Lost River Model for
TNDL Development
August 2005
Prepared for:
U.S. Environmental Protection Agency Region 10
U.S. Environmental Protection Agency Region 9
Oregon Department of Environmental Quality
North Coast Regional Water Quality Control Board
Prepared by:
Water Resources and TMDL Center
J Tetra Tech, Inc.
-------
Model Configuration and Results
Lost River Model for TMDL Development
August 29, 2005
Prepared for:
U.S. Environmental Protection Agency Region 10
U.S. Environmental Protection Agency Region 9
Oregon Department of Environmental Quality
North Coast Regional Water Quality Control Board
Prepared by:
Tetra Tech, Inc.
-------
Model Configuration and Results
TABLE OF CONTENTS
ACKNOWLEDGMENTS 3
1.0 INTRODUCTION 4
2.0 MODELING APPROACH 6
2.1 MODEL SELECTION 6
2.2 MODEL ENHANCEMENTS 6
2.2.1 Piecewise Simulation 7
2.2.2 Tributary Partitioning. 7
2.2.3 Sediment Oxygen Demand 7
2.2.4 Aquatic Vegetation 8
2.2.5 Slope Impacts 9
2.2.6 Flow Control Structures 9
2.2.7 TMDLDevelopment Tool 10
2.3 MODEL CONFIGURATION 10
2.3.7 Segmentation/Computational Grid Setup 10
2.3.2 State Variables 14
2.3.3 Boundary Conditions/Linkages 14
2.3.4 Initial Conditions 33
2.4 MODELING ASSUMPTIONS AND LIMITATIONS 35
2.4.1 Assumptions 35
2.4.2 Limitations 36
3.0 MODEL TESTING 37
3.1 MONITORING LOCATIONS 37
3.2 HYDRODYNAMIC SIMULATION 40
3.2.7 Hydraulic Parameter Designation 40
3.2.2 Water Balance and Water Surface Elevation Calibration 40
3.2.3 Hydrodynamic Model Evaluation w ith Temperature Data 41
3.2.4 Further Hydrodynamic Model Evaluation with Conductivity Data 42
3.3 WATER QUALITY SIMULATION 42
3.3.7 Lost River at Keller Bridge (LRKB) 46
3.3.2 Lost River at HarpoldDam (LRHD) 46
3.3.3 Lost River at RM 27 (HPDS2) 47
3.3.4 Lost River at Wilson Reservoir (LRWRC) 47
3.3.5 Lost River at Anderson Rose Dam (LRAR/ARDMUS) 47
3.3.6 Lost River at East West Road (LREW) 48
3.3.7 Lost River at Tule Lake (TLTO) 48
3.3.8 P-Canal(PC) 49
3.3.9 Klamath Strait Drains at the State Line (KSDSR) 49
3.3.10 Klamath Straits Drain at Pump E (KSDTR) 49
3.3.11 Klamath Straits Drain at Pump F (KSDPSF) 49
3.3.72 Klamath Straits Drain at Highway 97 (KSD97) 50
3.3.73 Klamath Straits Drain at Railroad (KSDM) 50
-------
Model Configuration and Results
3.3.14 Diel Dissolved Oxygen Analysis 50
3.3.15 Macrophyte Analysis 52
3.4 MODEL SENSITIVITY ANALYSES 52
4.0 TMDL SCENARIO 55
5.0 DISCUSSION 60
6.0 REFERENCES 62
APPENDIX A_1999
APPENDIX A_2004
APPENDIX B_1999
APPENDIX B_2004
APPENDIX C_1999
APPENDIX C_2004
APPENDIX DJI999
APPENDIX D_2004
APPENDIX E_2004
APPENDIX F_1999
APPENDIX G
1999 Water Balance and Water Surface Elevation
Calibration
2004 Water Balance and Water Surface Elevation
Calibration
1999 Hydrodynamic Model Evaluation with
Temperature Data
2004 Hydrodynamic Model Evaluation with
Temperature Data
1999 Further Hydrodynamic Model Evaluation
with Conductivity Data
2004 Further Hydrodynamic Model Evaluation
with Conductivity Data
1999 Water Quality Calibration Results
2004 Water Quality Calibration Results
Evaluation of Macrophyte Mass and Diel DO
Comparison
Sensitivity Analysis
TMDL Scenario
-------
Model Configuration and Results
ACKNOWLEDGMENTS
The Lost River Hydrodynamic and Water Quality Model was developed in a relatively short
period of time. This was made possible by the generous support and responsiveness of many
people. The authors would like to acknowledge the following professionals, in particular, for
their contributions:
Jason Cameron
Thomas Cole
Ben Cope
Mark Filippini
Bill Hobson
Steve Kirk
David Leland
Gail Louis
Tim Mayer
John Rasmussen
Matt St. John
Daniel Turner
Bureau of Reclamation
U.S. Army Corps of Engineers
U.S. Environmental Protection Agency, Region 10
U.S. Environmental Protection Agency, Region 10
North Coast Regional Water Quality Control Board (formerly of)
Oregon Department of Environmental Quality
North Coast Regional Water Quality Control Board
U.S. Environmental Protection Agency, Region 9
U.S. Fish and Wildlife Service
Bureau of Reclamation
North Coast Regional Water Quality Control Board
Oregon Department of Environmental Quality
-------
Model Configuration and Results
1.0 INTRODUCTION
The Oregon DEQ (ODEQ) and the North Coast RWQCB (NCRWQCB) have both included the
Lost River on their corresponding 303d Lists as a result of observed water quality criteria
exceedances. The Lost River is actually composed of a series of riverine segments,
impoundments, drains, and canals that straddle the Oregon-California border from Clear Lake
Reservoir to the Klamath River. Impairments include dissolved oxygen, chlorophyll a,
temperature, fecal coliform, pH, and ammonia for various portions of the Lost River system
(including the Klamath Straits Drain) in Oregon and nutrients, temperature, and pH for segments
of the system in California. As such, the states are required to develop TMDLs for applicable
water quality parameters. The first steps in the TMDL development process have already been
conducted and included compilation of available data; evaluation of monitoring data to identify
the extent, location, and timing of water quality impairments; and development of a technical
approach to analyze the relationship between source pollutant loading contributions and in-stream
response. These steps were detailed in "Data Review and Modeling Approach - Klamath and
Lost Rivers TMDL Development," dated April 23, 2004. Subsequent steps include model
configuration, model testing (calibration and corroboration), and scenario analysis. This
document discusses the configuration of the Lost River model and presents modeling results for
the Lost River from Malone Dam through Klamath Straits Drain to the Klamath River for 1999
and 2004.
The Lost River originates at Clear Lake Reservoir in Northern California. It flows in a
northwesterly direction into Oregon, to the town of Bonanza. The river then returns south, passes
in close proximity to the city of Klamath Falls, and ultimately flows back into California. In
California, the Lost River flows into Tule Lake, the river's natural hydrologic termination. Since
the early 1900's, extensive flood diversion and irrigation facilities have been constructed
throughout the Lost River Basin, known as the U.S. Bureau of Reclamation's (BOR) "Klamath
Project" (WRE 1965). The "Klamath Project" has a significant impact on the hydrology and
water quality of the Lost River, because it essentially creates a series of impoundments (including
those at Malone Dam, Harpold Dam, Wilson Dam, and Anderson-Rose Dam) and free-flowing
river segments. Flow magnitudes change dramatically throughout the year along the entire length
of the Lost River due to irrigation practices and operation of the "Klamath Project." Modification
of the natural hydrology has also enabled the Lost River to be hydrologically-connected to the
Klamath River for the past century through a series of pumps, canals, drains, and impoundments
(from Tule Lake, through the P Canal, into the Lower Klamath Lake, and through the Klamath
Straits Drain).
Due to the complex physical nature of the Lost River and its profound influence on water
chemistry and biology, every attempt was made to obtain the most current and comprehensive
data to support model development, application, and analysis. Although data were accessed from
numerous sources and multiple, focused water quality monitoring efforts were conducted during
the summer of 2004, there are still extensive data limitations. The monitoring efforts
demonstrated that flow and water quality conditions can change dramatically from one day to the
next at different locations in the system (even without significant atmospheric changes). These
conditions imply that a detailed understanding of the time-variable nature of irrigation return
flows and withdrawals along the entire length of the Lost River (and associated water quality
characteristics), is critical. Unfortunately, these data sets are not currently available. The
modeling approach that was pursued made the best use of available data and provides a
framework which can be readily updated in the future as more data become available. While the
-------
Model Configuration and Results
model may not be able to predict sudden, localized changes in the system or flawlessly reproduce
the temporal variability of every water quality parameter, it can be used to evaluate temporal and
spatial trends and perform allocation analysis for TMDL development.
-------
Model Configuration and Results
2.0 MODELING APPROACH
2.1 Model Selection
In order to support TMDL development for the Lost River system, the need for an integrated
receiving water hydrodynamic and water quality modeling system was identified. The "Data
Review and Modeling Approach - Klamath and Lost Rivers TMDL Development" proposed
implementing the U.S. Army Corps of Engineers' CE-QUAL-W2 (W2) model for the entire Lost
River system from Malone Dam through the Lower Klamath Lake, as well as the Klamath Straits
Drain. This approach was approved by EPA Region 10, EPA Region 9, ODEQ, and NCRWQCB
in May, 2004.
W2 is a two-dimensional, longitudinal/vertical (laterally averaged), hydrodynamic and water
quality model (Cole and Wells 2003). The model is applicable to lakes, rivers, and estuaries that
do not exhibit significant lateral variability in water quality conditions. It allows application to
multiple branches for geometrically complex waterbodies with variable grid spacing, time-
variable boundary conditions, and multiple inflows and outflows from point/nonpoint sources and
precipitation.
The two major components of the W2 model include hydrodynamics and water quality kinetics.
Both of these components are coupled, i.e. the hydrodynamic output is used to drive the water
quality at every timestep. This makes it very efficient to execute model runs. The hydrodynamic
portion of the model predicts water surface elevations, velocities, and temperature. The W2
model uses the ULTIMATE-QUICKEST numerical scheme for advection computation. The
ULTIMATE-QUICKEST numerical scheme is a third order finite difference scheme. This
method reduces numerical diffusion in the vertical and horizontal directions to a minimum. In
areas of high gradients this scheme eliminates undershoots and overshoots which may produce
small negative concentrations. The water quality portion of W2 can simulate the constituents
required for Lost River TMDL development, including dissolved oxygen (DO), nutrients,
phytoplankton interactions, macrophytes, and pH. In addition, the model is equipped to simulate
other generic constituents.
2.2 Model Enhancements
The Lost River is a highly hydro-modified system that is predominantly fed by lake diversions. It
contains multiple impoundments along its length that cause the river to exhibit highly variable
hydraulic conditions, including free-flowing riverine segments and relatively stagnant
reservoirs/ponds. These conditions, along with significant return flow and withdrawal from
adjacent agricultural lands, lead to significant variability within short time periods. These
dynamics have a dramatic effect on water quality and lead to a unique, highly biologically-active
system that is inundated with phytoplankton and macrophytes during the spring, summer, and
fall. Although the W2 model is capable of addressing the issues identified above, a number of
enhancements to W2 version 3.2 were deemed necessary to expedite and strengthen the model for
the rigors of Lost River TMDL development. These are described below.
-------
Model Configuration and Results
2.2.1 Piecewise Simulation
The W2 code was modified to enable piecewise (i.e., waterbody-by-waterbody) simulation. This
modification was instituted primarily to improve computational efficiency during the model
calibration process. The Lost River is a complex hydraulic system that is divided into sections by
a series of dams and other physical features (e.g., tunnels and pumps). To most accurately
represent the system using W2, it needs to be divided into smaller "sections" composed of
discrete computational segments. This is described in subsequent sections. Because the system is
so large and is broken up into a large number of segments, it is extremely time-consuming to run
the entire model throughout the testing process. The simulation of an upstream waterbody, for
example, does not need the information for downstream segments (unless they are linked using an
internal head boundary). In the existing W2 framework, even if modelers and decision makers
are only concerned with the most upstream waterbody, the model would have to solve the
governing equations for all waterbodies. Additionally, when all waterbodies are simulated
together, any time step constraints that apply to one waterbody (such as the need to use an
extremely short timestep to avoid numerical instability), would apply to all of them, and this can
significantly lengthen the simulation time. To achieve more efficient model calibration and
scenario evaluation, a piecewise modeling capability was incorporated into W2 which allows the
user to setup a model for all waterbodies of interest, but only run select waterbodies. The model
tracks and stores previous runs and enables the user to incorporate these results as boundary
conditions (when interested only in downstream segments). And it essentially disconnects
unnecessary downstream segments when the user is interested only in upstream segments.
2.2.2 Tributary Partitioning
A user-defined, distributed tributary partitioning function was incorporated into the W2 code to
more accurately represent diffuse flows (e.g., surface runoff, return flows, and withdrawals) into
and out of the river. In the existing W2 model, distributed flows (i.e., any lateral flow not
represented by a discrete tributary) are applied to each modeled segment based on the segment's
length. While this is typically a reasonable assumption, the Lost River's land-based inflows and
withdrawals are highly managed and thus cannot be represented based on segment length (or even
watershed area). To allow for a more flexible distribution of the distributed flow, a function was
incorporated into W2 to provide users the ability to specify a spatial distribution pattern for the
distributed flows at each branch.
2.2.3 Sediment Oxygen Demand
To improve W2's representation of SOD, a Monod type SOD formulation was incorporated into
the code, as presented by Chapra (1997). Chapra suggested that SOD is nonlinearly related to the
water column DO concentration, such that the higher the DO, the higher the SOD. This
formulation more accurately represents the dependence of SOD on water column DO
concentration. In the existing W2 model, SOD is related to water column DO such that when DO
is lower than a threshold, SOD is "turned off" until the DO recovers (above the threshold). SOD
is then suddenly turned on at that point. The current formulation introduces abrupt SOD behavior
into the system. Therefore, it was improved upon using a Monod type formulation to achieve a
smooth transition. The formulation is as follows:
SOD=DO/(DO + D0half_saturation) *SODC (Eq. 2-1)
-------
Model Configuration and Results
where: SOD is the effective SOD in g/m2/day
DO is the water column dissolved oxygen concentration
DOhaiLsatwation is the half saturation coefficient that represents the oxygen level at
which SOD becomes half of the specified zero-order SODC coefficient.
With this formulation, the effective SOD changes continuously with the overlying water column
DO and provides for a more reasonable approximation of natural processes.
2.2.4 Aquatic Vegetation
The Lost River is a biologically productive system that exhibits excessive growth of floating
(phytoplankton) and rooted aquatic vegetation (macrophytes - which for the purpose of this
discussion include epiphyton and periphyton). A recent survey of the Lost River during July
2004 by MaxDepth Aquatics indicated that the dominant taxon present in the river system was
Ceratophyllum demersum (coontail). Other common species included Lemna minor (duckweed),
several species of pondweed (Potomogeton pectinatus, P. crispus, and P. nodosus), Elodea
canadensis, Heteranthera dubia, and Cladophora sp. All of these taxa are tolerant of high
turbidity and are common species found in eutrophic lakes and slow-moving waters. Because the
macrophytes, in particular, play such an integral role in the dynamics of the river, it is critical that
they are properly addressed in the modeling framework. The formula incorporated is the same as
in the EFDC model (Park et all995) and the WASP model (Shanahan and Alam 2001).
The current version of W2 is capable of simulating phytoplankton and epiphyton/periphyton
(macrophytes). The current representation of macrophytes does not consider substrate
availability or flow velocity limitation. Additionally, light conditions are calculated on a layer-
average basis and the impact of macrophyte growth on hydrodynamic simulation is not handled.
In an effort to more accurately represent macrophyte dynamics, a series of enhancements were
made to W2 to address these limitations. No modifications were made to the phytoplankton
representation, and it was used for this application.
Substrate availability limitation was added to the model. The substrate availability limitation
accounts for the impact of bed composition on macrophyte communities. With this factor
considered, macrophyte growth can be limited by bed type (i.e., the specified bed particle size
distribution must be suitable for growth). In general, periphyton growth is limited in areas
dominated by fine sediments whereas macrophytes (generally more prevalent in this system) may
exhibit limited or no growth in areas characterized by cobbles and large particles. The fraction of
mud reported in the MaxDepth Aquatics survey was used to characterize substrate availability
throughout the system.
Hydrodynamic interactions with macrophyte growth were considered in this modeling effort.
High in-stream velocities can have a limiting effect on the macrophyte community. The
magnitude of the limiting effect was represented using the half saturation coefficient specified in
the model input file. In general, the higher the value for the half saturation coefficient, the
smaller the impact of velocity on macrophyte growth. The growth of macrophyte communities
can increase the bottom roughness of river channels, and thus has an effect on hydrodynamic
behavior. A simplified Monod-type formulation was used in the model to relate the density of the
macrophytes to the stream bottom roughness coefficients. The half saturation coefficients for this
relationship are obtained through calibration where sufficient data were available. To most
accurately simulate the impact of macrophyte growth on hydrodynamics, detailed numerical
representation of their physical characteristics would be required, and this is outside the scope of
the current study.
-------
Model Configuration and Results
Growth thickness was also included in the model to better represent the actual light conditions
affecting macrophyte growth. The plant (mattress) height is specified in the updated model input
file, and the model calculates the vertical average light condition for the specified height/length of
the macrophytes. This is used in the growth calculation. The formulation itself is equivalent to
the original W2 formulation for the case when the macrophyte occupies the entire layer. It
improves predictions when macrophytes occupy only a portion of the layer and is especially
useful for periphyton simulation (since periphyton tend to grow on the bottom only).
A single macrophyte category was used in the model to represent all macrophyte species present
in the river. Representation of all species as a single category is supported by results of the recent
survey that indicate present species generally obtain most of their nutrients from the water
column in a similar manner. Ideally, a comprehensive ecological model could be incorporated
into the system to study inter-species competition and co-evolution, but this is precluded by data,
time, and resource constraints. It should be noted that phytoplankton was also simulated for the
Lost River, and this was done using the existing W2 algorithms.
2.2.5 Slope Impacts
The model's ability to represent SOD and macrophytes for a significantly sloping channel (i.e., in
the longitudinal direction - from upstream to downstream) was also improved. Using the original
W2 code, the impact of SOD and macrophytes on DO can be significantly under-predicted in
some sections of a sloping channel due to the heterogeneity in layer subtraction/addition for a
long, sloping channel. For example, in the Lost River section from Malone Dam to Harpold
Dam, the upstream segments may become very shallow during low flow conditions, however
downstream segments may maintain considerable depth (due to return flows and damming).
Under such a circumstance, the original W2 model would implement layer subtraction based on
the upstream segments (i.e., reduce the number of vertical layers in the model). This would result
in only one active layer throughout the entire modeling segment. Unfortunately, W2 currently
would not be able to change the bed area for this layer to accurately represent the bed dimensions
of this "new" artificial layer. Thus, the bed's impact on SOD and macrophyte prediction would
not be accurately represented. To better represent potential variability, bed area associated with
SOD and macrophytes is tracked in the updated model throughout the built-in layer subtraction
and addition process. Therefore, the impact of SOD and macrophytes is better represented. It
should be noted that the original W2 can generate reasonable results when the width of each
vertical model layer varies gradually (i.e., there is not an abrupt change in segment width between
any model layers in the vertical direction). However, when a narrow bottom layer is used to
contain low flow, the error in SOD and macrophytes is significant.
