Appendix 1). 482

Hydrodynamic information generated at discrete points in the Eulerian 483

mesh is interpolated to locations anywhere within the physical domain 484

where fish may be. This conversion of information from the Eulerian mesh 485

to a Lagrangian framework allows the generation of directional sensory 486

inputs and movements in a reference framework similar to that perceived 487

by real fish. Movement is treated as a two-step process: first, the fish eval-488

uates agent attributes within the detection range of its sensory system and, 489

second, it executes a response to an agent by moving (Bian 2003). The 490

volume from which a fish acquires decision-making information is repre-491

sented as a 2-D sensory ovoid. A virtual fish’s sense of direction at each 492

time increment is based on its orientation at the beginning of the time in-493

crement. Directional sensory inputs are tracked relative to the horizontal 494

orientation of the fish because fish response to laterally-located versus 495

frontally-located stimuli can be different (Coombs et al. 2000). The senso-496

ry ovoid has a vertical reference because fish detect accelerations and grav-497

itation through the otolith of its inner ear (Paxton 2000). It also senses 498

three-dimensional information on motion (Braun and Coombs 2000). In 499

this individual-based model (IBM) a symmetrical (spherical) sensory 500

ovoid is used. 501

Movement 502

Two fish swim speeds were used, the drift velocity set at 0.25 BL/s and the 503

cruising velocity of 1.5 BL/s. Fish speed variability was induced by calcu-504


lating a random seed from a normal distribution centered on 0 with a 505

standard deviation of 1 termed RRR. Swim speed variability was simulat-506

ed by first calculating a deviation as 507

𝜎 = 𝑅𝑅𝑅 ∗ 𝐶𝑟𝑢𝑠𝑖𝑒 − 𝐷𝑟𝑖𝑓𝑡 508

where cruise is the cruising velocity and drift is the drift velocity. Next the 509

swim speed is computed as 510

𝑆𝑝𝑒𝑒𝑑 = 𝐵𝐿 ∗ (𝐶𝑟𝑢𝑠𝑖𝑒 + 𝜎) 511

Many behaviors can be implemented within the ELAM. For this study on-512

ly one behavior, a biased random walk in the downstream direction was 513

used. The 2015 Fremont weir fish movement data suggest no additional 514

behaviors are represented. 515

The Ornstein-Uhlenbeck (OU) process was used to simulate sensing and 516

orientation in the fish, i.e. how straight or variable a fish track composed 517

of multiple sequential points is. The process was implemented by first 518

calling a random seed from a wrapped uniform distribution. Two coeffi-519

cients, lamda_xy and c_xy are used to calibrate computed fish positions 520

using measured fish positions as a guide. Sensing describes the ability of 521

the fish to locate the proper swim direction. For example, lamda_xy = 1 522

would be perfect sensing ability and the fish would always know which 523

movement direction was correct. On the other hand, c_xy represents the 524

orientating ability with a value of 0 being perfect. 525

4.7 Fish movement modeling procedure

There were 13 separate hydraulic models representing the base condition 526

and 12 scenarios. The base condition matched the location, discharge and 527

stage under which late fall and winter run chinook were tagged and re-528

leased in 2105. Thus the base condition was used to calibrate the fish 529

movement model. The calibration was done using 2D depth averaged hy-530

draulic models. This was done in lieu of 3D hydraulic models for two rea-531

sons. First, the telemetry data is also 2D due to technology limitations of 532

the telemetry gear that was used. Second, since there were are twelve sce-533

narios to be considered, developing 3D models was time and cost prohibi-534

tive. Additional 3D models may be developed in the future if required for 535

particular questions. 536


For calibration, fish were released in the model at Knights Landing. A to-537

tal of 500 particles or fish were placed in a lateral cross section. The fish 538

length was set to the mean size of fish released as part of Steele (2016) 539

equaling 124 mm. No differentiation in the fish movement model is made 540

between late fall chinook and winter run chinook. Fish moved down-541

stream, passed through the Fremont weir reach, and exited the model at 542

Verona. 543

Fish movement model data was post processed to produce speed over 544

ground (SOG) and spatial distributions (kernel densities) using JMP 2012. 545

The estimates were compared to the measured data, adjustments made to 546

model parameters, and the model rerun until measured and computed 547

values were similar. The two coefficients lamda_xy and c_xy were adjust-548

ed to approximate the speed over ground and spatial distribution through 549

the project reach. Coefficient lamda_xy was set to 0.1 and c_xy was set to 550

2.0. Speed was insensitive and spatial distribution was sensitive to the pa-551

rameters. 552

The calibrated model was then run for the twelve proposed scenarios and 553

the proportion of fish entering the notch versus exiting the model domain 554

at Verona was computed. Ten to thirty runs each with 500 fish were com-555

pleted in order to estimate model variability. Each run was made with a 556

different random seed to start the model. Higher levels of variability were 557

possible by adjusting calibrated model parameters but results begin to dif-558

fer from measured results. Thus, for the final runs we only modified the 559

random seed. 560

Estimates of entrainment percentages for each scenario were made for the 561

maximum anticipated notch flow ranging from 1,000 to 12,000 cfs. Addi-562

tional analysis was done for Scenarios 1 and 2 representing an intake and 563

shelf style notch respectively. The analysis required running across a 564

range of anticipated notch flows and estimating the entrainment for each. 565

In addition, Scenario 10 and 10B involved three separate structures and a 566

complicated rating curve. Additional analysis for 10 and 10B across a 567

range of flows was also done. 568


5 Results

5.1 Spatial distribution

Spatial distribution was assessed qualitatively by overlying measured fish 569

positions from Steele et al (2015) with modeled fish tracks (Figure 5). 570

Tracks overlapped and have similar cross channel distributions. 571

Figure 7 Measured and Modeled Fish Locations 572

573

5.2 Kernel density estimates

Kernel densities for the measured and modeled fish distributions were cal-574

culated (Figure 5). Bivariate density estimation models a smooth surface 575

that describes how dense the data points are at each point in that surface. 576


The plot adds a set of contour lines showing the density (Figure 8). Op-577

tionally, the contour lines are quantile contours in 5% intervals with thick-578

er lines at the 10% quantile intervals. This means that about 5% of the 579

points are below the lowest contour, 10% are below the next contour, and 580

so forth. The highest contour has about 95% of the points below it. 581

Figure 8. Contour lines showing the density speed estimates for modeled (A) and 582 measured fish positions (B) 583

This nonparametric density method first requires dividing each axis into 584

50 binning intervals, for a total of 2,500 bins over the whole surface, 2) 585

counting the points in each bin, 3) decide the smoothing kernel standard 586

deviation (handled in JMP), 4) run a bivariate normal kernel smoother us-587

A

B


ing an FFT and inverse FFT to do the convolution, and 5) create a contour map 588

on the 2,500 bins using a bilinear surface patch model. 589

5.3 Speed Estimates

Speed over ground was computed for measured and modeled fish. Mod-590

eled fish estimates were based on 500 individual particles. Fish were re-591

leased at Knights Landing Bridge and exited the domain at Verona. The 592

resulting data set was subsampled to capture track data corresponding to 593

the measured fish position data. Fish speed was computed for each fish 594

and represented as a box plot (Figure 9). Modeled fish speed was 0.71 m/s 595

and measured fish speed was 0.67 m/s with arrange of 0 to 2.o m/s. 596

Figure 9. Box plot of fish speed for modeled (A) and measured (B) fish speed over 597 ground estimates. 598

5.4 Entrainment across all scenarios

Entrainment, as depicted in Figure 11, varied as a function of notch type 599

(intake versus shelf), location (west, central, or east weir) and notch flow 600

volume (cfs). Scenarios 1(intake) and 2 (shelf) had entrainment rates of 601

approximately 8% with Scenario 2 slightly superior to Scenario 1. Both 602

Scenarios 1 and 2 have a maximum notch flow of 6,000 cfs. In contrast, 603

Scenarios 3 and 4, while in the same location as Scenarios 1 and 2, have 604

entrainment estimates of approximately 5 and 1% respectively. However, 605

it is important to note that Scenario 3 and 4 have higher invert elevations 606

and lower notch flows when compared to Scenario 1 and Scenario 2 607

A B


Scenario 5 is located in the central portion of the Fremont Weir but is oth-608

erwise similar to Scenario 2. Scenario 5 entrains approximately 4%. Sce-609

nario 5 is the only single notch structure evaluated for the central Fremont 610

weir location. Scenarios 10 and 10B structures are in a similar location 611

and are described below. 612

Scenarios 6 through 8 are all located on the east portion of Fremont weir. 613

Scenarios 6 and 7 entrain approximately 5%, and Scenario 8 entrains ap-614

proximately 2%. Like Scenarios 1 and 2, scenarios 6 and 7 are a direct 615

comparison of an intake versus shelf. Like Scenarios 1 and 2, scenarios 6 616

and 7 have similar entrainment estimates. Compared to Scenarios 1 and 2, 617

scenarios 6 and 7 have lower entrainment estimates. Scenario 8 is directly 618

comparable to scenario 3 with the exception of its location on the east por-619

tion of Fremont weir. Both scenarios 3 and 8 have approximately 2% en-620

trainment. 621

Scenario 9 is a combination of scenarios 3 and 6 with one structure located 622

on the west portion and one located on the east portion of Fremont weir. 623

Scenario 9 has an approximately 2% entrainment rate similar to either 624

scenario 3 or scenario 6 alone. 625

626

Scenario 10 was similar to scenario 5, 6, and 7 at a flow of 3402 cfs. Sce-627

nario 10B was modified based on input provided by CA DWR (Figure 10). 628

Figure 10. Modifications completed for Scenario 10B based on

email from Josh Urias to David Smith, 12/2/2016


The modification required generating a new spatial model and running the 629

2D hydraulic model to produce the new flow fields. We attempted to cap-630

ture as much of the input as possible. We modified the bathymetry and 631

resloped the bank. We flattened the bathymerry signal from the existing 632

piles and we softened the edges of the intake structure to round them. The 633

resulting flow field and subsequent entrainment estimates were improved 634

over Scenario 10 with approximately 10% of the fish entrained at 3402 cfs. 635

Figure 11. Mean entrainment estimates for each scenario at maximum flow with 636 standard deviations. Scenario number is placed above each error bar. 637