2.2.6 Flow Control Structures
To more accurately simulate flow downstream of impoundments and to avoid numerical
instability during the low-flow season, the existing W2 flow control structure equation was
enhanced to include a leakage term. This added leakage term allows some water to flow from
upstream segments to downstream segments, even when the upstream water level is lower than
the dam crest. Dam leakage allows downstream segments to realistically remain wet during low
impoundment water surface elevation periods. This function was added because there is always a
small amount of leakage water downstream of impoundments in the Lost River (personal
communication with BOR). Additionally, the hydrodynamic model will not run when there is no
water, thus a minimum flow must be maintained in the system.
-------
Model Configuration and Results
2.2.7 TMDL Development Tool
To expedite TMDL development, a TMDL Development Tool was developed and incorporated
into the W2 model. The tool allows the user to specify the load reduction ratio for each loading
source and constituent in an external control file. The model then directly uses this information
to adjust the boundary conditions for scenario analysis.
2.3 Model Configuration
Model configuration involved setting up the model computational grid (bathymetry) using
available geometric data, designating the model's state variables, and setting boundary
conditions, initial conditions, and hydraulic and kinetic parameters for the hydrodynamic and
water quality simulations. This section describes the configuration process and key components
of the model in greater detail.
2.3.1 Segmentation/Computational Grid Setup
The computational grid setup defines the process of segmenting the entire Lost River into smaller
computational segments for application of the W2 finite difference scheme. In general,
bathymetry is the most critical component in developing the grid for the system.
For this modeling study, the Lost River was divided into 12 waterbodies based on the presence of
major hydraulic features and the location of monitoring data in the system. Each of these
waterbodies, which are listed in Table 2-1 and shown graphically in Figure 2-1, was represented
using unique geometric and hydrological characteristics in the model. It should be noted that all
longitudinal dimensions (in meters) in the table are approximate values, as measured using CIS.
Longitudinal dimensions presented for waterbodies 7 through 12 are estimates, because
waterbodies 7 and 9 are wide lakes that do not necessarily have a finite length. A combination of
USGS quadrangle maps for Oregon and California and RF3 reach file layers in CIS were used to
establish the longitudinal dimensions of the system.
10
-------
Model Configuration and Results
Table 2-1. Lost River Model Waterbodies
Waterbody Description Starting Ending Number Segment Number Layer
# River River of Length of Thickness
Meter Meter Segments (m) Layers (m)
(m) (m)
1
2
3
4
5
6
7
8
9
10
11
12
Malone-
Harpold
Harpold- Ranch
Ranch-Wilson
Reservoir
Wilson
Reservoir
Wilson Dam to
Anderson
Rose Dam
Anderson
Rose Dam to
Tule Lake
Tule Lake
P-Canal
Lower Klamath
Lake
Klamath Straits
Drain before
Pump E
Klamath Straits
Drain before
Pump F
Klamath Straits
Drain after
Pump F
0
38,638
43,535
58,695
63,253
92,653
104,722
112,732
116,757
128,655
135,255
143,327
38,638
43,535
58,695
63,253
92,653
104,722
112,732
116,757
128,655
135,252
143,327
146,346
80
10
30
9
55
24
1
8
1
13
15
6
483.0
489.7
505.3
506.4
534.5
502.9
8008.0
502.6
11898.0
507.2
538.1
503.2
5
4
4
5
5
4
2
3
2
5
5
5
1.0
0.96
0.84
1.0
1.0
1.0
1.0
1.0
1.0
1.15
0.93
0.93
11
-------
Model Configuration and Results
o
.a
OJ
-a
o
CD
2
CD
12
-------
Model Configuration and Results
Each of the 12 modeling waterbodies was further divided into computational segments (a.k.a.
segments) for greater detail in modeling. The number of segments and lengths of each segment
varied by modeling waterbody, however each segment was approximately 500 meters in length
(with the exception of Tule Lake and the Lower Klamath Lake which were each represented as a
single computational segment). This resulted in a whole number of segments for each segment.
It should be emphasized that Tule Lake and the Lower Klamath Lake are very complex features,
however due to the scope of the study and geometric, hydrologic, and water quality data
limitations, representation using a single computational segment for each was deemed
appropriate. Major data limitations precluding a more detailed representation of these lakes
include absence of bathymetric data and lack of spatially- and temporally-variable loading data.
These lakes can be further segmented in the future to examine spatial variability of water quality,
in the event that sufficient additional data are obtained. Table 2-1 also summarizes the lengths
and numbers of computational segments for each of the 12 waterbodies.
Within the W2 model, each computational segment can have multiple layers associated with it.
The number of layers and layer thickness for each computational segment is designated based on
physical characteristics and the need to adequately represent the vertical variation of water quality
while maintaining computational stability and limiting simulation time. The number of vertical
layers varied for each of the modeling waterbodies from 2 to 5 layers. Previous W2 applications
have used a vertical grid spacing (layer thickness) of 0.2 meters to 5 meters (Cole and Wells
2003). For this study, layer thicknesses were set to approximately 1 meter (and ranged from 0.84
meters to 1.15 meters) for the 12 waterbodies (Table 2-1).
The average width of each computational segment varied significantly for each waterbody (and
each layer) used to represent the Lost River system. Depth-variable widths were specified using a
combination of USGS quad maps; Lost River Channel Improvements Plans from the early 1900's
(from BOR) for several locations on the Lost River - Poe Valley, Langell Valley, Klamath Straits
Drain, and the Lower Lost River; and physical measurements conducted by NCRWQCB and
BOR in late 2003 and early 2004. The NCRWQCB and BOR physical measurement locations
are presented in Table 2-2.
Table 2-2. Physical Measurement Locations
Site ID Site Name
LRDM
LRGR
MCEL
LRCR
LRKB
LR70
BC
LRHDB
LRPV
LROG
LRWRC
DR1
LRDCTR
LRDR
LRSB
DR5
Lost River downstream of Malone Dam
Lost River at Gift Road Bridge
Miller Creek at East Langell Valley Road
Lost River at Cheese Factory Road Bridge
Lost River at Keller Bridge
Lost River at Highway 70 Bridge in Bonanza
Buck Creek @ Burgdorf Road/Casebeer Road
Lost River downstream of Harpold Dam @ bridge
Lost River @ Poe Valley, Lost River Ranch @ bridge
Lost River @ Olene Gap
Lost River in Wilson Reservoir @ Crystal Springs Road Bridge
Drain #1 (downstream of Lost River Diversion Channel)
Lost River Diversion Channel @ Tingley Road Bridge
Lost River @ Dehlinger Road
Lost River @ Stukel Bridge (Matney Way)
Drain #5 (Wong Road)
13
-------
Model Configuration and Results
Site ID Site Name
LR39
LRFR
LRMB
LRARB
LREW
PC
KSDM
NNC
ADY
Lost River @ Highway 39 Bridge
Lost River @ Falvey Road Bridge
Lost River @ S. Merrill Road Bridge
Lost River downstream of Anderson- Rose Dam @ Bridge
Lost River @ East/West Road Bridge
P Canal
Klamath Straits Drain West of Rail Road Tracks
North Canal section between RR tracks and Hwy 97
ADY Canal section, east side of Hwy 97 Bridge
Once the dimensions of the computational segments had been defined for the grid, the segment
orientation was specified. These values were specified in radians, with north represented as zero
radians. The segment orientation was measured for each segment in each waterbody and stored
in the bathymetry files.
2.3.2 State Variables
Selection of appropriate model state variables to represent processes and hydrodynamic and water
quality processes of concern is a critical factor in model configuration. For this study, state
variables were selected to most accurately predict TMDL impairments and related physical,
chemical, and biological processes. The following constituents were configured for the Lost
River model in W2. Refer to Cole and Wells (2003) for details regarding equations utilized and
interactions represented.
1) Conductivity
2) Temperature
3) ISS (inorganic suspended solids)
4) PO4 (dissolved inorganic phosphorus)
5) NH4 (ammonium)
6) NO3/NO2
7) LDOM (labile dissolved organic matter)
8) RDOM (refractory dissolved organic matter)
9) LPOM (labile particulate organic matter)
10) RPOM (refractory particulate organic matter)
11) DO (Dissolved Oxygen)
12) CBOD (Carbonaceous Biochemical Oxygen Demand)
13) Alkalinity
14) TIC (total inorganic carbon)
15) Phytoplankton
16) Macrophytes (epiphyton/periphyton)
2.3.3 Boundary Conditions/Linkages
To run the dynamic W2 model external forcing factors, known as boundary conditions, and
internal linkages must be specified for the system. These forcing factors are a critical component
in the modeling process and have direct implications on the quality of the model's predictions.
External factors include a wide range of dynamic information:
14
-------
Model Configuration and Results
Upstream external inflows, temperature, and constituent boundary conditions (US);
Tributary inflows, temperature, and constituent boundary conditions (TRIE);
Distributed tributary inflows, temperature, and constituent boundary conditions (DST);
Withdrawals (WD); and
Atmospheric conditions (including wind, air temperature, solar radiation).
Upstream external inflows essentially represent the inflow at the model's "starting" point.
Tributary inflows represent the major tributaries that feed into the Lost River. Distributed
tributary inflows represent the combination of all diffuse contributions to each of the waterbodies
(i.e., anything that is not considered a major tributary inflow, such as irrigation return flow). All
water removed from the system is combined within the Withdrawals category. The US, TRIE,
DST, and WD boundary conditions were specified for the Lost River model based on all available
data. The available data are sufficient to provide limited spatially- and temporally-variable
inputs. Thus, the boundary conditions generally represent "smoothed" or averaged conditions
over a period of time. Ideally, high-resolution time-variable inputs should be used to drive the
model, however, these inputs are currently not available. "Smoothed" conditions limit the
model's ability to predict localized and short-term effects, however, they enable the model to
reasonably predict trends.
Meteorological data are an important component of the W2 model. The surface boundary
conditions are determined by the meteorological conditions. The meteorological data required by
the W2 model are air temperature, dewpoint temperature, wind speed, wind direction, and cloud
cover. In general, hourly data are recommended (expressed in Julian Day) (Cole and Wells
2003). Hourly, unedited local climatological data were used from the Klamath Falls Airport for
the entire Lost River system. These data provided the most complete data set of required
meteorological parameters for the W2 model meteorological file. Cloud cover was calculated
using hourly sky conditions reported at this site (which are based on a scale from 0 to 8). Table
2-3 shows the lookup table used for calculating the cloud cover. The sky conditions reported
were converted to a scale often based on W2's meteorological data file requirements for cloud
cover.
Table 2-3. Cloud Cover Lookup Table
Cloud Cover Condition Cloud Cover
CLR (Clear)
FEW (Few)
SCT (Scattered)
BKN (Broken)
VV (Vertical Visibility into fog or snow)
OVC (Overcast)
0.05
0.25
0.50
0.75
0.90
0.95
Precipitation and evaporation inputs are not directly considered for the river sections in the Lost
River model formulations. They were assumed to be equivalent, on an annual basis, due to their
relatively small surface areas. For Tule Lake and Lower Klamath Lake evaporation is simulated
using the default parameters and equations in CE-QUAL-W2. It should be noted, however, that
although the model does not explicitly consider precipitation or evaporation inputs for the
riverine sections, it does simulate the impact of evaporation on heat balance.
15
-------
Model Configuration and Results
Internal linkages that must be specified include:
Downstream weir-based boundary conditions (DSW);
Upstream internal flow, temperature and constituent boundary conditions (USIFB);
Downstream internal head boundary conditions (DSIH); and
Upstream internal head boundary conditions (USIH).
Dams (e.g., Harpold and Anderson Rose), which represent the most downstream portion of many
Lost River waterbodies, are represented using downstream weir-based boundary conditions. The
equations instituted are described below. Internal boundary conditions for waterbodies that are
downstream of a dam (or segment represented with DSW) are represented using upstream
internal boundary conditions (or essentially, the outflows based on the DSW equations).
Downstream and upstream internal head boundary conditions are used to link free-flowing
waterbodies (i.e., those not divided by a physical structure).
The model was first configured and calibrated (tested) for 1999 due to data availability and the
exhibition of water quality criteria exceedences. The calibration was corroborated using 2004
data. The model boundary conditions primarily utilized information from 1999 (for both the
1999 and 2004 runs). Any exceptions are discussed in subsequent sections (initalics). Figure 2-
2 presents a diagram of the modeling environment and relative locations of boundary conditions.
Yellow circles represent divisions between the 12 modeling waterbodies. Blue arrows represent
TRIE inputs and the one US input. Red arrows represent DST inputs, which are not at a single
location, but rather distributed along the entire waterbody length. WDs are represented with
green arrows. Table 2-4 lists the primary boundary conditions and linkages by waterbody.
16
-------
Model Configuration and Results
A Canal
(represented
through DSTs)
TRIB(F1 Canal)
WD (Lost River Diversion Channel)
Lost River
Diversion Channel
(represented
through TRIBs)
O
I
TRIB (Station 48 Turnout)
TRIB (Drain 1)
TRIB (DrainS)
TRIB (Merrill STP)"'
Anderson-Rose
QKIamath Straits Dam
Drain
DST
WD
(Harpold Dam)
WD (J Canal)
DST
TRIB
.(Mller Creek)
DST
Lower
Wamath
Lake
o
P Canal /-) Tule Lake
' ^ and Sump
\
DST
\
Malone Darn
DST
DST
Figure 2-2. Lost River Model
Table 2-4. Boundary conditions and Linkages for the Lost River Waterbodies
Waterbody
1
1
1
1
1
1
1
2
2
2
2
3
3
3
4
4
4
4
Location
Malone Dam
From Malone to Harpold Dam
Miller Creek
Big Springs
Buck Creek
Lost River (LR) before Harpold Dam
LR at Harpold Dam
LR downstream of Harpold Dam
LR from Harpold to RM 27
E Canal
LR at RM 27
LR downstream of RM 27
LR from Ranch to Wilson Res
LR before entering Wilson Res
LR entering Wilson Res
LR at Wilson Reservoir
LR upstream of Wilson Dam
F-1 Canal
sBTHiTiTsTSIERJiTiTnRJil
US
DST
TRIB
TRIB
TRIB
WD
DSW
USIFB
DST
TRIB
DSIH
USIH
DST
DSIH
USIH
DST
WD
TRIB
17
-------
Model Configuration and Results
Waterbody Location Boundary Condition
4
5
5
5
5
5
5
5
5
6
6
6
7
7
7
8
8
8
9
9
9
9
10
10
10
11
11
11
12
12
12
LR at Wilson Dam
LR downstream of Wilson Dam
LR from Wilson Dam to Anderson Rose
Dam
LRat J-Canal
LR at Anderson Rose Dam
Station 48 Turnout
Drain #1
Drain #5
City of Merrill STP
LR downstream of Anderson Rose Dam
LR from Anderson Rose Dam to Tule
Lake
LR before entering Tule Lake
LR entering Tule Lake
LR at Tule Lake
LR at Tule Lake outlet
P Canal downstream of Tule Lake
P Canal
LR before entering Tule Lake
P canal entering Lower Klamath Lake
Lower Klamath Lake
ADY Canal
Lower Klamath Lake entering Klamath
Straits Drain
Klamath Straits Drain leaving Lower
Klamath Lake
Klamath Straits Drain from Lower Klamath
Lake to Pump E
Klamath Straits Drain at Pump E
Klamath Straits Drain leaving Pump E
Klamath Straits Drain from Pump E to
Pump F
Klamath Straits Drain at Pumps F
Klamath Straits Drain leaving Pump F
Klamath Straits Drain from Pump F to
Discharge at Klamath
Klamath Straits Drain at Klamath River
DSW
USIFB
DST
WD
DSW
TRIB
TRIB
TRIB
TRIB
USIFB
DST
DSW
USIFB
DST
DSW
USIFB
DST
DSW
USIFB
DST
TRIB
DSW
USIFB
DST
DSW
USIFB
DST
DSW
USIFB
DST
DSW
The following sub-sections provide a detailed description of the boundary conditions used to
represent each waterbody.
2.3.3.1 Waterbody #1: Malone Dam to Harpold Dam
US: The 1999 daily flow data downstream of Malone Dam from the BOR database were used to
form the upstream inflow boundary condition. During the irrigation period Malone Dam
discharge into the Lost River is effectively zero, with the exception of dam leakage (which was
represented in the model as 0.2 cms for the sake of model stability). Time-variable temperature,
monitored at LRDM (downstream of the Malone Dam diversion), were also used. Since the
18
-------
Model Configuration and Results
temperature data are relatively sparse, dates without data were obtained through linear
interpolation using the measured data.
With regard to water quality parameters, the measured data do not cover the complete list of the
model state variables, so certain assumptions were made in initially setting the boundary
conditions. It should be noted that the final values for these parameters were determined through
the calibration process. The constituents with measured data include NH4, NO3/NO2,
phytoplankton (in terms of chlorophyll-a), PO4, alkalinity, and conductivity. These values were
directly incorporated into the concentration boundary condition file. There are no 1999 data
available to specify boundary conditions for ISS, CBOD, LDOM, RDOM, LPOM, RPOM,
periphyton/macrophytes, or TIC. In the W2 model the LDOM, RDOM, LPOM, and RPOM are
used to track the organic matter internally generated by algae death, so the boundary
concentrations for these four organic matter constituents were set to 0.0 for all the dates.
Periphyton/macrophytes were set to 0.0 g/m2 since they are assumed to be non-transportable (and
represented as such in the model). CBOD and TIC are two important constituents for which no
data were available. Therefore their values were estimated to be equal to 4.0 mg/L through the
calibration process. This value is within the range of the 2004 monitoring data. Since no data are
available to characterize the temporal trends of CBOD and TIC concentrations for the boundary
conditions, they were set constant throughout the year. TIC was obtained through model
calibration for pH. No data were available for ISS either, therefore it was set to 6.0 mg/L, which
is within the range of the 2004 monitoring data.
For the 2004 simulation, 2004 monitoring data at station LRDM were used. The CBOD
concentration was derived from monitored BODS data and a calibrated decay rate. Flow data at
Malone Dam were only available after April 29, 2004. Data for the remaining period were
derived from available data at Harpold Dam for 2004 and 1999. This derivation involved
calculating the ratio between concurrently available flows at Harpold and Malone Dams. This
ratio was also used to derive flows for Miller Creek, Big Springs, and Buck Creek.
DSW: Although outflow rates are available at Harpold Dam (only in winter months due to
backwater effects from Lost River Ranch dam in other months), they were not directly applied as
a flow boundary condition in the model, because the goal of the effort is to develop a predictive
modeling system (that is able to route flow from upstream to downstream). Since the discharge
flow at Harpold Dam is generally stage-dependent, a stage-discharge relationship was derived
based on the observed flow and water surface elevation data. The equation is as follows:
Q=aHb+c (Eq. 2-2)
where: Q is the spillway flow rate (m3/s)
His the water depth above the spillway crest (m)
a, and b are the derived coefficient and exponent, respectively
c is a leakage term
For Harpold Dam, the values of a, b, and c are: 28.3, 1.95, and 0.1, respectively.
TRIE: In total, three tributaries were represented for waterbody #1. These include Miller Creek,
Bonanza Springs, and Buck Creek. Although Bonanza Springs is not a direct surface tributary to
the Lost River, past reports identify that it contributes significant flow to the river.