638 Scenario 11, with the flow of 12,000 cfs entrained the greatest number of 639

fish at approximately 25%. Scenario 12 is comparable to Scenario 2 as 640

both are intake style notches. Entrainment rates for both are approxi-641

mately 7%. 642

5.5 Flow and entrainment relationships

For scenario 1 (intake) and scenario 2 (shelf) entrainment was modeled for 643

a range of flows to establish the notch entrainment trends over the range 644

of expected operating conditions. Scenarios 1 and 2 were chosen because 645

each is located in the reach were fish were tracked in 2015. The hydro-646

graph from the time period of December 1 to December 30 2015 was used 647

as it contained the low and high river flows (represented as stages from 648

Fremont weir gage) need to capture the full range of notch entrainment 649

and was also used for the base model. The figures are entrainment esti-650

mates for simulated fish for scenarios 1 and 2 at Fremont across a range of 651

10B

11

1 2

12

10 6 7

9 8

5

4

3


notch flows. Each data point is the mean entrainment rate at each notch 652

flow. Error bars are the standard deviation based on a minimum of 6 runs 653

of 500 fish each. Entrainment increases with notch flow for both but the 654

transition from low entrainment (~1%) versus high entrainment (~8%) is 655

slower for the shelf. Both scenarios entrain similar percentages of fish but 656

Scenario 1 (intake type notch) uses less water to achieve maximum en-657

trainment. 658

Figure 12. Scenarios 1 and 2 659

660

661


Scenario 3 through 12 lack error bars for the entrainment estimates because the 662

runs are not done. Error bars are expected to be similar to what has already been 663

reported. Scenario 3 entrains relatively few fish over the range of flows evaluat-664

ed with the trend suggesting maximum entrainment of approximately 1 to 2% 665

from 1500 to 3000 cfs. 666

Figure 13 667

668

Scenario 4 is the smallest in terms of flow and entrainment across flows remains 669

below 1%. 670


Figure 14. 671

672

Scenario 5 has a peak entrainment of approximately 5 % and reaches a plateau 673

near 5000 cfs. 674

Scenario 6 reaches a peak entrainment of approximately 10% at approximately 675

3000 cfs or half of the rated maximum notch flow. This appears to be related to 676

the interaction of the excavated bench and stage that tend diminish near bank re-677

circulation zones and promote direct streamlines along the bank. 678

Figure 15. 679

680


Scenario 7 entrains approximately 3 to 4% across a wide range of notch flows but 681

has more variability across flows than other scenarios. 682

Figure 16. 683

684

Scenario 8 entrains approximately 3 to 4% and the entrainment trend sug-685

gest that a entrainment plateau has not been reached. 686

Figure 17. 687

688


Scenario 9 entrains approximately 1% and the entrainment trend suggest 689

that a entrainment plateau has been reached. 690

Figure 18. 691

692

Scenario 10 and 10B represent a different notch design on comparison to 693

the other designs. Flows from 37.5 cfs to 3648 cfs (27.5, 54.9, 197.1, 499, 694

1043, 1363, 2098, 2521, 3039, 3358, 3402, 3648 cfs) were run incremen-695

tally for both Scenario 10 and 10B covering the range of flows dictated by 696

the rating curve. For both scenarios a flow of 3402 cfs maximized en-697

trainment. All other flows entrained less than 1% of fish. This is likely re-698

lated to the complicated bank and bathymetry at this location and a 699

recirculation zone that is established in the bend. 700


Figure 19. 701

702

Scenario 11 shows a strong increase in entrainment rates with notch flow 703

and even at the midpoint of flow of 6000 cfs is entraining approximately 704

15% of the fish and reaching a maximum of approximately 24% at 12000 705

cfs. Scenario 11 and 12 are located deeper into the bend than other west 706

scenarios and has a different design lacking a two-step weir and instead 707

relying on a single invert elevation. The width of the structure is wide 708

(220 ft) and it attracts a large cross section of streamlines from the river. 709