19
-------
Model Configuration and Results
There are no data available to specify time series flow for Miller Creek, although a qualitative
description in a historical report stated that the flow in Miller Creek into the Lost River is
insignificant (USGS, 1999). Three flow data points are available from the USGS report, and the
average of these flows is 0.2 cms. Therefore, this value was assumed for the entire year in the
flow boundary condition file. Observed temperature and water quality constituent concentrations
for Miller Creek (using LRMC) were used to set up the temperature and concentration boundary
condition files. For the constituents without observed data for 1999 (ISS, CBOD, LDOM,
RDOM, LPOM, RPOM, macrophytes, and TIC), the same convention used for the US was
applied.
There are also insufficient data available to specify a complete flow time series from Bonanza
Springs. Woods and Orlob (1963) stated that Bonanza Springs contributes to the Lost River at an
average flow rate of 1.97 cms. Therefore, this flow rate was used for the entire year as the flow
boundary condition. The temperature and concentration boundary conditions were specified
utilizing observed data at BS. For the constituents without observed data (which are the same as
those listed for US), the same convention used for the US was applied.
No time series flow data are available for Buck Creek either, although three flow values from
USGS (1999) are available. The average of the three flow data points is 0.30 cms, thus this value
was assigned as the boundary condition for the entire year. The temperature and concentration
boundary conditions were specified based on observed data at BCBR. For the constituents
without observed data (same as above), the same convention used for the US was applied.
TRIE data were derived from the monitoring data at the Miller Creek station for 2004. These
data were applied to an extended period preceding and following (30 days in each direction) the
monitoring data. The BOD was calculated in the same way as for the US boundary. Flow
measurements for 2004 were averaged and used for the entire 2004 simulation period. TRIE
data for Big Springs and Buck Creek also used 2004 monitoring data. Once again the data were
extended backward for 30 days and forward for 30 days to create a relatively stable environment.
The 2004 measured flow data for Buck Creek were averaged to update the flow file with a
constant value for the 2004 simulation period.
DST: In general, the major sources of the distributed flows are agricultural irrigation
withdrawals, return flows, watershed runoff, groundwater interaction, and other unaccounted for
flows. In some cases these flows were negative and represented withdrawals from the system.
Ideally, these sources would be treated separately in the model since they represent different
spatial and temporal, as well as bio-chemical features. However, since no data are available to
support a detailed characterization/differentiation of these impacts, a "combined" approach was
applied to derive and configure the distributed flows. Using the "combined" approach, all the
distributed flow and pollutant sources were lumped together to form a single source/sink for the
waterbody.
For Waterbody #1, the initial estimate of the distributed flow rate was obtained by subtracting the
Malone Dam outflow and tributary flows from the Harpold Dam outflow. This initial estimated
flow was then iteratively refined through a comparison of the observed and simulated elevations
at the Harpold Dam, until a reasonable match between the data and model results were achieved.
There are no data available for assigning the time series of temperature and water quality
concentration to the distributed flow. It was assumed that during the summer low flow period
(which has been the focus of the study to date), the water quality at Lost River at Keller Bridge
(LRKB) is largely a direct reflection of the distributed flow water quality. During summer
periods, the flow from Malone Dam is generally insignificant since most of the water is used for
20
-------
Model Configuration and Results
irrigation. Therefore, the LRKB monitoring data for temperature and constituent concentrations
were used to represent the DST.
For 2004, LRKB data were also used as the basis for DST. The data were extended backward for
30 days and forward for 30 days to create a relatively stable environment (reducing the impact of
the 1999 boundary). The BOD andRPOM were calculated similarly to that for the US boundary.
WD: One withdrawal was configured for Waterbody #1, and this was the withdrawal at Harpold
Dam. The Harpold Dam withdrawal represents the pump-out operation, which is located
upstream of Harpold Dam, and all the pumping operations in the region. This pumping was
obtained through the model calibration process (since no pumping data were readily available),
with an aim at maintaining computational stability. A value of 4.0 cms was used in the model
during the irrigation season while a value of 0.0 cms was used for the remainder of the year.
2.3.3.2 Waterbody #2: Harpold Dam to RM 27 (Poe Valley Bridge)
USIFB: The upstream inflow, temperature, and constituents for Waterbody #2 are provided by
the flow from Waterbody #1 at Harpold Dam. In this study, Waterbody #1 was simulated
independently, and the resulting dam discharge flow rate, as well as the simulated temperature
and water quality, were saved in separate files. These were then read when simulating
Waterbodies #2 through #4.
DSIH: For Waterbody #2, the downstream boundary condition was set as an internal head
boundary condition at the Poe Valley Bridge (RM 27). This was continuously calculated and
updated throughout the model simulation process.
TRIE: One tributary was represented in the model for Waterbody #2, in order to account for the
contribution from the E Canal. Based on communication with ODEQ, there was assumed to be
relatively insignificant flow from the E Canal. Therefore, a 0.1 cms flow was assumed for this
inflow. The temperature and concentration boundary conditions were similar to those used for
upstream TRIE inputs. Considering the low flow rate from the E canal, this temperature and
concentration boundary condition do not have a significant impact on model results. The reason
for keeping this tributary as a boundary condition is for scenario evaluation, where the flow from
E Canal may require further evaluation.
DST: The distributed flow for Waterbody #2 was estimated in combination with Waterbodies #3
and #4. The distributed flow for Waterbodies #2, #3, and #4 were derived simultaneously,
because the only available water surface elevation data for flow balance calculation were
available for Wilson Reservoir. The initial estimate of the distributed flow rates for the three
waterbodies were obtained by scaling the distributed flow for Waterbody #1, using the segment-
length ratio between Waterbody #1 and Waterbodies #2, #3 and #4. These initial estimated flows
were then iteratively refined by comparing the observed and simulated elevations at Wilson Dam,
until a reasonable match between the data and model results were achieved. It was found that for
most of the dates negative flows needed to be assigned to distributed flows for Waterbodies #2,
#3, and #4, in order to achieve a reasonable match between observed and simulated surface water
elevation for Wilson Reservoir. These negative flow rates represent pumping activities,
vegetative evapotranspiration, seepage, measurement uncertainties, and other factors which are
not explicitly represented. Because using negative flows for the distributed inputs misrepresents
the load actually contributed by the watershed, the negative flows were removed from the DST
file and added to the withdrawal file (WD) for the Lost River Diversion Dam described for
Waterbody #4.
21
-------
Model Configuration and Results
There are no data available for assigning a time series of temperature and water quality
concentrations to the distributed flows, so it was assumed that the watershed runoff, agricultural
return flow, and ground water concentrations for Waterbody #2 are similar to those of Waterbody
#1.
WD: No withdrawals were configured for this waterbody.
2.3.3.3 Waterbody #3: RM 27 to Lost River before Wilson Reservoir
USIH: The upstream boundary condition for Waterbody #3 was provided through the internal
head boundary condition between Waterbodies #2 and #3 at Poe Valley Bridge (RM 27). This is
continuously calculated and updated throughout the model simulation process.
DSIH: For Waterbody #3, the downstream boundary condition was set as an internal head
boundary condition at the point before entering Wilson Reservoir. This was continuously
calculated and updated throughout the model simulation process.
TRIE: No tributary boundary conditions were configured for this waterbody.
DST: The derivation of distributed flow, temperature, and constituent boundary conditions for
Waterbody #3 were described in the section for Waterbody #2.
WD: No withdrawals were configured for this waterbody.
2.3.3.4 Waterbody #4: Lost River before Wilson Reservoir to Wilson Dam
USIH: The upstream boundary condition for Waterbody #4 was provided through the internal
head boundary condition between Waterbodies #3 and #4 at the Lost River before entering
Wilson Reservoir. This is continuously calculated and updated throughout the model simulation
process.
DSW: Historical data show that no spillage occurred at Wilson Dam during 1999, therefore a
constant downstream boundary condition was configured to represent leakage of 0.1 cms.
TRIE: One tributary, the F-l Canal, was represented in the model for Waterbody #4. Based on
communication with ODEQ, flow from the F-l Canal was assumed to be relatively insignificant,
therefore, a 0.1 cms flow was assumed. The temperature and concentration boundary conditions
were assigned based on upstream TRIE inputs. Considering the low flow rate from the F-l canal,
this temperature and concentration boundary condition do not have a significant impact on model
results. The reason for keeping this tributary as a boundary condition is for scenario evaluation,
where the flow from F-1 Canal may require further evaluation.
DST: The derivation of distributed flow, temperature, and constituent boundary conditions for
Waterbody #4 were described in the section for Waterbody #2.
WD: One withdrawal was configured for Waterbody #4 to represent the diversion to the Lost
River Diversion Channel. The historical observed diversion flow rate was converted into the
format required by W2 to form the corresponding boundary condition. This flow was adjusted
using the approach described in DST for Waterbody #2.
22
-------
Model Configuration and Results
The historical observed diversion flow rate for 2004 was used and adjusted as for 1999.
2.3.3.5 Waterbody #5: Wilson Dam to Anderson Rose Dam
USIFB: The upstream inflow, temperature, and constituents for Waterbody #5 are provided by
the flow from Waterbody #4 at Wilson Dam. In this study, Waterbodies #2, #3, and #4 were
simulated simultaneously, and the simulated temperature and water quality immediately upstream
of the dam were saved in separate files. These were then read during simulation of Waterbody #5
as the upstream boundary condition. The upstream flow boundary condition for Waterbody #5
was set the same as the DSW for Waterbody #4.
DSW: The downstream boundary condition for Waterbody #5 (at Anderson Rose Dam) was
defined using a rectangular weir equation. This was done because dimensions of Anderson Rose
Dam were available and thus flow could be related to crest length. A stage-storage relationship
wasn't available as was the case for Harpold Dam.
Q=2/3-Cd-L-j2g-H15 (Eq. 2-3)
where: Cj is the coefficient of discharge (-0.62)
L is the crest length
It should be noted that if one adds the seepage term c to Eq. 2-3, it is equivalent to Eq. 2-2. The
coefficients a and b in Eq.2-2 are the same as 2 / 3- Cd L- JZg and 1.5, respectively, in Eq. 2-
3. Using the crest length of the Anderson Rose Dam (324 ft) and the coefficient of discharge
(0.62) results in values for a and b equal to 180.72 and 1.5, respectively. A leakage term of 0.1
was used to account for leakage when water is below the dam crest.
TRIE: Four tributaries were represented in the Lost River model for Waterbody #5, including the
Station 48 turnout, Drain #1, Drain #5, and the City of Merrill STP.
Daily flow data are available for the Station 48 turnout and were used to specify the time series
boundary condition at this location. The temperature and concentration boundary conditions
were also specified based on observed data (at Station 48). For the constituents without observed
data (similar to those identified for upstream waterbodies), the same convention for setting up the
upstream boundary condition was followed.
It should be noted that the Station 48 turnout flow includes contributions from the Klamath River
and from the Lost River Diversion Dam (Wilson Reservoir contributions). For TMDL
development, distinguishing between these sources is important. If it is assumed that all the Lost
River Diversion Dam water goes into Station 48 and the remainder of water at Station 48 is from
the Klamath River (taking into account the contribution of the net flow from the Miller Hill
pump), the net flow provides an estimate of the amount of water from either the Klamath River
(when net flow is negative) or Lost River Diversion Dam (when net flow is positive) (BOR
personal communication). The equation below (Eq. 2-4) illustrates this:
NetFJow= LostRiverDiversionDam- Station^- MillerHillNet (Eq. 2-4)
23
-------
Model Configuration and Results
All the negative net flow, along with the corresponding constituent boundary conditions, at
Station 48 provide an estimate of the loading from the Klamath River. Figure 2-3 presents the
USGS Quads for this region.
Lost River
Diversion Dam
Figure 2-3. Junction of the Lost River, Lost River Diversion Channel, Drain #1, and Station 48
Turnout
No time-series flow data were available for Drain #1. The flow was therefore set equal to Drain
#5. The temperature and concentration boundary conditions were specified based on observed
data for Drain #1 (at DR1). For the constituents without observed data, the same convention for
setting up the upstream boundary condition was followed.
Four flow data points were available for Drain #5 during the 1999 irrigation period (USGS 1999).
The data points were averaged to generate the flow rate for Drain #5 in the model.
The temperature and concentration boundary conditions were specified based on observed data
for Drain #5 (at DR5). For the constituents without observed data, the same convention for
setting up the upstream boundary condition was followed.
The City of Merrill STP was also configured as a TRIE in the Lost River model. Temporally-
variable data were not available for this point source, therefore its contribution was represented
using an average flow of 0.14 mgd (0.006 cms) and water quality concentrations of 1.8 mg/L, 9.1
mg/L, 1.8 mg/L, 12.7 mg/L, 6.0 mg/L, and 830 uOhm/cm for PO4, NH4, NO2/NO3, CBOD, ISS,
24
-------
Model Configuration and Results
and conductivity, respectively (personal communication with ODEQ). Flow from this STP is
generally small compared to other contributing sources to the river at this location.
For 2004, daily flow data for the Station 48 turnout were used to specify the time series boundary
condition at this location. Monitoring data at this location were also used. The data were
extended backward for 30 days and forward for 30 days to create a relatively stable environment.
Drain #1 flow values (three) for 2004 were averaged and set for the 2004 simulation period. The
constituent and temperature data were updated with 2004 data collected atDRl.
Four flow values were available for Drain #5. These values were averaged to specify the flow
rate at Drain #5. The constituent concentrations and temperature at Drain #5 were updated
using 2004 data following the same method for deriving other updated boundary conditions.
DST: The distributed flows for Waterbody #5 were first derived using the same scaling approach
as for Waterbodies #2 through #4. This was then adjusted iteratively until the predicted Anderson
Rose Dam spillage matched the observed flow (namely in terms of average flow from May
through October). There are no data available for assigning the temperature and water quality
time series concentrations to the distributed flows, so it was assumed that the watershed runoff,
agricultural return flow and ground water concentrations for Waterbody #5 are similar to
conditions at Anderson Rose Dam. These conditions were therefore represented by the
monitoring data at the dam. It should be noted that for the entire stretch from Wilson Dam to the
Anderson Rose Dam, the ARDMUS station was the only station with considerable data. With
this configuration, the model tended to significantly underpredict the nitrate/nitrite for the
summer. To reduce this underprediction, the distributed boundary for nitrate/nitrite was
iteratively altered until the model reasonably reproduced the observed nitrate/nitrate at
ARDMUS.
A small A-canal contribution exists via the C-canal (approximately 0.08-0.42 cms) that feeds into
the downstream portion of this Waterbody (personal communication with BOR). According to
BOR an estimate of these spills may be available from the Klamath Irrigation District (KID).
Currently this contribution is considered within the DST. For TMDL analysis, the A-canal
contribution can be separated, if data can be obtained for the C-canal.
WD: One withdrawal was configured for this waterbody to represent the water diversion to the J
Canal. The historical observed diversion flow rate was converted into the format required by W2
to form the corresponding boundary condition.
The observed diversion flow rate in 2004 was converted into the format required by W2 to form
the corresponding boundary condition.
2.3.3.6 Waterbody #6: Anderson Rose Dam to Lost River before Tule Lake
USIFB: The upstream boundary condition for Waterbody #6 was provided by the internal flow
boundary condition between Waterbody #5 and #6 at Anderson Rose Dam.
DSW: The downstream boundary condition for Waterbody #6 (from Anderson Rose Dam to
Tule Lake), was set using a hypothetical flow control structure (using Eq. 2-2). An internal head
boundary was initially defined, however the abrupt change in the river's dimensions at this point
(i.e., from the Lost River to Tule Lake - since Tule Lake was represented as a single
25
-------
Model Configuration and Results
computational segment) caused model instability. As such, a hypothetical flow control structure
was used to essentially regulate flow into Tule Lake, yet still predict realistic flows into the lake.
The values for "a" and "b" were initially derived based on Manning's equation, and then refined
through calibration, in an effort to avoid computational instability. The final values were as
follows: a = 20.0, b = 1.67, and c = 0.0.
TRIE: No tributary boundary conditions were configured for this waterbody.
DST: The distributed flows for Waterbody #6 were derived based on the Tule Lake water
balance data in Tim Mayer's (2004) technical memorandum. It indicated that the average May to
October contribution to Tule Lake from Anderson Rose Dam spill is about 0.69 cms. Since the
simulated Waterbody #5 spill flow for the same period was 0.53 cms (close to, but smaller than
0.69 cms), a constant flow of 0.16 cms was assigned to the distributed flow boundary condition
for Waterbody #6 during this period. For other periods, the flow was obtained by scaling the
distributed flow for Waterbody #5 based on a segment length ratio. The temperature and
concentration boundary conditions were set based on monitoring data at station LREW.
WD: No withdrawals were configured for this waterbody.
2.3.3.7 Waterbody #7: Tule Lake
USIFB: The upstream boundary condition for Waterbody #7 was provided by the internal flow
boundary condition between Waterbodies #6 and #7.
DSW: Tule Lake was configured as a very coarse, one-segment model, with inflow and outflow
occurring in the same segment. The major outflow from Tule Lake includes irrigation diversions
through pumps R and 26, N-12 canal, Q and R canals, D plant pump, and evaporation (Mayer,
2004). However, no information was available to distinguish between rates for each of these
diversions. Since the primary outflow from Tule Lake to downstream Lost River segments is
through the D plant pump, the daily flow data at the D plant pump were used to represent the
outflow boundary condition. The remaining outflows were therefore implicitly represented in the
lumped DST.
The 2004 daily flow data at the D plant pump were used to set up an outflow boundary condition
for Tule Lake.
TRIE: No tributary boundary conditions were configured for this waterbody.
DST: The initial estimate of the distributed flows for waterbody #7 were derived based on the
Tule Lake water balance data in Tim Mayers' (2004) technical memorandum. This report listed
the monthly distributed inflow rate for Tule Lake for the period 1989 to 1998. These flow rates
were converted into cms and the monthly average values were used for each day of the month in
the flow time series. The initial estimate of the DST was iteratively adjusted to account for
ungaged outflow and inflow, and the final estimates were obtained once the simulated Tule Lake
surface water elevation correlated well with the measured elevation.
The DST also inherently includes contributions from the City of Tulelake STP. The STP
discharges to a normally dry drainage ditch adjacent to the plant, which is located approximately
10 miles from the Tule Lake Sump. Although the drainage ditch is hydrologically-connected to
the Tule Lake Sump, flows from the STP are relatively low and the discharge path to the sump is
not well-defined (personal communication with NCRWQCB). Therefore, it was deemed
26
-------
Model Configuration and Results
appropriate to consider the STP's contribution as a component of the DST for this waterbody.
The water quality and temperature of the DST was set to be the same as that of Waterbody #6,
considering their geographical proximity.
WD: No withdrawals were configured for this waterbody.
2.3.3.8 Waterbody #8: P-Canal
USIFB: The upstream boundary condition for Waterbody #8 was provided by the internal flow
boundary condition between Waterbody #7 and Waterbody #8, which is equivalent to the flow at
the D plant pump. Considering that Tule Lake was represented with extremely coarse spatial
resolution, the model result of Tule Lake is not considered to be sufficiently accurate to represent
the upstream boundary condition for Waterbody #8. Instead, the monitored data at station TLTO
is used to configure the USIFB water quality boundary condition to avoid transferring the
uncertainty in Tule Lake model result to downstream segments.
DSW: The downstream boundary condition for Waterbody #8 was set using a hypothetical flow
control structure (using Eq. 2-2). The values for "a", "b", and "c" were derived based on
Manning's equation as: a = 1.67, b = 1.72, and c = 0.0.
TRIE: No tributary boundary conditions were configured for this waterbody.
DST: The distributed flows for Waterbody #8 were set to zero based on the assumption that the
return flow, groundwater recharge, and other runoff in this relatively small drainage area are
discharged into either Tule Lake or Lower Klamath Lake.
WD: No withdrawals were configured for this waterbody.
2.3.3.9 Waterbody #9: Lower Klamath Lake
USIFB: The upstream flow and water quality boundary condition for Waterbody #9 was
provided by the internal boundary condition between Waterbodies #8 and #9.
DSW: Lower Klamath Lake was configured as a coarse, single-segment model with inflow and
outflow occurring in the same segment. Insufficient information was available for further
discretizing this water body into a higher resolution grid (to better represent spatial variability).
The major outflow from Lower Klamath Lake was represented by the discharge into Klamath
Straits Drain. Other withdrawals for irrigation were lumped into the DST since data were not
available for a detailed representation.
Flow was derived for 2004 based on the assumption that the outflow from Lower Klamath Lake
to Klamath Straits Drain is proportional to the pump rate at the E and EE pumps. Therefore,
ratios derived from 1999 data were applied to available 2004 data.
TRIE: ADY canal was the only tributary explicitly represented for Lower Klamath Lake. Flow
data provided by BOR were used to represent this tributary inflow. The concentration of NH4,
PO4, and NO3 were assigned based on Mayer, 2004. For other constituents, concentrations were
assumed to be similar to those for the Station 48 turnout (since contributions to both locations
originate in the Klamath River during the irrigation season).
27
-------
Model Configuration and Results
2004 flow data provided by BOR were used to represent this tributary flow. The concentration
and temperature data for ADY inflow was updated based on 2004 data using the same method
previously discussed.
DST: Distributed flow was used to balance the flow into and out of the lake. An iterative process
was implemented to estimate the distributed flow rate based on the time variable storage for
Lower Klamath Lake as reported in Burt and Freeman's report (Burt and Freeman, 2003).
Although this process resulted in some negative distributed flows, the flows were not assigned to
the outflows, as was the case for upstream segments. This is based on the assumption that the
negative flows are largely caused by withdrawals from the lake. Since the outflow from the lake
is used as the inflow for the downstream waterbody, the negative flows were not added to the
outflow in order to prevent introducing uncertainty in the downstream segment simulation.
Constituent concentrations and temperature for the distributed flow were set equal to those for
Waterbody #7, based on the assumption that return flow and groundwater recharge in this area are
similar to that of the Tule Lake area.
Distributed flow for 2004 was set to be the same as that of the 1999 since no data is available to
derive the Row balance for 2004. The water quality concentration was configured based on the
1999 condition but updated with the monitoring data at P-Canal for the summer period using the
same convention previously discussed. The temperature time series is set to be equal to that of
water body #7.
WD: No lateral withdrawals were configured for this waterbody.
2.3.3.10 Waterbody #10: Klamath Straits Drain at Pump E
USIFB: The upstream boundary condition for Waterbody #10 was provided by the internal flow
boundary condition between Waterbodies #9 and #10. . Considering that Lower Klamath Lake
was represented with extremely coarse spatial resolution, the model result of the Lake is not
considered to be sufficiently accurate to represent the upstream boundary condition for
Waterbody #10. Instead, the monitored data at station KSDSR is used to configure the USIFB
water quality boundary condition to avoid transferring the uncertainty in Lower Klamath Lake
model result to downstream segments.
DSW: The downstream boundary condition for Waterbody #10 was set using the flow time
series at pump E provided by BOR.
The downstream boundary condition for Waterbody #10 was set using the 2004 flow time series
at pump E provided by BOR.
TRIE: No tributary boundary conditions were configured for this waterbody.
DST: The distributed flow for Waterbody #10 was derived from the USIFB and DSW flows.
This was assumed to be an acceptable approach since monitoring data for distributed flows and
river stage were not readily available. The concentrations of the distributed flow were set based
on the data at KSDTR, and temperature is set to be the same as in the DST for waterbody #9.
The 2004 concentration time series is configured by using he monitoring data at KSDSR to
update the time series for 1999 model to represent the summer condition. The temperature time
series is set to be the same as for water body #9.
28
-------
Model Configuration and Results
WD: No lateral withdrawals were configured for this waterbody.
2.3.3.11 Waterbody #11: Klamath Straits Drain between E and F Pumps
USIFB: The upstream boundary condition for Waterbody #11 was provided by the internal flow
boundary condition between Waterbody #10 and #11. The simulated water quality and
temperature in Waterbody #10 were used directly as water quality and temperature boundary
conditions.
DSW: The downstream boundary condition for Waterbody #11 was set using the flow time
series at pump F provided by BOR.
The downstream boundary condition for Waterbody #11 was set using the 2004 flow time series
at pump F provided by BOR.
TRIE: No tributary boundary conditions were configured for this waterbody.
DST: The distributed flow for Waterbody #11 was derived from the USIFB and DSW flows.
This was assumed to be an acceptable approach since monitoring data for distributed flows and
river stage were not readily available. The concentration of the distributed flow were configured
based on the monitoring data at KSDPSF, and the temperature is set to be the same as waterbody
#10.
The 2004 concentration time series is configured by using the monitoring data at KSDM to
update the time series for 1999 model to represent the summer condition. The temperature time
series is set to be the same as for water body #10.
WD: No lateral withdrawals were configured for this waterbody.
2.3.3.12 Waterbody #12: Klamath Straits Drain between F Pump and Klamath
River
USIFB: The upstream boundary condition for Waterbody #12 was provided by the internal flow
boundary condition between Waterbodies #11 and #12. The simulated water quality and
temperature at the last segment in Waterbody #11 were used directly as water quality and
temperature boundary conditions.
DSW: The downstream boundary condition for Waterbody #12 was set using a hypothetical
flow control structure (using Eq. 2-2). The values for "a", "b", and "c" were derived based on
Manning's equation as: a = 1.696, b = 1.72, and c = 0.0.
TRIE: No tributary boundary conditions were configured for this waterbody.
DST: The distributed flows for Waterbody #12 were derived by scaling the distributed flow for
Waterbody #11 using the ratio between the lengths of Waterbodies #12 and #11. This assumes
that these two sections of the Klamath Straits Drain share similar flow generation and discharge
characteristics. The concentration of the distributed flow were configured based on the
monitoring data at KSD97, and the temperature is set to be the same as waterbody #11.
29
-------
Model Configuration and Results
The 2004 concentration and temperature time series were set to be the same as for waterbody
#12 since no data is available to derive the waterbody specific time series, and they two water
bodies are close to each other to warrant using the similar DST concentration and temperature
time series.
WD: No lateral withdrawals were configured for this waterbody.
2.3.3.13 Boundary Conditions Summary Table
Table 2-5 is presented below to summarize the extensive information presented in Sections
2.3.3.1 through 2.3.3.13. The Flow and Water Quality columns indicate the method used to
characterize the boundary conditions. "Data" refers to monitoring data. If presented alone, then
sufficient time-variable monitoring data were available to characterize conditions.
"Interpolation" indicates that limited monitoring data were available and thus values were
interpolated from available data. "Calibration" indicates that values were arrived at through
model calibration. In many situations, a combination of these methods was used. "Literature"
indicates that values were derived from the literature. "Weir Equation" denotes that an equation
was used to derive flow conditions. "Local Knowledge" indicates that values were designated
based on conversations with local experts and professionals familiar with the area and conditions.
"Model" indicates that the model generated values (typically for linkage between Waterbodies).
Table 2-5. Boundary Conditions and Linkages Summary Table
Waterbody Location Boundary Flow Water Quality
Condition
1
1
1
1
1
1
1
2
2
2
2
3
3
3
Malone Dam
From Malone to
Harpold Dam
Miller Creek
Big Springs
Buck Creek
Lost River (LR)
before Harpold
Dam
LR at Harpold Dam
LR downstream of
Harpold Dam
LR from Harpold to
RM27
E Canal
LR at RM 27
LR downstream of
RM27
LR from Ranch to
Wilson Res
LR before entering
Wilson Res
US
DST
TRIB
TRIB
TRIB
WD
DSW
USIFB
DST
TRIB
DSIH
USIH
DST
DSIH
Data
Calibration
Data,
Interpolation,
Calibration
Literature,
Interpolation,
Calibration
Data,
Interpolation,
Calibration
Calibration
Weir Equation
Model
Calibration
Local Knowledge
Model
Model
Calibration
Model
Data, Interpolation,
Calibration
Calibration
Data, Interpolation,
Calibration
Data, Interpolation,
Calibration
Data, Interpolation,
Calibration
Model
Model
Model
Calibration
Local Knowledge
Model
Model
Calibration
Model
30
-------
Model Configuration and Results
Waterbody Location Boundary Flow Water Quality
Condition
4
4
4
4
4
5
5
5
5
5
5
5
5
6
6
6
7
7
7
8
8
8
9
9
9
LR entering Wilson
Res
LR at Wilson
Reservoir
LR upstream of
Wilson Dam
F-1 Canal
LR at Wilson Dam
LR downstream of
Wilson Dam
LR from Wilson
Dam to Anderson
Rose Dam
LRatJ-Canal
LR at Anderson
Rose Dam
Station 48 Turnout
Drain #1
Drain #5
City of Merrill STP
LR downstream of
Anderson Rose
Dam
LR from Anderson
Rose Dam to Tule
Lake
LR before entering
Tule Lake
LR entering Tule
Lake
LR at Tule Lake
LR at Tule Lake
outlet
P Canal
downstream of Tule
Lake
P Canal
LR before entering
Tule Lake
P canal entering
Lower Klamath
Lake
Lower Klamath
Lake
ADY Canal
USIH
DST
WD
TRIB
DSW
USIFB
DST
WD
DSW
TRIB
TRIB
TRIB
TRIB
USIFB
DST
DSW
USIFB
DST
DSW
USIFB
DST
DSW
USIFB
DST
TRIB
Model
Calibration
Data
Local Knowledge
Data, Local
Knowledge
Model
Calibration
Data
Weir Equation
Data
Data,
Interpolation,
Calibration
Data,
Interpolation,
Calibration
Data
Model
Calibration
Weir Equation
Model
Calibration
Model
Calibration
Data
Model
Calibration
Data
Model
Calibration
Model
Local Knowledge
Model
Model
Calibration
Model
Model
Data, Interpolation,
Calibration
Data, Interpolation,
Calibration
Data, Interpolation,
Calibration
Data
Model
Calibration
Model
Model
Calibration
Model
Calibration
Model
Model
Calibration
Literature,
Interpolation,
31
-------
Model Configuration and Results
Waterbody Location Boundary Flow Water Quality
Condition
9
10
10
10
11
11
11
12
12
12
Lower Klamath
Lake entering
Klamath Straits
Drain
Klamath Straits
Drain leaving Lower
Klamath Lake
Klamath Straits
Drain from Lower
Klamath Lake to
Pump E
Klamath Straits
Drain at Pump E
Klamath Straits
Drain leaving Pump
E
Klamath Straits
Drain from Pump E
to Pump F
Klamath Straits
Drain at Pumps F
Klamath Straits
Drain leaving Pump
F
Klamath Straits
Drain from Pump F
to Discharge at
Klamath
Klamath Straits
Drain at Klamath
River
DSW
USIFB
DST
DSW
USIFB
DST
DSW
USIFB
DST
DSW
Data,
Interpolation
Model
Calibration
Data
Model
Calibration
Data
Model
Calibration
Data
Calibration
Model
Model
Calibration
Model
Model
Calibration
Model
Model
Calibration
Model
2.3.3.14 USIFB Flows
The USIFB (internal boundary conditions) for flow are presented below in Figures 2-4 and 2-5 to
demonstrate the highly variable nature of flow throughout the Lost River system throughout the
year.
32
-------
Model Configuration and Results
Flow (cms)
100.0
10.0
LRd/s ofHarpold
LR d/s of Anderson Rose
LRd/s of Wilson
LR into Tule Lake
Figure 2-4. USIFB Flows from Harpold Dam to Tule Lake
Flow (cms)
100.0
10.0 ---
1.0
0.1 ---
0.0
a>
O)
a>
a>
a>
a>
a>
O)
a>
a>
a>
O)
a>
O)
a>
O)
O)
O)
O)
O)
O)
O)
CO
P canal d/s of Tule Lake
KSD leaving LKL
KSD at pump F
Figure 2-5. USIFB Flows from Tule Lake through the Klamath Straits Drain
2.3.4 Initial Conditions
The W2 model requires specifying initial conditions in the control and bathymetry input files.
The control file specifies the initial temperature and constituents. Since there are no data
available to specify the initial conditions for all the constituents and water surface elevation, they
were specified based on best professional judgment. For this modeling study, the critical
33
-------
Model Configuration and Results
conditions were generally identified as the summer. Therefore, modeling to date has focused on
this period. The model simulation, however, begins on January 1st (of 1999 and 2004). While all
of 1999 was simulated, the end date for the 2004 simulation was day 240 (end of August). This
coincided with available flow and water quality data. As such, the initial conditions do not
significantly impact the model predictions during the critical period. Table 2-6 lists the initial
condition for temperature and all the simulated constituents.
Table 2-6. Initial Conditions
Constituent Initial Condition Value
Temperature
P04
NH4
N03/N02
Conductivity
Bacteria
LOOM
ROOM
LPOM
RPOM
Periphyton/macrophytes
DO
CBOD
ISS
Algae
TIC
Alkalinity
2.0 ฐC
0.1 mg/L
0.1mg/L
0.1 mg/L
300 us/cm
0.0 cfu
0.2 mg/L
0.2 mg/L
0.2 mg/L
0.2 mg/L
0.8 g/m2
10.0 mg/L
6.0 mg/L
0.0 mg/L
0.2 mg/L
1 2 mg/L
100 mg/L
34
-------
Model Configuration and Results
2.4 Modeling Assumptions and Limitations
All mathematical water quality models are a simplified representation of the very complex real
world. The Lost River system is certainly no exception. It is a highly modified environmental
system driven largely by irrigation operations, and it exhibits tremendous biological activity. Due
to a lack of quantitative data to describe many aspects of the system, a number of key
assumptions were made during model development. The combination of the lack of data and
assumptions made, also lead to inherent limitations associated with the effort.
2.4.1 Assumptions
The major underlying assumptions associated with Lost River model development are as follows:
Weather conditions do not vary significantly over the entire modeling domain. If they
do, the impact on resulting water quality is assumed not to be significant.
The impact of sediment transport and siltation on channel geometry is not significant,
therefore the same bathymetric configuration can be used for different scenario
simulations.
The initial condition and the boundary conditions set for the winter and early spring
period do not have a significant impact on the simulated water quality during the critical
summer and early fall periods. This assumption permits assigning the initial conditions
and winter/early spring boundary conditions using best professional judgment, without
impairing the model performance for the critical period.
Time series flow data were not available for all drains, tributaries, and withdrawals.
Reliable time series flow data were also not available for many monitoring locations
along the length of the Lost River. In light of the limitations, it was assumed that
tributary flows could be reasonably represented through interpolation based on limited
flow measurements. Additionally, drains and withdrawals were assumed to be
reasonably derived through the calibration process. Where flow monitoring data were
available along the length of the river, the data were generally assumed to be appropriate
(due to the absence of data indicating otherwise), except where backwater effects were
prevalent.
The distributed flows for P-canal (Waterbody #8) were set to zero based on the
assumption that the return flow, groundwater recharge, and other runoff in this relatively
small drainage area are discharged into either Tule Lake or Lower Klamath Lake.
Water quality associated with the distributed flow inputs to the model was initially
specified based on monitoring data within the Lost River itself. Due to the lack of
quantitative data for characterizing agricultural pumping, return flow and other unknown
sources and sinks, it was assumed that the water quality associated with the distributed
flow is similar to the water quality in the Lost River where the distributed flow
discharges. Tetra Tech obtained data from BOR for a number of pumps in the basin
(including D, E, EE, F, and FF), however data for the numerous remaining pumps in the
basin could not be obtained. All irrigation districts were contacted, however data were
not available.
One phytoplankton species and one macrophyte species are sufficient for representing the
overall primary production and nutrient interactions in the system.
Topographic shading effects on water temperature and algal growth are insignificant. In
scenario runs, the effects of riparian vegetation shading can be accounted for by using a
scaling factor for solar radiation intensity.
35
-------
Model Configuration and Results
Alkalinity is conservative (as stated in CE-QUAL-W2 manual). Therefore, no internal
sources or sinks were considered.
All the organic matter in the water column (and that from other sources) has the same
stochiometric ratio.
The impact of zooplankton and benthic creatures do not have a significant impact on the
algal dynamics and nutrient recycling.
The water quality gradient within Tule Lake and Lower Klamath Lake is insignificant,
therefore each can be considered as a single, mixed segment.
2.4.2 Limitations
Lost River model limitations include the following:
The capability of a model is constrained by the availability and quality of data. Built on
limited data, the Lost River model is not expected to be able to mimic the exact timing
and location of all water quality conditions. However, the model can be used to represent
the overall water quality trends in response to external loading and internal system
dynamics.
The model does not explicitly represent the spatial and temporal distribution of
agricultural return flows and pump operation due to a lack of quantitative data.
Therefore, it cannot be used to evaluate the potential impact of changing specific
pumping schemes or the locations and timing of return flows. It can, however, be used to
evaluate the overall impact of varied pumping flow/return flow, as well as the associated
loadings (in a lumped manner). The goal of the model is to predict the general response
of the river and its impoundments to spatially and temporally variable load inputs (though
not necessarily discrete inputs) and to evaluate the impact of hypothetical load changes
relative to current and historical conditions. The model can also be used to evaluate
water quality criteria develop TMDLs.
The winter and early spring boundary conditions for the distributed tributary flows were
based on conditions for the summer (where more data are available), and thus might not
be reliable. This, however, won't significantly decrease the reliability of the model for
the summer critical period simulation.
The model does not simulate multiple species for phytoplankton and macrophytes.
Therefore, this model is currently not suitable for evaluating competition among multiple
species or evolution of the aquatic algal communities and their interaction with nutrients.
The model does not simulate water quality processes within the Lost River Diversion
Channel, however, it can be used to transfer to and from a Klamath River model.
Due to the lack of a direct linkage between organic matter loading and SOD and benthic
nutrient flux, the model in its present stage, is not suitable for evaluating the long-term
impact of load reductions on SOD.
Neither zooplankton nor benthic animals are simulated in the model, hence, there may be
some uncertainty in the simulation of algal dynamics and nutrient cycling.
Bacteria in CE-QUAL-W2 is simulated as a general constituent with simple first-order
die-off.
Tule Lake and Lower Klamath Lake were treated as well-mixed segments, thus the model
cannot be used to accurately evaluate the local water quality conditions (i.e., spatial or
depth-variability) associated with the water bodies. This also introduces uncertainties in
representing segments downstream of the lakes, including the P Canal and the Klamath
Straits Drain.
36
-------
Model Configuration and Results
3.0 MODEL TESTING
Once the Lost River W2 model was configured, a calibration was performed at multiple locations
throughout the system. Calibration refers to the adjustment or fine-tuning of modeling
parameters to produce an adequate fit of the observations. The sequence of calibration for the
Lost River W2 model involved checking the water budget first using the water surface elevation,
then calibrating hydrodynamics using temperature and conductivity data, and finally calibrating
water quality using available monitoring data.