Figure 20. 710

711


Scenario 12 entrains approximately 5 % of the fish. The trend suggests a 712

plateau is reached at around 3000 cfs. 713

Figure 21. 714

715


Discussion

The ELAM was calibrated using fish telemetry data collected in 2015 716

(Steele et al. 2016) and the CFD simulations (Lia 2016). Once complete, 717

additional CFD runs were made for proposed notches that represented dif-718

ferent locations and notch designs. This allows the considerable 719

The broad pattern of entrainment across all scenarios finds that entrain-720

ments estimates vary from a low of approximately 1% to a high of approx-721

imately 25%. Ratio of entrainment flow to river flow correspondingly was 722

2 to 27%. These numbers broadly agree with several studies completed at 723

the Georgianna Slough junction with the Sacramento River. Perry et al. 724

(2014) measured the fraction of fish measured in 2011 entering Georgian-725

na Slough, which ranged from 1 to 30% with 20 to 30% entering when a 726

non-physical barrier was not operating. The flow split between Georgian-727

na Slough and the Sacramento River was approximately 20% during the 728

study period. Entrainment into Georgianna Slough is strongly dependent 729

on tides and flows. The 2011 year was dominated by high non-reversing 730

flows, conditions under which entrainment probabilities decline dramati-731

cally (Perry et al. 2015). Perry et al. (2015) summarized data from a wide 732

range of sources and estimated an entrainment probability from negative 733

to approximately 55% across a number of low flow years. The mean flow 734

ratio between Georgianna Slough and the Sacramento River was 22% with 735

a low of 15 and a high of -17% (more water going down Georgianna Slough 736

than the Sacramento River). Perry (2010) found mean daily flow ratios 737

between Georgianna Slough and the Sacramento River from 2007 to 2009 738

varied from approximately 30% to 80% and entrainment probabilities 30 739

to 55%. Finally, Cavallo et al. (2015) summarized data from Sacramento 740

River diversions (including Perry 2010) and concluded entrainment rates 741

varied from 10% to 60% with diversion ratios of approximately 18% to 742

60%. 743

We plotted summary data from Peery (2010) and Cavallo et al. (2015) with 744

the ELAM entrainment estimates to contextualize our findings (Figure 13). 745

The data suggest that our entrainment estimates trend well with meas-746

ured entrainment values within the Sacramento River. However, the di-747

version ratios proposed at the Fremont Weir notch are generally less than 748

the reported data. In addition the slope relating river diversion ratio to 749


entrainment differs with the ELAM estimates being the most sensitive to 750

river diversion ratio. However, the entrainment estimates we developed 751

overlap suggesting that the ELAM entrainment estimates are reasonable. 752

The Ferment weir notch scenarios differ from Georgianna Slough in im-753

portant ways. First, the proportion of water entrained varies from approx-754

imately 1% (Scenario 4) to 27% (Scenario 11). Only Scenario 11 approaches 755

the ratios of flow diverted at Georgianna Slough. The remainder are con-756

siderably less. Georgina Slough is also tidal and can be slow or even re-757

versed whereas the Fremont Weir reach is rarely less than 0.75 m/s. This 758

suggests the exposure time of a fish to the diversion point is less in the 759

Fremont Weir. Finally, cross channel distributions of fish in the Fremont 760

Weir reach and the nearby USACE test reach at river mile 85.6 and 43.7 761

are insensitive to discharge (Sandstrom et al 20xx, Singer et al. 20xx, Steel 762

et al 2016, Steele et al in prep, Woods et al. in prep) with most fish tending 763

toward center channel. In comparison, cross channel distributions at 764

Georgianna Slough vary with discharge and stage. Entrainment at any of 765

the Fremont Weir notches may not be as dynamic or of similar magnitude 766

as it is to Georgianna Slough. 767

Figure 13. Plot of ELAM estimates with comparable estimates from the Sacramento 768 River. Cavallo et al (2015) line estimated by pulling values from graph and thus is an 769

approximation. 1:1 line denotes when entrainment is proportional to entrainment 770 flow. 771