In this study, the model was tested for two separate years (1999 and 2004) to increase model
reliability. 1999 was the year with the most concurrent data for model configuration. The
monitoring data for this year also exhibit significant water quality impairment during the summer
critical period, and thus provide an excellent basis for testing the model's capability of capturing
extreme conditions, which are of concern for TMDL development. 2004 was selected because a
summer sampling effort was conducted by ODEQ, NCRWQCB, EPA Region 10, and EPA
Region 9 to support modeling. The Lost River model was first calibrated for 1999, and then 2004
data were used to corroborate the model.
3.1 Monitoring Locations
In order to fully calibrate the Lost River model, a significant amount of time-varying monitoring
data, with sufficient longitudinal resolution (and vertical resolution in the impoundments), are
required. The data obtained for the Lost River are listed in Table 3-1. Stations in this list that are
located on the main-stem of the Lost River and its impoundments supported model calibration.
These locations are depicted in Figure 3-1 and discussed in subsequent sections. Other stations in
the list were used to prescribe boundary conditions.
37
-------
Model Configuration and Results
Table 3-1 Modeling Support Data and Data Sources
Data Type Station/Location Start Date End Date Frequency Source
Flow
Water Quality
Malone Dam
Keller Bridge
Harpold Dam
Pump E-EE
Pump F-FF
Miller Cr
Lost River Diversion
Channel
Station 48
Anderson Rose Dam
J-Canal
Plant-D
ADY Refuge
KSFLOW @ Stateline
BCBR (BC)
BS
DR1
DR5
KSD @ Stateline
KSD @Tunnel
LRAR (ARDMUS)
LREW
LRHDB (LRHD)
LRKB
LRMD (LRDM)
LRWRC (WDUS)
LRMC (MC)
ST48
LREW
TLTPD
TLTO
LKLO
KSDSR
KSDTR (Pump E-EE)
KSD97 (KSDM)
KSDPSF(Pump F-FF)
GB
2/2/87
4/1 7/87
1/1/87
1/1/87
1/1/87
2/1/87
1/1/87
1/16/87
1/16/87
1/16/87
1/16/87
3/23/91
1/1/87
3/2/99
5/1 2/99
5/1 2/99
1/28/99
3/23/99
3/23/99
3/2/99
1/28/99
3/2/99
5/1 2/99
5/1 2/99
1/28/99
3/2/99
5/1 2/99
1/28/99
1/13/99
3/23/99
1/13/99
7/30/99
3/23/99
1/13/99
3/10/99
6/15/04
9/11/2004
5/22/02
5/09/04
10/17/04
9/23/04
10/1/03
3/28/04
9/1 2/04
9/1 2/04
9/1 2/04
9/1 2/04
2/22/04
10/17/04
7/27/04
6/15/04
7/27/04
7/27/04
7/27/04
11/30/99
7/27/04
7/27/04
7/26/04
7/27/04
7/27/04
7/27/04
7/27/04
7/27/04
7/27/04
06/03/03
11/14/00
09/20/00
7/27/04
11/30/99
7/27/04
06/20/01
7/27/04
Daily
Daily
Daily
Daily
Daily
Daily
Daily
Daily
Daily
Daily
Daily
Daily
Daily
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
BOR
BOR
BOR
BOR
BOR
BOR
BOR
BOR
BOR
BOR
BOR
BOR
BOR
BOR, ODEQ,
NCRWQCB
BOR, ODEQ,
NCRWQCB
BOR, ODEQ,
NCRWQCB
BOR, ODEQ,
NCRWQCB
BOR, ODEQ,
NCRWQCB
BOR
BOR, ODEQ,
NCRWQCB
BOR, ODEQ,
NCRWQCB
BOR, ODEQ,
NCRWQCB
ODEQ, NCRWQCB
BOR, ODEQ,
NCRWQCB
BOR, ODEQ
BOR, ODEQ
BOR, ODEQ,
NCRWQCB
BOR, ODEQ,
NCRWQCB
BOR
BOR
BOR
BOR, ODEQ,
NCRWQCB
BOR
BOR
BOR, ODEQ
ODEQ, NCRWQCB
38
-------
Model Configuration and Results
Data Type
Station/Location
LRCF
LRKB
KBNE
KB
we
YD
LRY
PC
Start Date
6/15/04
6/15/04
6/15/04
6/15/04
6/15/04
6/15/04
6/15/04
6/15/04
End Date
7/27/04
7/27/04
7/27/04
7/27/04
7/27/04
7/27/04
7/27/04
7/27/04
Frequency
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Grab
Source
ODEQ, NCRWQCB
ODEQ, NCRWQCB
ODEQ, NCRWQCB
ODEQ, NCRWQCB
ODEQ, NCRWQCB
ODEQ, NCRWQCB
ODEQ, NCRWQCB
ODEQ, NCRWQCB
I- SO M
0 Calibration Stations
/\/Modeling Segments
/\/' Lost River Basin Streams (rf3)
Figure 3-1. Calibration Locations for Lost River Modeling
39
-------
Model Configuration and Results
3.2 Hydrodynamic Simulation
3.2.1 Hydraulic Parameter Designation
Default hydraulic parameters were used to run the model initially. With each model run, these
parameters were adjusted to achieve a unique set of coefficients that best represented the system
under all conditions. Cole and Wells (2003) reported that previous experience has shown that the
default values produce remarkably accurate temperature predictions for a wide variety of systems,
provided accurate geometry and boundary conditions were specified (1995). Table 3-2 shows the
calibration coefficients for the hydrodynamics and temperature.
Table 3-3 Calibration Coefficients for Hydrodynamic Simulation
Coefficient Name Value
Longitudinal eddy viscosity
Longitudinal eddy diffusivity
Manning's coefficient
Wind sheltering coefficient
Solar radiation absorbed in surface layer
Sediment temperature
Coefficient of bottom heat exchange
[AX]
[DX]
[MANN]
[WSC]
[BETA]
[TSED]
[CBHE]
1 m'/s
1 m'/s
-0.02
0.8
0.45
11.5ฐC
0.3
In general, the bathymetry and a balanced water budget for the Lost River system were the most
crucial factors in the hydrodynamic simulation. Most parameters were found to have little or no
effect on hydrodynamics, with the exception of Manning's Coefficient, n [MANN]. It was found
that too high or too low a value for MANN caused model instability. Therefore, a moderate value
or approximately 0.02 was used, and this reasonably represented the physical characteristics of
the system, while maintaining model stability. The wind-sheltering coefficient [WSC] was set to
a constant value since there is no visible vegetation throughout the majority of the watershed that
could potentially modify the wind velocity.
3.2.2 Water Balance and Water Surface Elevation Calibration
Historical water surface elevation data (daily values) were available at Harpold Dam and Wilson
Dam, therefore, these two locations were used to derive the flow balance for the first four
waterbodies of the Lost River model. In addition to the water surface elevation data, data on the
release rates from Harpold Dam were available from BOR. The hydrodynamic portion of the W2
model was first run to verify the water budget. This involved comparing predicted reservoir
elevations with observed water surface elevations.
The simulation was implemented in a piece-wise manner. As the first step, Waterbody #1 was
independently simulated using the boundary conditions described in the model configuration
section. With the upstream and tributary boundary conditions and the downstream weir equation
fixed, the major unknown flow sources/sinks were the distributed flows (e.g., irrigation water
withdrawal or return flows, as well as watershed runoff). The major task for the flow balance
calibration was thus to derive this unknown component of the flow by simulating water surface
elevation at Harpold Dam and trying to match the observed elevations. Starting from the initially
estimated distributed flow, the values were iteratively adjusted until a reasonable match between
the simulated and observed elevation were obtained. Figures A_1999-l, A_1999-2, and A_2004-
40
-------
Model Configuration and Results
1 (for both 1999 and 2004) in Appendices A_1999 and A_2004 display the simulated water
surface elevation and dam discharge calibration versus the observed data at Harpold Dam.
After the water balance for Waterbody #1 was completed, the model simulated dam discharge at
Harpold Dam was incorporated into the simulation model for Waterbodies #2 to #4 as the
upstream flow boundary condition. For these three waterbodies, the major unknown
sources/sinks of flow were also from the distributed flows. With the upstream inflow and
tributary flows set, and the Lost River Diversion Channel withdrawal fixed, the distributed flow
rates for the three waterbodies were iteratively adjusted until the simulated water surface
elevation at Wilson Dam matched the observed data reasonably well. Figures A_ 1999-3 and
A_2004-2 (for 1999 and 2004) in Appendices A_1999 and A_2004 plots the simulated Wilson
Reservoir surface elevation against the observed data.
The water balances for Waterbodies #5 and #6 were calibrated using Tule Irrigation District data
(for 1999) and based on Mayer's (2004) technical memorandum. Figures A_1999-4 and A_2004-
3 plot the spillage at Anderson Rose Dam. For Waterbody #7, the distributed flow was obtained
by calibrating the simulated water surface elevation at Tule Lake against the observed data
provided by BOR. Figure A_1999-5 plots the simulated Tule Lake elevation versus the observed
data.
For Waterbody #8, a special water balance calculation was not implemented. The distributed
flow rate was set to zero due to the relatively small drainage area and the assumption that most
distributed flow in the area is accounted for in the Tule Lake and Lower Klamath Lake modeling
segments.
Major inflows for Waterbody #9 were configured using the P Canal discharge and the ADY
Canal inflow. Outflow was configured using the observed flow rate at the Klamath Straits Drain
(at the state line). With the known temporal variation of storage volume from the Burt and
Freeman (2003) report, the distributed flow was derived by matching the simulated volume with
the observation data. Figure A_ 1999-6 plots the comparison.
There were no water surface elevation data available for Waterbodies #10, #11, and #12.
Distributed flow for these three waterbodies was derived based on the assumption that flow in the
Klamath Straits Drain is approximately balanced. That is, the distributed flow for Waterbodies
#10 and #11 were calculated by subtracting the downstream outflow rates from the upstream
inflow rates. This difference in flow was adjusted minimally, in order to maintain computational
stability. For Waterbody #12, the distributed flow was calculated by scaling the distributed flow
for Waterbody #11 (based on the ratio of the lengths of each waterbody).
3.2.3 Hydrodynamic Model Evaluation with Temperature Data
After the water budget was calibrated, the next step was to reproduce the observed temperature
data in the system. A piece-wise evaluation of temperature predictions was conducted as for the
flow calibration process. However, for temperature, no adjustment to default parameters was
necessary. The simulated water temperatures were plotted against the measured data for 1999
and 2004 and are shown in Figures B_1999-l to B_1999-10 and B_2004-l to B_2004-8.
41
-------
Model Configuration and Results
3.2.4 Further Hydrodynamic Model Evaluation with Conductivity Data
The performance of the hydrodynamic model in simulating mass balance and transport was
further evaluated using conductivity as a conservative tracer. The comparison of model
predictions to observations is shown in Figures C_1999-l to C_1999-10 and C_2004-l to
C_2004-8.
The relatively poor correlation for Tule Lake and other downstream stations is likely caused by
the extremely coarse resolution used for Tule Lake and Lower Klamath Lake, which limit the
model in predicting the flashy nature of the system. In Figure C_l999-6, the data show a peak
conductivity in May and June with a decline after June. The model, however, predicts an
increase in conductivity from May to October before it starts to decline. This is due to the
configuration of Tule Lake as a completely mixed segment, which causes any loading entering
Tule Lake to be instantly diluted by its significant volume. Therefore, when the conductivity in
the distributed flow is set equal to that in Tule Lake, the simulated concentration experiences a
significant time lag in reflecting the loading impact. At the same time, the summer evaporation
causes the conductivity to increase, resulting in the simulated conductivity showing a rising trend.
If sufficient data were available to configure a higher resolution representation of Tule Lake, the
model would likely simulate a more appropriate response at locations receiving external loading.
3.3 Water Quality Simulation
Once the temperature and conductivity calibrations were completed, the next step was to perform
the water quality model simulation and calibration. Water quality model results, such as observed
dissolved oxygen, ammonia, nitrite/nitrate nitrogen, orthophosphate, chlorophyll a, and pH were
the key calibration parameters. The water quality calibration was also a piece-wise process,
which involved first calibrating the upstream waterbodies, and then using the resulting flow and
predicted concentration time series (together with the watershed and other tributary inputs) to
drive the downstream waterbody simulations.
The calibration of the water quality model was implemented through tuning major kinetic
parameters such as algal growth rate, death rate, nitrification/denitrification rates, CBOD/organic
matter decay rates, and SOD rates. The overall goal was to most accurately match observed data
while maintaining consistency among all the waterbodies. As a result, the kinetic parameter
values were kept the same for all the waterbodies, with the exception of Tule Lake and Lower
Klamath Lake (where macrophyte kinetics and benthic flux parameters varied) and SOD rates.
This approach provides confidence in applying the calibrated parameters to other time periods
and for use in alternative scenarios. Tables 3-4 through 3-6 list the calibrated kinetic values for
Waterbodies #1 through #12. Note that values in parentheses are for Tule and Lower Klamath
Lakes.
Table 3-4. Nutrient Input Parameters Used for the Lost River
Parameter Description Units Value Typical
Literature
Value1
PO4R
PARTP
ORGP
Sediment release rate of
phosphorous
Phosphorous partitioning
coefficient for suspended solids
Fraction of phosphorous in
fraction of
SOD
-
-
0.0030
0.000
0.011
0.001 to 0.03
0.000
0.011
42
-------
Model Configuration and Results
Parameter Description Units Value Typical
Literature
Value1
ORGN
NO3DK
NO3T1
NO3T2
N03K1
NO3K2
NH4DK
NH4R
NH4T1
NH4T2
NH4K1
NH4K2
organic matter
Fraction of nitrogen in organic
matter
Nitrate decay rate
Lower temperature for nitrate
decay
Upper temperature for nitrate
decay
Lower temperature rate multiplier
for nitrate decay
Upper temperature rate multiplier
for nitrate decay
Ammonium decay rate
Sediment release rate of
ammonium
Lower temperature for ammonium
decay
Upper temperature for ammonium
decay
Lower temperature rate multiplier
for ammonium decay
Upper temperature rate multiplier
for ammonium decay
-
day'
ฐC
ฐC
-
-
day"'
fraction of
SOD
ฐC
ฐC
-
-
0.080
0.05
5.0
25.0
0.10
0.99
0.08
0.001
5.0
25.0
0.10
0.99
0.080
0.05 to 0.1 5
5.0
25.0
0.10
0.99
0.00 to 0.80
0.000 to
0.400
5.0
25.0
0.10
0.99
Cole and Wells (2003); Chapra, S.C. (1997)
Table 3-5. Phytoplankton Input Parameters used for the Lost River
Parameter Description Units Values of Algal Typical
Groups Literature
Value1
AG
AR
AE
AM
AS
AHSP
AHSN
ASAT
AT1
AT2
ATS
AT4
AK1
Growth rate
Dark respiration rate
Excretion rate
Mortality rate
Settling rate
Phosphorous half-saturation
coefficient
Nitrogen half-saturation
coefficient
Light saturation
Lower temperature for
minimum algal rates
Lower temperature for
maximum algal rates
Upper temperature for
minimum algal rates
Upper temperature for
maximum algal rates
Lower temperature rate
multiplier for minimum algal
rates
day"'
day'
day"'
day'
day"'
g.nr*
g.m"J
W.rrTJ
ฐC
ฐC
ฐC
ฐC
1.1
0.10
0.01
0.03
0.20
0.002
0.01
100
5.0
12.0
25.0
30.0
0.1
0.2 to 9.0
0.01 to 0.92
0.01 to 0.044
0.03 to 0.30
0.001 to 13. 20
0.001 to 1.520
0.01 to 4.32
10 to 150
N/A
N/A
N/A
N/A
N/A
43
-------
Model Configuration and Results
Parameter Description Units Values of Algal Typical
Groups Literature
Value1
AK2
AK3
AK4
ALGP
ALGN
ALGC
ACHLA
Lower temperature rate
multiplier for maximum algal
rates
Upper temperature rate
multiplier for minimum algal
rates
Upper temperature rate
multiplier for maximum algal
rates
Phosphorous to biomass ratio
Nitrogen to biomass ratio
Carbon to biomass ratio
Algae to chlorophyll-a ratio
-
-
-
-
0.99
0.99
0.1
0.011
0.080
0.45
110
N/A
N/A
N/A
0.011
0.080
0.45
110
Literature values are from the CE-QUAL-W2 Users Manual which compiled data from a range of sources. The only
exception is the stoichiometric coefficient, which was derived from Chapra, 1997.
Table 3-6. Macrophyte Input Parameters used for the Lost River
Parameter Description Units Values of Algal Typical
Groups Literature
Value1
EG
ER
EE
EM
EB
EHSP
EHSN
ESAT
ET1
ET2
ET3
ET4
EK1
EK2
EK3
EK4
Growth rate
Dark respiration rate
Excretion rate
Mortality rate
burial rate
Phosphorous half-saturation
coefficient
Nitrogen half-saturation
coefficient
Light saturation
Lower temperature for
minimum macrophyte rates
Lower temperature for
maximum macrophyte rates
Upper temperature for
minimum macrophyte rates
Upper temperature for
maximum macrophyte rates
Lower temperature rate
multiplier for minimum
macrophyte rates
Lower temperature rate
multiplier for maximum
macrophyte I rates
Upper temperature rate
multiplier for minimum
macrophyte rates
Upper temperature rate
multiplier for maximum
macrophyte rates
day"1
day1
day"1
day"'
day"1
g.m"J
g.m"3
W.m"J
ฐC
ฐC
ฐC
ฐC
0.75
0.07
0.01
0.02
0.001
0.002
0.014
150
5.0
18.0
25.0
30.0
0.1
0.99
0.99
0.1
N/A
N/A
N/A
N/A
N/A
N/A
N/A
75-150
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
44
-------
Model Configuration and Results
EP
EN
EC
1 ECHLA 1
P h os ghojjojjsjo^
Nitrogen to biomass ratio
Carbon to biomass ratio
_jQ,lg,aj3jto^
0.011 ; N/A
0.080 ; N/A
0.45 N/A
__^ 55
N/A
Literature values are from the CE-QUAL-W2 Users Manual which compiled data from a range of sources and the
example models. The only exception is the stoichiometric coefficient, which was derived from Chapra, 1997.
Different SOD rates were assigned for different waterbodies and refined through the calibration
process. SOD monitoring data at Harpold Dam (a value of 3.8 g/m2/day) was used to derive SOD
rates from Malone Dam to Harpold Dam. It was assumed that the SOD increased linearly from
upstream to downstream, with an initial value of 1.0 g/m2/day set for the most upstream segment.
The SOD rate measured at Wilson Dam (2.5 g/m2/day) was used in a similar manner to derive
SOD rates from Harpold Dam to Wilson Dam. For the remaining waterbodies, a base value of
2.0 g/m2/day was initially estimated and then adjusted through calibration. Figure 3-2 presents
the SOD rates used for each modeling segment. The waterbody divisions are indicated at the top
of the plot.
Waterbody
2-4
10-12
50 100 150 200
Modeling Segment (From Upstream to Downstream)
250
Figure 3-2. SOD Rate Variability in the Model
Appendices D_1999 and D_2004 present the model-simulated nutrient, DO, and chlorophyll a
concentrations along with the observed data at the monitoring stations described above. Note in
the figures that the "max", "min", and "mean" refer to the vertical maximum, minimum, and
mean values of the corresponding constituents. If these values are indiscernible on the plots, then
there is no vertical stratification occurring; otherwise, vertical stratification exists. Time-series
plots of modeled versus observed data were the primary method of calibration for the Lost River
45
-------
Model Configuration and Results
Model. They provide more insight into the nature of the system than statistical comparisons -
particularly in light of the major data limitations associated with the Lost River. The following
text briefly discusses the calibration performance at each station.
3.3.1 Lost River at Keller Bridge (LRKB)
As shown in Figures D_1999-l and D_2004-l, the model results at LRKB follow the general
trends demonstrated by the observation data. For DO, the model simulated relatively high
concentrations during the winter, spring, and fall periods, but reached lower values during
summer. Also, the model results show that there is no vertical stratification in the section.
Since DO is plotted at a sub-daily frequency, the diel fluctuation is also apparent from the
simulation results. Diel fluctuation at LRKB is not significant, however. The simulated nutrient,
chlorophyll a, and pH values appear to follow the general observed trends very well, indicating
that the model reasonably represents mass balance and water quality interactions for this section.
The simulated spikes of NO3 and PO4 during March (both 1999 and 2004) are due to boundary
conditions at Malone Dam, where data show high concentrations of these two constituents. Due
to lack of data for 2004, boundary conditions during spring and winter of the 2004 simulation
were set equal to those in 1999, causing the spikes to occur in both years.
The calibration analyses indicated that chlorophyll a variations are dominated by the distributed
flow boundary condition. This was apparent from both the 1999 and 2004 simulations.
The 2004 model simulated significantly smaller diel fluctuations of DO at LRKB because the
growth of macrophytes is limited by nitrogen (and thus don't result in a major DO swing). This
relatively low nitrogen level in the model is caused by the boundary condition representation. In
the 2004 model, the DST boundary condition was set using monitoring data for several drains
(which were assumed to represent the distributed flow water quality). This monitoring showed
low nitrogen concentrations in the drains. Therefore the resulting in-stream nitrogen
concentrations were relatively low (thereby limiting macrophyte growth). It is expected that if
the distributed water quality boundary condition was set using the same approach used for the
1999 model (estimated based on LRKB data), the simulated nitrogen limiting condition might be
alleviated.
3.3.2 Lost River at Harpold Dam (LRHD)
Similar to LRKB, the model results at LRHD follow the general observed trends reasonably well
(Figures D_1999-2 and D_2004-2). No vertical stratification is apparent from either the model
results or the observed data at this location either. For DO, the model simulated relatively high
concentrations during winter, spring, and fall, and simulated relatively low values during summer.
The diel fluctuation of DO and NH4 is more prominent than in LRKB, suggesting a stronger
biological impact on water quality. This conclusion is supported by the aquatic vegetation survey
conducted in 2004, which shows that the macrophyte biomass at LRHD is about 6 times as high
as that at LRKB. The model results show that the simulated peak macrophyte biomass at LRHD
is about 7 times as high as that at the LRKB. The major factor affecting the spatial distribution of
macrophytes is flowthe relatively high flow velocity at LRKB has limited the growth of
macrophytes while the relatively stagnant conditions at Harpold Dam supports macrophyte
growth.
46
-------
Model Configuration and Results
The simulated spikes of NO3 and PO4 during March are due to boundary conditions at Malone
Dam, where data show high concentrations of these two constituents in early March. As shown
in Figure D_1999-2, the model under predicts NO3 during summer 1999. This is because the
distributed flow concentration was set equal to those at the LRKB, which has lower summer NO3
concentrations than at LRHD. Since the distributed flow boundary condition dominates during
the summer, the simulated NO3 would directly reflect the relatively low NO3 in the distributed
flow boundary condition. Improved simulation might be achieved by using spatially variable
distributed flow boundary conditions for upper and lower section of Waterbody #1, however
sufficient data are not currently available. The same phenomena exist for the 2004 model.
3.3.3 Lost River at RM 27 (HPDS2)
Modeling results at RM 27 are presented in the Appendix, however no calibration discussion is
presented because only one data point was available for comparison.
3.3.4 Lost River at Wilson Reservoir (LRWRC)
LRWRC is located inside Wilson Reservoir. The model results are compared with the observed
data in Figures D_1999-3 and D_2004-4. The plots demonstrate that the model successfully
reproduces the vertical stratification as well as the general seasonal trend for DO. The observed
data show several very low DO concentrations during the end of 1999, however the model cannot
reproduce this phenomenon since during winter the thermal stratification disappears. Hence the
entire water column should be well mixed with oxygen replenished from the atmosphere. The
very low DO in this period is likely caused by a highly site-specific feature that cannot be
characterized and represented by the model, or it is caused by erroneous monitoring data.
Another possible cause is that Wilson Reservoir reached a "quick inverse" stratification during
the winter, when water at 4 ฐC remained at the bottom and colder water floated on top, forming a
stratified condition. To develop a model that is capable of simulating this type of delicate thermal
structure requires highly accurate bathymetric, flow, and atmospheric data. Such data are not
available for the current modeling study.
As for the nutrients, the data show significant temporal fluctuation while the model simulates a
relatively smooth transition. The large fluctuation of nutrient concentrations can most likely be
attributed to sporadic loading to the system, which cannot be fully characterized in existing data
sets. Therefore, it is not expected that the model can reproduce these highly time variable
features of the system. The model does, however, represent the general trends seen in the data
very well. Again, the simulated spikes of NO3 and PO4 during March are due to the boundary
conditions at Malone Dam, where data show high concentration of these two constituents.
3.3.5 Lost River at Anderson Rose Dam (LRAR/ARDMUS)
The model results at LRAR/ARDMUS follow the general observed trends reasonably well
(Figures D_l999-4 and D_2004-5). For DO, the model simulated relatively high concentrations
during winter, spring, and fall, and simulated relatively low values during summer, with moderate
vertical stratification and diurnal fluctuation. The model captures the chlorophyll a seasonal
variability well. The simulation results for this portion of the Lost River system illustrate the
influence of the Lost River Diversion Channel. For example, the peak in chlorophyll a that
occurs during the summer of 1999 was primarily caused by loading from the Station 48
discharge. The simulated spikes of NH4 during March for both 1999 and 2004 are due to the
distributed flow boundary condition, where data show high NH4 concentration during March
47
-------
Model Configuration and Results
1999. Due to lack of data for 2004, boundary conditions during spring and winter of the 2004
simulation were set equal to those in 1999, causing the NH4 spikes to occur in both years.
There is only one data point available at ARDMUS for comparison to model results, making it
difficult to draw any definitive conclusions on the accuracy of the model simulations. In general,
it is observed that the 1999 model performs better than the 2004 in predicting the low summer
DO. The reason for the poor performance of the 2004 model is most likely the uncertainty in
flow and concentration boundary conditions. Since 2004 was considered as a model corroboration
run, no additional effort was made to adjust loading and parameters to improve the performance.
3.3.6 Lost River at East West Road (LREW)
The model results at LREW are compared with the observed data in Figures D_1999-5 and
D_2004-6. For DO, the model simulated relatively high concentrations during winter, spring, and
fall, as well as relatively low values during summer. The diel fluctuation of DO and NH4 is very
prominent, again suggesting very strong biological impact on water quality. The MaxDepth
survey conducted in 2004 showed dense macrophyte vegetation in this area. The model
reproduces the observed DO and NH4 trends reasonably well. The 2004 model, however, was not
able to predict the extremely high DO (16.0 mg/L) during June 2004. It was hypothesized that
the system (i.e., the boundary conditions set in the model, which were based on limited 2004
monitoring data for inflows to the river) lacked sufficient nutrient loading to sustain extensive
macrophyte growth. An alternative boundary condition loading scenario was run to test this
theory, and it is described in the Diel Dissolved Oxygen Analysis section of the report. The
alternative scenario demonstrated that the model responds to the increased nutrient load by
predicting a wider range in DO fluctuation, and thus represents the dynamics of the system.
Since 2004 was considered a model corroboration run, and since the boundaries were initially set
based on observation data, no attempt was made to replace all boundary condition inputs for the
system to improve the performance. It should also be noted that in the 1999 results, when the
nutrient loading reached high levels during fall, the model was able to predict extremely high DO
of over 20.0 mg/L. No vertical stratification is apparent from either the model results or
monitoring data.
3.3.7 Lost River at Tule Lake (TLTO)
The model results for Tule Lake are compared with the observed data at the Tule Lake outlet
(TLTO) in Figure D_l999-6. For DO, the model simulated the observed trends reasonably well.
The simulated NH4 and NO3/NO2 concentrations generally follow the observed data. However,
simulated NH4 is significantly lower than the observed concentrations, likely due to the
representation of Tule Lake as a single segment. Using this representation, the entire load
coming into the lake is instantaneously mixed throughout the entire lake, while in reality
significant spatial gradients may exist. A direct consequence is that the response of nutrient
concentrations to biological activity is significantly faster than response to external loading.
Therefore, the NH4 is quickly depleted by the algal growth but is more slowly replenished from
external loading. In return, the depleted NH4 limits biological activity in the lake. A better
representation might be achieved using a higher resolution model. However, no attempt was
made to further refine this model given time and data limitations.
48
-------
Model Configuration and Results
3.3.8 P-Canal (PC)
The model results for P-Canal were compared to the observed data for 2004 (Figure D_2004-7).
Since no distributed flow boundary condition was configured for P-Canal due to its relatively
small drainage area, water quality in the canal is mainly controlled by the Tule Lake outflow
conditions. For example, the lower predicted DO and NH4 is inherited from the model
uncertainty in Tule Lake. For better predictions, the processes between Tule Lake and P-Canal as
well as the representation of Tule Lake itself would need to be improved.
3.3.9 Klamath Strait Drains at the State Line (KSDSR)
The model results for the Klamath Straits Drain at the state line were compared to the observed
data at the same location (Figures D_1999-7 and D_2004-8). For DO, the model simulated
relatively high values during the winter, spring, and fall, and represented the relatively low values
during the summer. The model results show insignificant diel fluctuation of DO and NH4. This
is due to the very low chlorophyll a and macrophyte biomass resulting from the extremely
unfavorable light conditions in the drain. The simulated NH4 and NO3/NO2 results generally
follow the observed data. However, significant disparities exist. Possible reasons for these
disparities include:
The observed data were collected in the Klamath Straits Drain, which is immediately
downstream of Lower Klamath Lake. Water quality at this location is significantly
impacted by the lake, which is represented very generally (as one segment) in the model.
The DST boundary condition was set based on the DST of Tule Lake as well as the data
at pumps E and F, and these show different trends from that of KSDSR.
Uncertainty in flow balance.
Possible improvement may be achieved using spatially variable DST boundary conditions for
different sections of this waterbody. No attempt was made to further refine these results given
data limitations. The 2004 results indicate that the model captures trends reasonably well, except
for PO4, which is most likely caused by uncertainty in the boundary conditions.
3.3.10 Klamath Straits Drain at Pump E (KSDTR)
The model results for the Klamath Straits Drain at Pump E are compared with the observed data
in Figure D_1999-8. The model reproduces the observed trends for DO, nutrient concentrations,
and pH well. No vertical stratification is apparent from either the model results or monitoring
data. The model results show insignificant diel fluctuation of DO and NH4. This is due to the
very low chlorophyll a and macrophyte biomass resulting from the extremely unfavorable light
conditions in the drain. It should be noted that the model's background light extinction
coefficient was set to a high value (3.5/m) through the calibration process to account for the
observed "dark water" (low light penetration) conditions.
3.3.11 Klamath Straits Drain at Pump F (KSDPSF)
The model results for the Klamath Straits Drain at Pump F are compared with the observed data
in Figure D_1999-9. The model reproduces the observed trends for DO, nutrient concentrations,
and pH well. No vertical stratification is apparent from either the model results or monitoring
49
-------
Model Configuration and Results
data. The model result shows insignificant diel fluctuation of DO and NH4. This is due to the
very low chlorophyll a and macrophyte biomass resulting from the extremely unfavorable light
conditions in the KSD.
3.3.12 Klamath Straits Drain at Highway 97 (KSD97)
The model results for the Klamath Straits Drain at Highway 97 are compared with the observed
data in Figure D_1999-10. The model once again reproduces the observed trends for DO,
nutrient concentrations, and pH well. No vertical stratification is apparent from either the model
results or monitoring data. The model results show insignificant diel fluctuation of DO and NH4.
Similar chlorophyll a and macrophyte biomass conditions exist as for previous sections of the
drain.
3.3.13 Klamath Straits Drain at Railroad (KSDM)
The 2004 model results for the Klamath Straits Drain at railroad are compared to the observed
data in Figure D_2004-9. The results indicate that the model generally predicts the observed
condition well. Conditions are similar to previous sections of the Klamath Straits Drain.
3.3.14 Diel Dissolved Oxygen Analysis
Diel DO was measured at LROG, LRDR, LREW, and KSD97 from June 14 to June 17, 2004.
The data reflect the most delicate dynamics in the waterbody including temperature, biological
activities, and nutrient interactions, as well as benthic flux at specific times and locations. It is
clear from monitoring data for the Lost River that the diel variation of in-stream water quality is
significantly impacted by short-term patterns in local loadings (watershed return flows). Flow
and loading to the river can be flashy and highly variable over short time periods (day-to-day or
even hour-to-hour) rather than constant. A highly accurate reproduction of the observed diel DO
data (as well as temperature and conductivity) requires accurate specification of all the major
boundary conditions at sufficient spatial and temporal resolution. Unfortunately, no data
representing such high resolution are currently available for any portion of the Lost River, or
more importantly, to characterize the entire system.
As discussed previously, the model was configured for 2004 using data collected during two
sampling events (one in June and one in July), each lasting for two consecutive days. Model
predictions were compared to the raw DO data, as well as the temperature, pH, and conductivity
data. These results are presented in Figures E_2004-l through E_2004-4. The model simulates
the general trends, however the magnitude of the DO swings is not closely matched at each
location, particularly at LREW. This is a direct result of limitations with regard to setting the
boundary conditions, as discussed in Section 3.3.6. It was hypothesized that the system lacked
sufficient nutrient loading to sustain extensive macrophyte growth. Therefore, an alternative
boundary condition loading scenario was run to test this theory for Waterbody #6 (which contains
LREW).
In the original model run for 2004, the nutrient concentrations for the DST to Waterbody #6 were
based on measured data at LREW. With the values being relatively low (between 0.05 and 0.3
mg/L for NH4, 0.02 and 1.7 mg/L for NO2/NO3, and 0.1 and 0.4 mg/L for PO4), macrophyte
growth was limited (at least for the summer) and diel DO fluctuation was underpredicted. For the
alternative loading scenario, NH4, NO2/NO3, and PO4 concentrations in the DST file were
increased. No other changes were made. Concentrations were designated within a range of
50
-------
Model Configuration and Results
values identified in a recent monitoring study of conditions in the canals and drains surrounding
Tule Lake (Danosky and Kaffka 2002). Although the monitoring locations are not in the same
locations as the return flows contributing to Waterbody #6, it was assumed that conditions would
be relatively comparable. The Danosky and Kaffka report summarized a data collection effort
during 1999 for 18 surface water locations and 10 tile drain locations. Samples were collected at
various locations every 10 days from April through October and one to two times a month for the
rest of the year. Table 3-7 summarizes data from the report (which were used as bounds for the
alternative loading scenario). Concentrations in Table 3-7 reflect the minimum, maximum,
average, and standard deviation of samples across all locations for any given date.
Table 3-7. Alternative Loading Scenario Nutrient Concentrations (Danosky and Kaffka 2002)
^^| "mq/U3 (mq/U (mq/L) ^H
Minimum
Maximum
Average
Standard Deviation
0.01
14.37
3.79
4.45
0.00
16.23
2.67
3.95
0.15
1.94
0.80
0.54
Results of the alternative loading scenario are presented in Figure E_2004-5. The figure includes
comparisons of diel DO and pH data with modeling results. It is apparent from these plots that
the increased nutrient loading (nitrogen, in particular) provides the macrophytes with the ability
to sustain a larger biomass, thus resulting in a wider swing in DO levels. The scenario suggests
that the model represents the physical, chemical, and biological attributes of the system
reasonably well, and that it is important to provide accurate boundary loading conditions to match
the observed diel DO data. A more detailed discussion is provided herein regarding the model's
limitations with respect to diel simulation, as well as its capability of catching general trends:
Mathematical models are constrained by the availability, quality, and resolution of input
data. Built on very limited data, the Lost River model is not expected to be able to mimic
the exact timing and location of all water quality conditions. The model is expected,
however, to be able to represent general water quality trends in response to external
loadings and internal dynamics of the system.
While the Lost River watershed covers a large area and exhibits spatially-variable
weather conditions, this model was built using only the data at the Klamath Falls weather
station (the most complete dataset available). This station can represent the overall
conditions in the watershed, however, it does not reflect site-specific conditions for each
modeled Waterbody. This has large implications on the temperature and water quality
simulation.
The in-stream water quality at most locations in the Lost River system is primarily
determined by characteristics of the local inflow. Therefore, the diel variation of in-
stream water quality can be significantly impacted by time-variable patterns associated
with local loadings. Flow and loading to the Lost River can vary significantly over a
short time period. For example, the conductivity at LRARB has been observed to change
by more than 300 uS/cm from one day to the next. This indicates that to accurately
simulate the timing and magnitude of the diel fluctuation, or any short term variability of
water quality in the river, accurate specification of boundary conditions at a similar
resolution is necessary (i.e., spatially-variable loading at each segment on a daily or sub-
daily timestep).
Another factor impacting the model's capability of reproducing diel fluctuation in water
quality is bathymetry information. Bathymetry is particularly important for the
reservoirs, because an accurate representation of bathymetry determines the model's
51
-------
Model Configuration and Results
capability of representing not only volume and flow but also the distribution of
macrophytes. For example, if the bathymetry has a shallower bed area than in reality, the
distribution of macrophytes and their impact on DO may be inaccurately simulated, and
vice versa. Currently, there is no detailed bathymetry data for the reservoirs.
The W2 model represents the river/reservoir system as a longitudinal-vertical 2-D
system, where the water quality in each computational cell represents the average
conditions in that cell. The observed data, however, represents highly localized
conditions in an area immediately surrounding the monitoring device. While the model
can predict the overall mass transport and balance on a large scale, it cannot necessarily
mimic highly localized features.
3.3.15 Macrophyte Analysis
As discussed earlier in the report, an aquatic vegetation survey was conducted during July 2004.
The lumped macrophyte biomass, in terms of dry weight per square meter, were reported at 10
locations along the Lost River system. To confirm the model's capability of matching the general
spatial pattern of macrophyte distribution, the minimum and maximum values of the model-
simulated average macrophyte biomass per square meter for July 2004 were plotted against the
observed data in Figure E_2004-6. Note that the observed data are represented using the
measured dry weight per square meter multiplied by the average trans-section coverage. The
model performs reasonably well in reproducing the observed pattern for most locations. The
model performs worst at the Lost River at Gift Road location, where the survey showed an
extremely high density of macrophyte growth. One reason for this may be that there is a
localized high nutrient loading which is not reflected in any of the monitoring data. Another
reason is that the macrophytes may extract significant amounts of nutrients from the benthic mud
to sustain such a high density of growth. Similarly, the model significantly underpredicts the
macrophyte growth at LREW and P-Canal. This is most likely due to lack of sufficient nutrient
boundary conditions. It is expected that the model performance for macrophyte simulation can be
improved by: (1) configuring more macrophyte species to represent more detailed dynamics in
terms of light competition, nutrient utilization, and response to flow conditions; (2) refining
boundary conditions after further data collection; or (3) developing a comprehensive ecological
model.