772


The difference in slope between the ELAM and the Georgianna Slough 773

may also be partially explained through differences in the river environ-774

ment. The Fremont weir is strongly advective and fish movement though 775

this reach reflects that. In comparison, the tidal junction at Georgianna 776

Slough induces upstream movement, station holding along the bank and 777

in general more complicated swim paths. Of the studies, Perry et al. 778

(2014) is the most comparable to the Fremont Weir because reversing 779

flows were rare. The ratio between Georgianna Slough and the Sacramen-780

to River was approximately 16% and entrainment was approximately 22% 781

when a non-physical barrier was not operating. This compares with a ratio 782

of 27%flow for 25% entrainment for Scenario 11 (the largest notch evaluat-783

ed). 784

We may underestimate entrainment in for scenarios 1,2, 3,4, 9 11 and 12 785

all located in the western portion of the notch. This is because the spatial 786

distributions of the modeled fish are deviate from the measured distribu-787

tion with the measured fish having a larger outside bend density. Broadly 788

the kernel density estimate overlap and agree but entrainment is sensitive 789

to lateral position in the channel. The difference is likely due not repre-790

senting secondary circulations in the 2D hydraulic model. We believe this 791

is acceptable because of the following reasons. First, developing 3D time 792

varying RANS simulations for all 12 alternatives was infeasible. Working 793

in 2D allowed all the spatial domains to be represented. Future design 794

work (as opposed to planning work) may need to consider 3D simulations. 795

Second, the bias introduced by the lateral distribution is equal across all 796

alternatives. Third, the ELAM estimates are comparable to other entrain-797

ment estimates from the Sacramento River suggesting whatever potential 798

underestimation we report is likely within the range of variation we expect 799

to see within existing measured entrainment data sets. 800

There are some additional caveats to this study as we presented model re-801

sults that will apply to future engineering design and analysis. 802

5.6 Accuracy and precision in planning studies

This study has provided entrainment estimates for a range of scenarios. 803

The results should be viewed cautiously for several reasons. First, there is 804

no fish entrainment data for any notch that was modeled. We simply cali-805

brated to existing conditions (Base scenario) and extended that calibration 806

to the 12 notch scenarios. Each notch scenario reported has an error bar 807

associated with it which captures the variability of the entrainment as 808


modified by varying ELAM boundary conditions slightly. Thus each sce-809

nario entrainment estimate is an ensemble estimate which is considered a 810

best practice for physical system numerical modeling. However, since the 811

real entrainment rate is unknown the raw estimates should not be viewed 812

as absolute numbers. Rather, the entrainment estimates should be used 813

as relative entrainment rates to highlight differences across scenarios. 814

This is consistent with USACE best practice. Future work should include 815

more detailed modeling and after construction measurement of notch per-816

formance. 817

5.7 Behavior

Fish have a near limitless level of behaviors that can be implemented and 818

our representation is inherently limited by incomplete understanding. 819

The behavior quantified in Steele (2015) was simple but undoubtedly other 820

behaviors which might influence movement were occurring but were not 821

measured. In addition, the notch will change the local environment and 822

expose fish to acceleration gradients in excess of what is found in the river. 823

Elevated acceleration gradients generally repel migrating juvenile salmon. 824

In addition, data and behavior for fry sized salmon are largely unavailable. 825

USACE studies (need refs) suggest very limited numbers of fry size salmon 826

near banks in this reach. Susceptibility of fry size salmon to the notch may 827

be greater than smolts or, if fry size fish are migrating similarly to parr and 828

smolts then entrainment estimates may correspond to results in this study. 829

5.8 Notch flow and design

Across all scenarios larger notch flows entrain greater fish numbers, alt-830

hough not proportionally to the volume through the notch. West located 831

notches entrain more fish than central and east and intake perform better 832

than shelfs. However, intakes and shelfs both performed poorly, regardless 833

of notch flows, when intake channels were angled from the mainstem. 834

A primary exception to notch flows being the most important design crite-835

ria is demonstrated with Scenario 10B. Scenario 10B was a late modifica-836

tion of Scenario 10 and those modifications improved notch performance. 837

These findings highlight the importance of hydrodynamics along the up-838

stream bank and angle of the intake off of the Sacramento River are im-839

portant design factors for optimizing fish entrainment. Addi-tionally, the 840

substantial biological response resulting from stakeholder-generated sce-841


nario design changes suggest this model can further analyze advance op-842

timization exercises and higher-order design drawings. 843

5.9 Unknown factors that influence entrainment

When a notch is constructed it may closely resemble the scenarios exam-844

ined in this study or it may deviate. We captures many details of each 845

scenario including structural changes and bankline, bathymetry and over-846

bank changes. As the design goes forward additional details will be added 847

and these details may begin to deviate from what was analyzed as part for 848

this study. 849

5.10 2D data in 3D river

Depth information for fish is unavailable. The measured positions there-850

fore are in 2D. Not having depth information induces uncertainty in the 851

measured positions. As fish move deeper, as may occur in the river bend, 852

the estimated path length measured in 2D diverges from the 3D path 853

length. This bias is inherent in the fish position data used for this study. 854

5.11 Impact of bank structures on secondary circulations

Secondary circulations are one factor driving the lateral distribution of fish 855

in the Sacramento River with the likely result of shifting fish positions to-856

ward the outside bank. When one of the scenarios is implemented and 857

constructed, we would expect that the existing secondary circulation pat-858

terns in the vicinity of the notch will change. For example, bend way weirs 859

are put along the outside bends of river expressly to disrupt secondary cir-860

culations. The end result may be that the constructed structure diminish-861

es the tendency of to skew lateral distributions to the outside bend. 862


6 Bibliography (being updated, issues with

reference manager)

Acierto. 2015. 863

Bian. 2003. 864

Bowman, a. and Foster, P. 1992. Density Based Exploration of Bivariate Data. Dept. of 865 Statistics, Univ. of Glasgow. 866

Coombs. 2000. 867

Coombs. 1999. 868

Coombs, Braums. 2000. 869

Foster, Bowman and. 1992. 870

Goodwin. 2001. 871

Greimann. 2008. 872

Greimann. 2010. 873

Jones, Marsh and. 1988. 874

lai. 2010. 875

Lai. 2016. 876

Lai. 2011. 877

Paxton. 2000. 878

Railsback. 1999. 879

Singer, Aalto, and James. 2008. 880

Sommer. 2001. 881

1993. "Statistics and computing ." 171-177. 882

Steel. 2015. 883

Steel. 2016. 884

Wu. 2000. 885

886


Appendix 1. ELAM documentation 887

THEORETICAL: 888

1. Nestler, J. M., R. A. Goodwin, and D. P. Loucks. 2000. Coupled Ecological 889 Modeling for Improved Water Resources and Ecosystem Management. 890 Proceedings of the ASCE 2000 Joint Conference on Water Resources En-891 gineering and Water Resources Planning and Management, Minneapolis, 892 MN 30 Jul- 1 Aug 2000. 893

2. Nestler, J. M., R. A. Goodwin, and L. Toney. 2002. Simulating movement 894 of highly mobile aquatic biota: foundation for population modeling in an 895 ecosystem context. Technical Note ERDC/EL TN-EMRRP-EM-02 Janu-896 ary 2002, U.S. Army Engineer Research and Development Center, Vicks-897 burg, MS 39180-6199. 898

3. Gustafson, E., J. Nestler, L. Gross, K. Reynolds, D. Yaussy, T. Maxwell, and 899 V. Dale. 2003. Evolving approaches and technologies to enhance the role 900 of ecological modeling in decision-making, ed., V. Dale. in Ecological 901 Modeling for Resource Management, Springer Verlag Publishers, New 902 York. 903

4. Nestler, J. M., R. A. Goodwin, and D. P. Loucks. 2005. Coupling of biolog-904 ical and engineering models for ecosystem analysis. ASCE Journal of Wa-905 ter Resources Planning and Management 131(2): 101-109. 906

5. Goodwin, R. A., D. L. Smith, J. M. Nestler, J. J. Anderson, L. J. Weber, and 907 R. L. Stockstill. 2006. Agent-based approach enhances conventional 908 aquatic habitat description and species utilization methods. In American 909 Society of Civil Engineers Proceedings of the World Environmental and 910 Water Resources Congress 2006: Examining the Confluence of Environ-911 mental and Water Concerns” held in Omaha, Nebraska, May 21-25 2006. 912 Pp 1-8. doi: 10.1061/40856(200)90 913

6. Goodwin, R. A., J. M. Nestler, J. J. Anderson, D. Peter Loucks. 2006. Fore-914 casting 3-D fish movement behavior using a Eulerian-Lagrangian-Agent 915 Method (ELAM). Ecological Modeling 192: 197-223. 916

7. Goodwin, R. A., Smith, D. L., Nestler, J. M., Anderson, J. J., Weber, L. J., 917 and Stockstill, R. L. 2006. Agent-based approach enhances conventional 918 aquatic habitat description and species utilization methods. Proceedings of 919 the World Environmental & Water Resources Congress, American Society 920 of Civil Engineers, 21 – 25 May 2006, Omaha, Nebraska. 921

8. Nestler, J. M., Goodwin, R. A., Smith, D. L., and Anderson, J. J. 2007. A 922 Mathematical and Conceptual Framework for Ecohydraulics,” In Wood, 923 P. J., D. M. Hannah, and J. P. Sadler, eds, Hydroecology and Ecohydrolo-924 gy: Past, Present, and Future, John Wiley & Sons, Ltd. pp 205-224. 925


9. Goodwin, R. A., Nestler, J. M., Anderson, J. J., and Weber, L. J. 2007. A 926 new tool to forecast fish movement and passage. HydroReview. 26(4), 58-927 71. 928

10. Nestler, J. M., R. A. Goodwin, J. J. Anderson, and D. L. Smith. 2007. Un-929 derstanding hydrodynamics from the fish’s point of view, Part II: Integrat-930 ing Flow Field Distortion, Sensory Biology, and Geomorphology. 931 Proceedings of the 6th International Symposium on Ecohydraulics, 18 - 23 932 February 2007, Christchurch, New Zealand 933

11. Goodwin, R. A., Nestler, J. M., Anderson, J. J., and Cheng, J.-R. 2007. Un-934 derstanding hydrodynamics from the fish’s point of view, Part I: Integrat-935 ing CFD modeling, individual movement, and spatial/cognitive ecology. 936 Proceedings of the 6th International Symposium on Ecohydraulics, 18 - 23 937 February 2007, Christchurch, New Zealand 938

12. Nestler, J. M, R. A. Goodwin, D. L. Smith, and Toni Toney. 2007. Innova-939 tive Integration of Engineering and Biological Tools Aids Hydraulic 940 Structure Design for Restoring Threatened and Endangered Fish. In Pro-941 ceedings of the Fifth LACCEI International Latin American and Caribbe-942 an Conference for Engineering and Technology (LACCEI’2007), 943 “Developing Entrepreneurial Engineers for the Sustainable Growth of 944 Latin America and the Caribbean: Education, Innovation, Technology and 945 Practice” 29 May – 1 June 2007, Tampico, México. 946

13. Goodwin, R. A., J. M. Nestler, J. J. Anderson, D. L. Smith, D. Tillman, T. 947 Toney, L. J. Weber, S. Li, J. R. Cheng, and R. M. Hunter. 2008. The Nu-948 merical Fish Surrogate: Converting Observed Patterns in Fish Movement 949 and Passage to a Mechanistic Hypothesis of Behavior for Engineering De-950 sign Support. U.S. Army Engineer Research and Development Center, Wa-951 terways Experiment Station, Vicksburg, MS. 952