3.4 Model Sensitivity Analyses
Since a mathematical model is a simplified representation of the real world, its prediction is often
subject to considerable uncertainty from a variety of sources. These sources include over-
simplification of modeling assumptions and formulations, noise-distorted data, and model
parameter values. It is important to gain a better understanding of a model's reliability by
analyzing the uncertainty associated with a model. Sensitivity analysis is a prime method of
measuring a model's uncertainty and reliability.
In this study, the sensitivity of the DO concentration was evaluated through a number of
mechanisms. The first mechanism evaluated the impact of key parameters on DO at two
locations (LRKB and LRHD). The analysis was performed by adjusting a single parameter at a
time by 20% (higher and lower) and evaluating the DO response of the waterbody (in terms of
DO change). The analysis was performed for the following parameters:
52
-------
Model Configuration and Results
Sediment oxygen demand (SOD)
Phytoplankton growth rate
CBOD decay rate
Macrophyte growth rate
SOD is the total oxygen consumption incurred by the biochemical processes in the sediment
layer. Generally, SOD is positively correlated to nutrient and organic loadings to a specific water
body (Chapra, 1997). Figures F_1999-l and F_1999-5 in Appendix F_1999 show that the DO
concentration is mildly sensitive to a change in SOD. LRHD shows a greater response to SOD
adjustment than LRKB.
The sensitivity of DO to phytoplankton growth rate is shown in Figures F_1999-2 and F_1999-6.
DO concentrations are insensitive to the change in phytoplankton growth rate, meaning that
phytoplankton plays a minor role in DO dynamics (for these locations).
The sensitivity of predicted DO concentration to CBOD decay rate is presented in Figures
F_1999-3 and F_1999-7. As shown, the DO results are relatively insensitive to the change in
decay rate, with a range of DO response within 0.2 mg/L. One reason for this insensitivity is due
to the relatively low value of CBOD concentration in the river.
The DO is most sensitive to the macrophyte growth rate as shown in Figures F_ 1999-4 and
F_1999-8. Results show that the DO concentrations react nonlinearly to macrophyte growth rate
changes and both increase and decrease due to either an increase or decrease in rate. This
exemplifies the complexity of nutrient and macrophyte interactions and its importance in DO
dynamics.
The second sensitivity evaluation addressed the impact of phosphorus reductions versus nitrogen
reductions on DO levels. This evaluation was performed by making two nutrient loading
reduction simulations and comparing them to existing conditions. The only difference between
the simulations was that the first simulation involved a reduction of phosphorus and nitrogen
(approximately 90%) while the second involved a reduction of only nitrogen (also approximately
90%). The results of these simulations are presented for two locations (LRHD and LRWRC) for
1999 in Figures F_1999-9 and F_1999-10 and are labeled "Nitrogen and Phosphorus Reduction"
and "Nitrogen Reduction Only," respectively.
It's apparent from the plots that both nutrient reduction simulations result in the same prediction.
Thus, DO for the Lost River is more sensitive to nitrogen reduction than to phosphorus reduction.
This observation is corroborated by the monitoring data, which show similar magnitudes of PO4
and NH4 concentrations in the water column. Macrophytes need about six times more nitrogen
than phosphorus for growth, therefore, nitrogen tends to be a limiting factor once macrophytes
grow to a certain level.
A third sensitivity evaluation addressed the potential impact of riparian shading on DO levels.
This evaluation involved a comparison of DO levels under existing conditions to levels under an
increased shade simulation. A 30% reduction of the solar radiation value was implemented for all
Waterbodies except in Wilson, Tule, and Lower Klamath Lakes to grossly represent increased
riparian shading. This 30% increase was assumed to represent the maximum possible shading for
the system and is most applicable to relatively narrow, riverine portions and narrow
impoundments (as opposed to wide lakes such as Wilson, Tule, and Lower Klamath). It should
be noted that the increased shade simulation does not explicitly consider vegetation
53
-------
Model Configuration and Results
height/density, the path of the sun or impact of variable shading over the course of the day,
orientation or geometry of the Waterbody (i.e., width of the river/impoundment), or topographic
shading impacts. Results of the simulations are presented in Figures F_1999-ll through F_1999-
19.
Increased riparian shading for the narrow Waterbodies resulted in noticeable changes to only a
few areas. Specifically, minimum DO levels at Harpold Dam increased by as much as 1.0, while
maximum DO levels decreased by as much as 2.0 mg/L. Minor increases in minimum values and
decreases in maximum values were apparent at Anderson-Rose Dam and East-West Road. Most
other locations showed little or no impact from increased shading.
54
-------
Model Configuration and Results
4.0 TMDL SCENARIO
After calibrating and validating the model and conducting model sensitivity simulations, the
model was applied to TMDL development using 1999 as the basis. The TMDL scenario involved
making load reductions to external boundaries (including the upstream boundary at Malone Dam
and all tributary and distributed flow inputs) in an upstream to downstream manner while
ensuring that water quality criteria were met in all riverine sections. Load reductions (and
incoming concentrations) were made for nitrogen and BOD, and these reductions varied
throughout the system. Reductions were not made to phosphorus since sensitivity analyses and
monitoring data indicated the DO levels were relatively insensitive to phosphorus reductions. DO
criteria were not achieved in all the Oregon impoundments for this TMDL scenario based solely
on nitrogen and BOD reductions. As such, an additional DO loading required to achieve DO
criteria for Wilson and Anderson-Rose Reservoirs and the Klamath Straits Drain was identified
subsequent to the nutrient and BOD loading reduction simulation.
The water quality criteria evaluated for compliance during the TMDL scenario are presented in
Table 4-1. No margin of safety (MOS) was explicitly considered in the modeling. Multiple
compliance points were evaluated throughout the system to ensure that water quality criteria were
being met in critical locations. These stations, shown graphically in Figure 2-1, are as follows:
Lost River at Gift Road (LRGR)
Lost River at Keller Bridge (LRKB)
Harpold Dam (LRHD)
Lost River at Stevenson Park (LRSP)
Lost River at Olene Gap (LROG)
Wilson Dam (LRWRC)
Lost River at Dehlinger Road (LRDR)
Lost River at Hwy 39 n/w of Merrill (LR39)
Anderson-Rose Dam (LRAR)
Lost River at Stateline Road - OR/CA border (LRSR)
Lost River at East-West Road (LREW)
Tule Lake (TLTO)
Lower Klamath Refuge/Lake (LKL)
Klamath Straits Drain at Stateline Road - OR/CA border (KSDSR)
Klamath Straits Drain at Township Road (KSDTR)
Klamath Straits Drain at Hwy 97 (KSD97)
Klamath Straits Drain at (KSDM)
55
-------
Model Configuration and Results
Table 4-1. Water Quality Criteria Evaluated for Compliance During TMDL Scenario
Parameter
Dissolved Oxygen
California
5.0 mg/L as an absolute
minimum
Oregon
4.0 mg/l as an absolute
minimum
6.5 mg/l as a 30-day mean
minimum
5.0 mg/l as a 7-day minimum
mean
Ammonia
No objective
See Table 20, OAR 340-41-
0965 (2)(p)(B)
chlorophyll-a
No objective
<0.015 mg/L
PH
7.0-9.0
6.5-9.0
The piece-wise simulation technique developed for this model enabled an efficient TMDL
scenario analysis from upstream to downstream. Waterbody #1 was analyzed first, with
reductions made to the boundary condition concentrations for BOD and nitrogen, as well as SOD.
After several iterations and determining the external loading reduction required to achieve the
water quality targets at Gift Road, Keller Bridge, and Harpold Dam (critical locations within
Waterbody #1), the discharge at Harpold Dam was configured as the upstream boundary
condition to Waterbody #2 and #3. A similar procedure was implemented to meet water quality
criteria in this segment (at Stevenson Park and Olene Gap). This procedure was followed in an
upstream to downstream manner until water quality criteria were achieved in all waterbodies (at
all critical locations), with the exception of Wilson and Anderson-Rose Reservoirs and the
Klamath Straits Drain. In general, the DO criteria were the most stringent criteria. Therefore,
chlorophyll-a and ammonia criteria were achieved once DO criteria were met. Model simulation
results at all compliance points are presented along with water quality criteria in Appendix G.
The nitrogen and BOD load reductions required, by Waterbody, to achieve water quality criteria
are presented in Table 4-2. It should be noted that pH and ammonia toxicity targets were slightly
exceeded for the TMDL scenario in a number of locations due to model and boundary condition
uncertainty. These exceptions to water quality criteria achievement are further described below.
Table 4-2. Nitrogen and BOD Load Reductions Required to Achieve Water Quality Criteria*
Waterbody # Reduction %
1
2
3
4
5
6
7
8
9
10
11
12
50
50
50
50
50
50
49
49
49
49
49
49
With the exception of Wilson and Anderson-Rose Reservoirs and the Klamath Straits Drain
56
-------
Model Configuration and Results
DO criteria were nearly impossible to achieve in the hypolimnion of Wilson and Anderson-Rose
Reservoirs or within the Klamath Straits Drain based solely on BOD, nitrogen, and SOD
reductions. Therefore once the model had been run to achieve water quality criteria in all other
waterbodies, the required DO loading to meet the DO criteria was calculated for Wilson and
Anderson-Rose Reservoirs and the Klamath Straits Drain. This calculation was based on the
most stringent DO criteria and average impoundment volumes.
Minimum modeled DO concentrations under the TMDL scenario for the minimum, 30-day, and
7-day Oregon DO criteria are presented in Table 4-3 at compliance points within Wilson and
Anderson-Rose Reservoirs and the Klamath Straits Drain. The necessary DO concentration
increase at each compliance point for each DO criteria is also presented, along with the DO
criteria. The necessary DO increase is the difference between the DO criteria and the minimum
modeled DO concentration, and it represents the greatest divergence from the DO criteria at any
given time throughout the year.
Table 4-3. DO Concentrations in Impoundments Not Meeting All DO Criteria Based Solely on
Load Reductions
Impoundment
Minimum Modeled DO Criteria (mg/L)
DO (mg/L)
Necessary DO
Increase (mg/L)
Min 30- 7-
day day
30- 7-
day day
LRWRC
LRAR
KSDSR
KSDTR
KSD97
KSDM
0.87
2.15
5.2
4.56
5.59
5.56
2.62
6.02
5.75
5.73
6.20
6.17
1.12
3.42
5.36
4.93
5.81
5.79
4.00
4.00
4.00
4.00
4.00
4.00
6.50
6.50
6.50
6.50
6.50
6.50
5.00
5.00
5.00
5.00
5.00
5.00
3.13
1.85
N/A
N/A
N/A
N/A
3.88
0.48
0.75
0.77
0.30
0.33
3.88
1.58
N/A
0.07
N/A
N/A
The necessary DO increase was then converted to mass units to comply with ODEQ TMDL
development requirements. This calculation was made for Wilson Reservoir, Anderson-Rose
Reservoir, and three impounded regions of Klamath Straits Drain (Waterbody #10 - from
KSDSR to KSDTR; Waterbody #11 - from KSDTR to KSD97; and Waterbody #12 - from
KSD97 to KSDM). The necessary DO increase was multiplied by the average volume of the
associated impoundment for the critical period (in terms of DO depression) from May through
October to obtain a mass. Table 4-4 presents the required DO mass (in tons and kg) for each
impoundment. It is recognized that DO changes over time and that the "static" or "instantaneous"
mass presented reflects a worst case DO condition and an average volume during the critical
season. It is assumed that the engineering solutions to improve the DO concentrations will
further explore and evaluate the temporal variability of DO in identifying necessary DO inputs.
57
-------
Model Configuration and Results
Table 4-4. DO Loading Requirements for Impoundments Not Meeting All DO Criteria Based
Solely on Load Reductions
Location Required Required DO
DO Mass Mass (tons)
(kg)
Wilson Reservoir
Anderson- Rose Reservoir
Klamath Straits Drain - Waterbody #10
Klamath Straits Drain - Waterbody #1 1
Klamath Straits Drain - Waterbody #1 2
16,929.71
496.96
152.64
70.89
32.56
16.93
0.50
0.15
0.07
0.03
As noted above, a number of the water quality criteria were exceeded under the TMDL scenario
due to model and boundary condition uncertainty. Ammonia toxicity model predictions were
found to exceed limits in the spring at LRDR, LR39, LRAR, and LRSR. These high values can
be attributed to the distributed boundary condition for NH4 between Wilson Dam and Anderson
Rose Dam. Unfortunately, there was only one data point used to derive this boundary condition
(and it was a high value-almost 1.6 mg/L, at Anderson Rose Dam). As such, the spring ammonia
toxicity predictions for this stretch of river are not tremendously reliable. It should also be noted
that based on a review of the monitoring data for this period, there were no apparent ammonia
toxicity issues.
As with ammonia toxicity, pH exceeded criteria at a number of locations. The pH exceedances
are largely attributed to background conditions (and in many cases are high throughout the year).
pH was not found to be too sensitive to adjustments in nutrients, and thus was assumed not to be
driven predominantly by biological processes. A few potential exceptions to this statement are at
LR39, LRSR, and LREW during October. In these cases, the pH appears to be driven, at least
partially, by biological activity. The predicted pH values during this period, however, are not
entirely reliable due to modeling uncertainty. Specifically, insufficient alkalinity data and no TIC
data were available to support boundary condition settings. It should also be noted that
monitoring data for 1999 at these stations do not show pH exceedances.
Chlorophyll-a concentrations in Appendix G appear to exceed the criteria in a number of
locations, however, these plots represent instantaneous levels. Oregon's chlorophyll-a criterion
targets a 3-month average chlorophyll-a concentration, along with depth averaging in the photic
zone. As such, the apparent exceedances at LRWRC and LRAR are actually below the 15 ug/L
criterion when averaged over 90 days.
A number of important assumptions were made in the process of running the TMDL scenario:
Nitrogen and BOD boundary conditions were reduced equally (within each waterbody).
SOD was reduced by the same percentage as boundary condition reductions (e.g., a 20%
boundary condition reduction would result in a 20% SOD reduction). This is based on
the linear assumption, as described in Chapra 1997, and has been widely used in TMDL
development when sediment diagenesis is not explicitly simulated. The SOD reduction
ratio for all the downstream waterbodies was calculated based on the lumped loading
from all tributaries, distributed loadings, as well as the contribution from upstream
waterbodies. For example, the reduction of loading to Waterbody #1 also influences
SOD reduction in Waterbody #2. The loading used to calculate the SOD reduction is on
an annual basis, and ammonia, nitrite/nitrate, and CBOD were used as the corresponding
58
-------
Model Configuration and Results
constituents. Since the loading reduction ratio of each of the constituents can be
different, the average of the calculated reduction ratio for the constituents was used as the
SOD reduction ratio.
Boundary condition DO was kept at incoming levels (i.e., calibration conditions) when
above water quality criteria; otherwise, it was set to Oregon's most stringent water
quality criteria (6.5 mg/L - which is based on the Oregon 30-day average criteria for
DO). This is based on the assumption that implementation in the watershed will enable
DO levels for incoming water to achieve the water quality criteria. If the incoming water
does not achieve the criteria, in-stream DO levels will violate the criteria at some
locations (primarily where watershed return flow dominates).
No change in temperature was made for boundary conditions from the calibrated model.
The maximum algae concentration was forced to the 15 ug/L standard or lower (only
when it was higher than the standard). This is based on the assumption that
implementation in the watershed will reduce algae concentrations for the incoming water
(based on corresponding nutrient reductions).
59
-------
Model Configuration and Results
5.0 DISCUSSION
The Lost River is an extremely complex system consisting of different waterbodies with distinct
physical, chemical, and biological features. In general, the riverine sections represent a system
that moves relatively quickly, and they demonstrate a rapid response to external loading. The
impounded sections represent a drastically different type of system - one that is relatively
stagnant and shows a much slower response to external loading. A modified W2 modeling
framework, specific to the Lost River, was developed to evaluate the temporal and spatial
variability of loading and in-stream hydrodynamic and water quality conditions.
Although significant data limitations posed a challenge to model development and application,
the approach implemented proved successful in deriving dominant boundary conditions and
making relatively accurate in-stream hydrodynamic and water quality predictions. The modeling
study suggests that macrophytes are the dominant factor controlling the diel DO and nutrient
fluctuation, and specifically minimum DO levels. This is corroborated by the intensive aquatic
vegetation survey conducted during the summer of 2004. Even though phytoplankton levels can
be high in some riverine locations, it is primarily due to localized external contributions and has a
less significant impact on internal dynamics and DO levels. In the impoundments, internal
growth of phytoplankton generally plays a more significant role due to the stagnant environment,
which is more favorable to algae growth.
The modeling also suggests many segments of the Lost River are limited more by nitrogen than
by phosphorus, with respect to macrophyte development. This model-based observation is
corroborated by the monitoring data, which show similar magnitudes of PO4 and NH4
concentrations in the water column. Macrophytes need about six times more nitrogen than
phosphorus for growth, therefore, nitrogen tends to be a limiting factor once macrophytes grow to
a certain level. This observation is also consistent with the Lost River aquatic vegetation survey
conducted in 2004 by MaxDepth.
The current modeling framework utilizes the best available data for the Lost River and provides a
sound technical basis for TMDL analysis and evaluation of water quality standards.
Improvements to the framework can be grouped into two major categories: further data
collection and model improvements. The most apparent data gaps are associated with defining
the temporal and spatial variability of all inputs to and withdrawals from the Lost River and its
impoundments. Characteristics data for Wilson Reservoir, Tule Lake, Lower Klamath Lake are
also needed to improve the model.
Representing competition for nutrients and light between phytoplankton and macrophytes, as well
as between different phytoplankton and macrophyte species, is seen as the most important model
improvement to accurately predict primary productivity in the Lost River. In the current model, a
single species of phytoplankton and macrophytes was used to represent the broader range of
vegetation present. In reality, different species within each group may show significantly
different biological behavior. For example, floating macroalgae, such as duckweed, could cover
the water surface and shade the entire water column below. Even the modified version of W2
used for this application is not able to represent the impact of duckweed on phytoplankton and
other macrophytes. Another issue not explicitly represented is the life cycle of macrophytes.
Some macrophytes, such as duckweed, do not die during the winter. Instead, they transform into
a dormant form and sink to the bottom of the water column until the next year. This is another
feature that most existing water quality models, including W2, cannot represent. Accurately
60
-------
Model Configuration and Results
simulating the multi-year life cycles of macrophytes is important for enhancing the predictability
of the model, especially when the long term effect of nutrient loading reductions are concerned.
61
-------
Model Configuration and Results
6.0 REFERENCES
Burt, C., and B. Freeman (2003). Hydrologic Assessment of the Upper Klamath Basin: Issues
and Opportunities, Draft Report, Prepared for BOR.
Cole, T.M., and S.A. Wells (2003). CE-QUAL-W2: A two-dimensional, laterally averaged,
Hydrodynamic and Water Quality Model, Version 3.1, Instruction Report EL-03-1, US Army
Engineering and Research Development Center, Vicksburg, MS.
Chapra, S.C. (1997). Surface Water-Quality Modeling. McGraw-Hill, New York.
Danosky, E. and S. Kaffka (2002). Fanning Practices and Water Quality in the Upper Klamath
Basin, Final Report to the California State Water Resources Control Board, 205j Program.
Mayer, T (2004). Irrigation Water Sources for TID. USFWS. Technical Memorandum.