14. Goodwin R. A., M. Politano, J. W. Garvin, J. M. Nestler, D. Hay, J. J. An-953 derson, L. J. Weber, E. Dimperio, D. L. Smith, and M. Timko. 2014. Fish 954 Navigate Large Dams by Modulating Flow Field Experience. Proceed-955 ings of the National Academies of Science 111(14): 5277-5282. 956 DOI:www.pnas.org/cgi/doi/10.1073/pnas.1311874111 957

15. Nestler, J. M., M. J. Stewardson, D. Gilvear, J. Angus Webb, and D. L. 958 Smith. 2016. Does Ecohydraulics have guiding principles? Paper 26780 959 in, Webb JA, Costelloe JF, Casas-Mulet R, Lyon JP, Stewardson MJ (eds.) 960 Proceedings of the 11th International Symposium on Ecohydraulics. Mel-961 bourne, Australia, 7-12 February 2016. The University of Melbourne, 962 ISBN: 978 0 7340 5339 8. 963

16. Nestler, J. M., M. J. Stewardson, D. Gilvear, J. A. Webb, and D. L. Smith. 964 2016. Ecohydraulics exemplifies the emerging “Paradigm of the Interdis-965 ciplines”. Journal of Ecohydraulics DOI 966 10.1080/24705357.2016.1229142. Available at: 967 http://dx.doi.org/10.1080/24705357.2016.1229142 . 968

http://dx.doi.org/10.1080/24705357.2016.1229142


17. Nestler, J. M., C. Baigun, and I. Maddock. 2016. Achieving the aquatic 969 ecosystem perspective: Interdisciplinary integration describes instream 970 hydraulic processes. In River Science: Research and Management for the 971 21st Century, First Edition pp 84-102..Edited by David J. Gilvear, Mal-972 colm T. Greenwood, Martin C. Thoms and Paul J. Wood. 2016 John 973 Wiley & Sons, Ltd. Published 2016 by John Wiley & Sons, Ltd. 416 pp. 974 DOI: 10.1002/9781118643525.ch5 975

METHODOLOGICAL: 976

18. Nestler, J. M., R. A. Goodwin, and R. S. Chapman. 2000. “Development of a 977 Numerical Fish Surrogate for Improved Selection of Fish Passage Design 978 and Operational Scenarios for Lower Granite Dam: Phase I,” ERDC/EL 979 TR-00-12, U.S. Army Engineer Research and Development Center, Water-980 ways Experiment Station, Vicksburg, MS. 981

19. Nestler, J. M. and R. A. Goodwin. 2000. Simulating Population Dynamics 982 in an Ecosystem Context Using Coupled Eulerian-Lagrangian Ecological 983 Models (CEL HYBRID Models). ERDC/EL TR-00-04. U.S. Army Engi-984 neer Research and Development Center, Vicksburg, MS. 985

20. Nestler, J. M., R. A. Goodwin, and L. Toni Schneider. 2000. Linking Bio-986 logical Models and Spatial Descriptions of Environmental Complexity 987 with Coupled Models. Technical Note ERDC TN-EMRRP-EM-01 July 988 2000, U.S. Army Engineer Research and Development Center, Vicksburg, 989 MS 39180-6199. 990

21. Goodwin, R. A., and J. M. Nestler. 2000. Coupled Eulerian-Lagrangian 991 Hybrid (CEL Hybrid) Ecological Modeling and Hydroinformatics. Pro-992 ceedings of the 4Th International Conference on HydroInformatics, Cedar 993 Rapids, IA, 23-27 July 2000 (CD). 994

22. Goodwin, R. A., J. M. Nestler, L. Weber, Y. G. Lai, and D. P. Loucks. 2001. 995 Ecologically Sensitive Hydraulic Design for Rivers: Lessons Learned in 996 Coupled Modeling for Improved Fish Passage. Proceedings of ASCE 997 Specialty Conference on Wetlands Engineering and River Restoration 998 2001 held on 25-31 August 2001 in Reno, Nevada 999

23. Goodwin, R. A., J. M. Nestler, D. P. Loucks, and R. S. Chapman. 2001. 1000 Simulating mobile populations in aquatic ecosystems. ASCE Journal of 1001 Water Resources Planning and Management 127(6): 386-393. 1002

24. Goodwin, R. A., Anderson, J. J., and Nestler, J. M. 2004. Decoding 3-D 1003 movement patterns of fish in response to hydrodynamics and water quality 1004 for forecast simulation. Proceedings of the 6th International Conference on 1005 Hydroinformatics 2004, Liong, Phoon, and Babovic, eds., World Scientific 1006 Publishing Company, 21 – 24 June 2004, Singapore. 1007


25. Parasiewicz, P., J. Nestler, N.L. Poff and A. Goodwin. 2008. Virtual Refer-1008 ence River: A Model for Scientific Discovery and Reconciliation. 2008. 1009 In: M. S. Alonso, I. M. Rubio (ed) Ecological Management: New Re-1010 search , Nova Science Publishers, Inc pp. 189-206. ISBN: 978-1-60456-1011 786-1. 1012

26. Daraio, J. A, Weber, L. J., Newton, T. J., and Nestler, J. M. 2010. A meth-1013 odological framework for integrating computational fluid dynamics and 1014 ecological models applied to juvenile freshwater mussel dispersal in the 1015 Upper Mississippi River. Ecological Modelling 221 (2):201-214 1016 http://dx.doi.org/10.1016/j.ecolmodel.2009.10.008 1017

27. Daraio, J. A., L. J. Weber, T. J. Newton, and J. M. Nestler. 2010. A meth-1018 odological framework for integrating computational fluid dynamics and 1019 ecological models applied to juvenile freshwater mussel dispersal in the 1020 Upper Mississippi River. Ecological Modelling 221 (2):201-214 1021 http://dx.doi.org/10.1016/j.ecolmodel.2009.10.008 1022

28. Nestler, J. M. and R. A. Goodwin. 2000. Patent Number 6,160,759 Method 1023 for Determining Probable Response of Aquatic Species to Selected Com-1024 ponents of Water Flow Fields. 1025

APPLICATION: 1026

29. Nestler, J. M., Goodwin, R. A., T. Cole, D. Degan, and D. Dennerline. 2002. 1027 Simulating movement patterns of blueback herring in a stratified southern 1028 impoundment. Transactions of the American Fisheries Society 131:55-69. 1029

30. Goodwin, R. A., Nestler, J. M., Anderson, J. J., and Weber, L. J. 2004. 1030 Forecast simulations of 3-D fish response to hydraulic structures. Proceed-1031 ings of the ASCE World Water & Environmental Resources Congress 1032 2004, 27 June – 1 July 2004, Salt Lake 1033

31. Goodwin, R. A., Nestler, J. M., Anderson, J. J., and Weber, L. J. 2004. Vir-1034 tual fish to evaluate bypass structures for endangered species. Proceedings 1035 of the 5th Symposium on Ecohydraulics, 12 – 17 September 2004, Madrid, 1036 Spain.City, Utah. 1037

32. Goodwin, R. A., Nestler, J. M., Anderson, J. J., Kim, J., and Toney, T. 2005. 1038 Evaluating Wanapum Dam Bypass Configurations for Outmigrating Juve-1039 nile Salmon Using Virtual Fish: Numerical Fish Surrogate (NFS) Analysis. 1040 ERDC/EL TR-05-7, U.S. Army Engineer Research and Development Cen-1041 ter, Vicksburg, Mississippi 39180-6199. 1042 http://el.erdc.usace.army.mil/elpubs/pdf/trel05-7.pdf 1043

33. Weber, L.J., Goodwin, R.A., Li, S., and Nestler, J.M. 2006. Application of 1044 an Eularian-Lagrangian-Agent-Method to rank scenario designs of a juve-1045 nile fish passage facility. Journal of Hydroinformatics. 8(4):271-295. 1046

http://dx.doi.org/10.1016/j.ecolmodel.2009.10.008

http://dx.doi.org/10.1016/j.ecolmodel.2009.10.008

http://el.erdc.usace.army.mil/elpubs/pdf/trel05-7.pdf




34. Nestler, J. M., R. A. Goodwin, D. L. Smith, and J. J. Anderson. 2008. Op-1047 timum fish passage designs are based on the hydrogeomorphology of natu-1048 ral rivers. River Research and Applications: 24: 148-168. 1049

35. Daraio, J.A., L.J. Weber, S. J. Zigler, T. J. Newton, and Nestler, JM. 2010. 1050 Simulated effects of host fish distribution on juvenile unionid mussel dis-1051 persal in a large river. River Research and Application: Article first pub-1052 lished on line: 2010, DOI: 10.1002/rra.1469. 1053

36. Nestler, J. M., P. Pompeu, R. A. Goodwin, D. L. Smith, L. Silva, C. R. M. 1054 Baigún, and N. O. Oldani. 2012. The River Machine: A Template for 1055 Fish Movement and Habitat, Fluvial Geomorphology, Fluid Dynamics, 1056 and Biogeochemical Cycling. River Research and Application 28(4): 490-1057 503 (wileyonlinelibrary.com) DOI: 10.1002/rra.1567 1058

37. Smith, D. L., R. A. Goodwin, Y. Long, and J. M. Nestler. 2014. Fish Path Se-1059 lection and Fatigue in Complex, Continuous Velocity Fields (Extended ab-1060 stract). Proceedings : 10th International Symposium on Ecohydraulics 1061 2014 , Norwegian University of Science and Technology, Trondheim, 1062 Norway, 23rd - 27th June 2014. SINTEF Energi AS:Paper 323. 1063

38. Smith, D. L., J. M. Nestler, T. Threadgill, R. A. Goodwin. 2016. Movement 1064 analysis of fish near a low head dam on the Mississippi River. Proceed-1065 ings of RiverFlow 2016 held at St. Louis University, St. Louis, Missouri, 1066 July 12-15, 2016. 1067

Documents

SCENARIO ANALYSIS OF FREMONT WEIR NOTCH · PDF fileerdc tr l-17-t scenario analysis of fremont weir notch – integration of engineering designs, telemetry, and flow fields d r david