Park, K., A. Y. Kuo, J. Shen and J.M. Hamrick (1995). A three-dimensional hydrodynamic-
eutrophication model (HEM-3D): description of water quality and sediment process submodels.
Special Report in Applied Marine Science and Ocean Engineering No. 327. School of Marine
Science Virginia Institute of Marine Science, College of William and Mary, January 1995.
Shanahan, P. and M. M. Alam (2001). The Water Quality Analysis Simulation Program, WASPS
Part A: Model Documentation (Updated) Version 5.2-MDEP, Hydraulic & Water Resources
Engineers, Inc., November 16, 2001.
Water Resources Engineers (WRE) (1965). A Proposed Hydrologic-Water Quality Model of the
Lost River System. Report of Preliminary Investigation Conducted for the Klamath Basin Study,
Division of Water Supply and Pollution Control, Region Nine, United States Public Health
Service.
Woods, P.C, and G.T. Orlob (1963). The Lost River System - A Water Quality Management
Investigation. Water Resources Center Contribution Number 68, University of California,
Berkeley, February, 1963.
USGS, (1999). Water Quality, Benthic Macroinvertebrate, and Fish Community Monitoring in
the Lost River Sub-basin, Oregon and California. USGS, Johnson Controls World Services Inc.,
U.S. Bureau of Reclamation.
62
-------
Model Configuration and Results
Appendix A_1999
Water Balance and Water Surface Elevation Calibration
A 1999-1
-------
Model Configuration and Results
1253
1252 -
1248
M
M
0
M Harpold Dam Elevation (1999)
F M A M
A S 0 N D
A_l 999-2 Harpold Dam Outflow (1999)
A 1999-2
-------
Model Configuration and Results
1248
1245
M
M
0
N D
A_1999-3 Wilson Reservoir Dam Elevation (1999)
M
M
0
D
A_l 999-4 Anderson Rose Spill (1999)
A 1999-3
-------
Model Configuration and Results
1231
1231 -
ฃ 1230 -
CD
1230 -
1229
M
M
0 N D
A_l 999-5 Tule Lake Elevation (1999)
M
A
M
A
A_l 999-6 Lower Klamath Lake Volume (1999)
A 1999-4
-------
Model Configuration and Results
Appendix A_2004
Water Balance and Water Surface Elevation Calibration
A 2004-1
-------
Model Configuration and Results
M
M
0 N
D
A_2004-l Harpold Dam Outflow (2004)
1250
1249 -
1245
M A
M
A
A_2004-2 Wilson Reservoir Dam Elevation (2004)
A 2004-2
-------
Model Configuration and Results
M
M
0 N
A_2004-3 Anderson Rose Spills (2004)
A 2004-3
-------
Model Configuration and Results
Appendix B_1999
Hydrodynamic Model Evaluation with Temperature Data
B 1999-1
-------
Model Configuration and Results
0
JFMAMJJA
B_l 999-1 Lost River at Keller Bridge (LRKB) -1999
Max Min Mean o Obs
JFMAMJJA
B_l 999-2 Lost River at Harpold Dam (LRHDB) -1999
N
D
B 1999-2
-------
Model Configuration and Results
Max- Mm Mean O Obs
MAM
A
0 N D
B_l 999-3 Lost River at Crystal Springs (LRWRC) -1999
Max Min Mean O Obs
JFMAMJJASOND
B_l 999-4 Lost River at Anderson Rose Dam (LRAR) - 1999
B 1999-3
-------
Model Configuration and Results
35 7
Max Mm Mean c Obs
M
B_l 999-5 Lost River at East West Road (LREW) - 1999
N
M A M J J A S
B_l 999-6 Lost River at Tule Lake Tunnel Outlet (TLTO) - 1999
N
B 1999-4
-------
Model Configuration and Results
Max- Mm Mean O Obs
A
M
0
D
B_l999-7 Klamath Straits Drain at Stateline Road (KSDSR) - 1999
Max- Mm Mean O Obs
B_l999-8 Klamath Straits Drain at Township Road (pump E) (KSDTR) - 1999
B 1999-5
-------
Model Configuration and Results
25
20
-------
Model Configuration and Results
Appendix B_2004
Hydrodynamic Model Evaluation with Temperature Data
B 2004-1
-------
Model Configuration and Results
M
B_2004-l Lost River at Keller Bridge (LRKB) -2004
N
Max Mm Mean c Obs
J F M A M J J
B_2004-2 Lost River at Harpold Dam (LRHD) -2004
B 2004-2
-------
Model Configuration and Results
Max- Mm Mean O Obs
JFMAMJJASOND
B_2004-3 Poe Valley Bridge at RM 27 (HPDS2) -2004
JFMAMJJAS
B_2004-4 Lost River at Crystal Springs (WDUS/LRWRC) - 2004
B 2004-3
-------
Model Configuration and Results
25
20
-------
Model Configuration and Results
J F M A
B_2004-7 P-Canal (PC) - 2004
M
A
0
N
D
Max Mm Mean r Obs
JFMAMJJASO
B_2004-8 Klamath Straits Drain at Stateline Road (KSDSR) - 2004
N
D
B 2004-5
-------
Model Configuration and Results
JFMAMJJA
B_2004-9 Klamath Straits Drain at Railroad (KSDM) - 2004
0
N
D
B 2004-6
-------
Model Configuration and Results
Appendix C_1999
Further Hydrodynamic Model Evaluation with Conductivity Data
C 1999-1
-------
Model Configuration and Results
1200
1000 -I
I 800-I
& 600
o
I
o
O
Max Min Mean o Obs
400 -J
200- ^\r*^
O
O
1200
1000 -
800 -
600 -
400
200
Max Min Mean o Obs
M
M
C_l 999-2 Lost River at Harpold Dam (LRHD) -1999
JFMAMJJASOND
C_l 999-1 Lost River at Keller Bridge (LRKB) -1999
OND
C 1999-2
-------
Model Configuration and Results
1200
1000 -
800 -
o
o
400 -
.Max Min Mean o Obs
M A
M
A
C_l 999-3 Lost River at Crystal Springs (LRWRC) -1999
1200
0
J F M A M J J A
C_l 999-4 Lost River at Anderson Rose Dam (LRAR) - 1999
C 1999-3
-------
Model Configuration and Results
A
M
A
C_l 999-5 Lost River at East West Road (LREW) - 1999
1200
1000
| 800
& 600 -
o
I 400 -I
o
O
200 -
Max Min Mean o Obs
oฐ a
o o
o
o
o
00
JFMAMJJAS
C_l 999-6 Lost River at Tule Lake Tunnel Outlet (TLTO) - 1999
0
C 1999-4
-------
Model Configuration and Results
2000
J F M A M J J A S
C_l 999-7 Klamath Straits Drain at Stateline Road (KSDSR) - 1999
2000
1800 -
J F M A M J J A S 0
C_l 999-8 Klamath Straits Drain at Township Road (pump E) (KSDTR) - 1999
C 1999-5
-------
Model Configuration and Results
J F M A M J J A S
C_l999-9 Klamath Straits Drain at Pump Station F (KSDPSF) - 1999
JFMAMJJA
C_1999-10 Klamath Straits Drain at Highway 97 (KSD97) - 1999
0
D
C 1999-6
-------
Model Configuration and Results
Appendix C_2004
Further Hydrodynamic Model Evaluation with Conductivity Data
C 2004-1
-------
Model Configuration and Results
C_2004-l Lost River at Keller Bridge (LRKB) -2004
J F M A M J J
C_2004-2 Lost River at Harpold Dam (LRHD) -2004
0
D
C 2004-2
-------
Model Configuration and Results
E
0
-1 '
^>
-1 '
o
^
T3
C
O
O
1700
1000 -
800 -
600 -
400 -
200 -
n
^ ^^ ^ -2 ^^jlfW
JFMAMJJASOND
C_2004-3 Poe Valley Bridge at RM 27(HPDS2) 2004
-ionn
E
o
"o
o
O
1000 -
800 -
600 -
400 -
200 -
n -
o 0
J F M A M J J A S 0
C_2004-4 Lost River at Crystal Springs (WDUS/LRWRC) -2004
C 2004-3
-------
Model Configuration and Results
1200
1000
I 800
& 600
O
O
400
200
Max
Min
Mean Obs
J F M A M J J A
C_2004-5 Anderson Rose Dam Upstream (ARDMUS) - 2004
F
_0
W
^
>^
ductivil
o
o
I^UU
mnn J
snn -
snn
/inn
onn
n -
^^-JV^VA^^W ^^^^^^^^^^
^^^ S~\ " ^
QJ J ^^^
M
M
0 N D
C_2004-6 Lost River at East West Road (LREW) - 2004
C 2004-4
-------
Model Configuration and Results
F
0
O>
^
>,
^>
o
^
o
O
i?nn
mnn -i
snn
snn
9nn J
n _
May Min Mpan o Oh^
~"\
IT
^n ^ Q
^^^ VX L \z>
^~
MAM
A
C_2004-7 P-Canal (PC) - 2004
1600 -
1/inn
E -lonn
0 izuu -
o>
-E- mnn
>^
_-4^
> ann
o
^
"0 snn
o
o
/inn
onn
n -
^^\^_^
M ay Min M pan o Oh^
, Q
0\
\ ^v
Vifir
>ง5
M
M
0 N
C_2004-8 Klamath Straits Drain at Stateline Road (KSDSR) - 2004
C 2004-5
-------
Model Configuration and Results
Max Mm Mean o Obs
J F M A M J J A
C_2004-9 Klamath Straits Drain at Railroad (KSDM) - 2004
C 2004-6
-------
Model Configuration and Results
Appendix D_1999
Water Quality Calibration Results
D 1999-1
-------
Model Configuration and Results
20
18 -\
2
Max
Obs
Min
DO Criteria
Mean
DOSAT
50
40 -
-
Q.
2 20 -I
_
O
10 -
1
0.9 -
_ 0.8 -
fo.7-1
5 o.e
z
.| 0.5 -
o
E 0.4 -I
E
5 0.3 ^
.2
tฐ 0.2 -
0.1 -
0
M
M
O
Max
Min
Mean O Obs - - Chl-a Criteria
O
M
M
O
Max Min Mean O Obs
JFMAMJJA
D_l 999-1 Lost River at Keller Bridge (LRKB) -1999
o
D 1999-2
-------
Model Configuration and Results
14
12
-Max
Mean
pH Criteria 2
-Min
O Obs
- pH Criteria 1
6 -
4
2
0
0.8
O
0.7 -
0.6 -
0.5 -
0.4 -
0.3 -
0.2 -
0.1 -
0
1.5
1.25 -
1 -
0.75 -
0.5 -
0.25
0
JFMAMJ JASOND
Max Min Mean o Obs
MAM
ASOND
- Max Min Mean o Obs
-j-
ifiiQjjiil
JFMAMJ JASOND
D_l 999-1 Lost River at Keller Bridge (LRKB) -1999 continued
D 1999-3
-------
Model Configuration and Results
0.6
0.5 -
0.4
| 0.3
o
E
< 0.2
0.1 -
0.0
1.4
1.2
1.04
1 0.6 -\
I 0.4 ^
0.2 -
0.0
1.4
_ 1.2 -
a 1.0 -
*
z 0.8 ^
55
0.6 H
-SOdayAvg
CCC if Early Stage Fish Present
M
M
Highest 4-day avg in 30 days
O
2.5xCCCEarly Stage Fish Present
M
-CMCAcute
M
O
Mean
JFMAMJJASO
D_l 999-1 Lost River at Keller Bridge (LRKB) -1999 continued
D 1999-4
-------
Model Configuration and Results
JFMAMJJA
D_l 999-2 Lost River at Harpold Dam (LRHD) -1999
D 1999-5
-------
Model Configuration and Results
Q-
14
12 -
10 -
r
8 -
e"
4
2 -I
0
-Max
Mean
pH Criteria 2
-Min
O Obs
- pH Criteria 1
0.8
1)
O
0.7 -
0.6 -
0.5 -
0.4
0.3 -
0.2 -
0.1 -
0
M
M
Max Min Mean O Obs
M
M
O N D
OND
JFMAMJ JASOND
D_l 999-2 Lost River at Harpold Dam (LRHD) -1999 continued
D 1999-6
-------
Model Configuration and Results
0.6
0.5
O)
0.4
<5 0.3 -
c
o
E
10.2 -j
0.1-1
0.0
1.4
1.2 -
0.8 -
o 0.6 -
E
< 0.4 -
ro
"o
H 0.2
0.0
SOdayAvg
CCC if Early Stage Fish Present
M
M
Highest 4-day avg in 30 days
O N
. 2.5xCCCEarly Stage Fish Present
M
CMCAcute
M
O
D
Mean
1.4
1.2 -
E 1-0-1
0.8 -
0.6 -
E
< 0.4 -
ro
H 0.2 -I
JFMAMJJASON
D_l 999-2 Lost River at Harpold Dam (LRHD) -1999 continued
D
D 1999-7
-------
Model Configuration and Results
20 -,
18 -
16 -
14
ง 12
O)
10 -
6
Max
Min
Mean Obs - - DO Criteria
DOSAT
n
4 - -
2
o
M
M
O
D
Min
Mean o Obs - - Chl-a Criteria
Max Mm Mean o Obs
JFMAMJJASON
D_l 999-3 Lost River at Crystal Springs (LRWRC) -1999
D 1999-8
-------
Model Configuration and Results
14
12
-Max
Mean
pH Criteria 2
Min
O Obs
- pH Criteria 1
M
M
O
D)
E
0.8
0.7 -
0.6 -
0.5 -
0.4
*
O
ฐ- 0.3 H
1.5
O)
E
0.75 -
0.25 -
Max Min Mean r Obs
e>
--GG-GG---GG-GG-- -G--
M
M
O
Max Min Mean o Obs
M
M
O
D_l 999-3 Lost River at Crystal Springs (LRWRC) -1999 continued
D 1999-9
-------
Model Configuration and Results
1.4
I-2 -
1.0 -1
ro
| 0.6 -j
E
^ 0.4 -
0.2 -
0.0
1.4
1.0 ^
0.6 ^
0.2 -
SOdayAvg
CCC if Early Stage Fish Present
M
M
Highest 4-day avg in 30 days
O
2.5xCCCEarly Stage Fish Present
V
/v
M
- -CMCAcute
M
O
Mean
JFMAMJJASON
D_l 999-3 Lost River at Crystal Springs (LRWRC) -1999 continued
D 1999-10
-------
Model Configuration and Results
M
M
O N
150
=5, 100 -
-Max
Min
Mean
O Obs
- - Chl-a Criteria
>s
Q.
O
O
50 -
J FMAMJ JASOND
1.8 -
E 1.4 -j
H
0.8 -
0.4 -
0.2 -
0
Max Min Mean o Obs
O
JFMAMJJASOND
D_l 999-4 Lost River at Anderson Rose Dam (LRAR) - 1999
D 1999-11
-------
Model Configuration and Results
14
-Max
Mean
pH Criteria 2
-Min
O Obs
- pH Criteria 1
1
0.9 -
0.8 -
0.7 -
)
-0.5 -
0.3 -
0.2 -
0.1 -
0
2.25
2
1.75 -
1.5 -
1.25 -
1 -
0.75 -
0.5 -
0.25 -
0
M
M
O N D
-o-
Max Min Mean Obs
M
M
O N D
Max Min Mean o Obs
JFMAMJJASOND
D_l 999-4 Lost River at Anderson Rose Dam (LRAR) - 1999 continued
D 1999-12
-------
Model Configuration and Results
0.0
-SOdayAvg
CCC if Early Stage Fish Present
M
M
Highest 4-day avg in 30 days
O
2.5xCCCEarly Stage Fish Present
M
CMCAcute
M
O
-Mean
JFMAMJJASON
D_l 999-4 Lost River at Anderson Rose Dam (LRAR) - 1999 continued
D 1999-13
-------
Model Configuration and Results
25 -
20 -
T3
CD
I 10
5 -
-Max
Obs
Min
DO Criteria
Mean
M
M
O N
100
90 -
80 -
~ 70 -
i 60 -
ro
>- 50 -
Q.
ง 40
O 30 -
20 -
10 -
Max
Min
Mean
Obs
Chl-a Criteria
O
Q
-_*_-^
0
M
M
O N D
1.2 -
Max - Mm Mean o Obs
0
J FMAMJ JASOND
D_l 999-5 Lost River at East West Road (LREW) - 1999
D 1999-14
-------
Model Configuration and Results
14
12
10
Max
Mean
-pH Criteria 2
Min
O Obs
- pH Criteria 1
O
MAM
S O N D
0.8
O
0.7 -
0.6 -
0.5 -
0.3 -
0.2 -
0.1 -
Max Min Mean r Obs
JFMAMJJASOND
2.52
1.89 -
D)
E
1.26 -
0.63 -
0
Max Min Mean o Obs
v-
O
00
JFMAMJJASON
D_l 999-5 Lost River at East West Road (LREW) - 1999 continued
D 1999-15
-------
Model Configuration and Results
0.0
0.0
-SOdayAvg
CCC if Early Stage Fish Present
M
M
Highest 4-day avg in 30 days
O
2.5xCCCEarly Stage Fish Present
M
CMCAcute
M
O
Mean
J FMAMJ JASO
D_l 999-5 Lost River at East West Road (LREW) - 1999 continued
D 1999-16
-------
Model Configuration and Results
20
O)
>s
6
-a
18 -
16 -
14 -
12 -
10 -
8 -
6 -
4
2
O Obs
DO Criteria
DOSAT
MAM
OND
"5)
100
90
80
70
60 -
>, 50-
Q. 40
.
O
20 -
A
i\.
/ V
y ^-^^^^
M
M
OND
1
0.9 -
_ 0.8 -
lo.7-I
^r
z
.| 0.5 -
o
E 0.4 ]
E
" 0.3 -
ro
0.2-
0.1 -
0
Max Min Mean o Obs
O O u VJ.Q Q
JFMAMJJASO
D_l 999-6 Lost River at Tule Lake Tunnel Outlet (TLTO) -1999
D 1999-17
-------
Model Configuration and Results
1 n
8 -
0. . ,
R
A
O
n
J
n 8
n 7
Oc
Oc
~D)
E n A
O
o_ n ^
0.2 -
n 1
n -
J
1 5
-1
E n 7^
X
0
n ^
n 9^
0
- pH Criteria 2 - pH Criteria 1
^^ ^^nnOOn^^
;k - O-O-(^-(=r)j^.-i(^-O -^ XJ->% zฃ - - UJU-^ - - .
ซ^ ^^x- _^_,^^^ .^...^..i (j
FMAMJ JASOND
/~^^^ O O
O O
O OQ OO O O
FMAMJ JASOND
I^ZI
ooooo^ooooooo^o ^p^^
JFMAMJJASOND
D_l 999-6 Lost River at Tule Lake Tunnel Outlet (TLTO) -1999 continued
D 1999-18
-------
Model Configuration and Results
-SOdayAvg
CCC if Early Stage Fish Present
CCC if Early Stage Fish Absent
0.0
M
M
-Highest 4-day avg in 30 days
2.5xCCCEarly Stage Fish Present
O
2.5xCCCEarly Stage Fish Absent
M
CMCAcute
M
O N D
Mean
0.0
J FMAMJ JASOND
D_l 999-6 Lost River at Tule Lake Tunnel Outlet (TLTO) -1999 continued
D 1999-19
-------
Model Configuration and Results
20 -
18 -
16 -
|> 14 -
c 12 -
0) '^
o)
ฃ10-
T3 Q
~ฐ K
0) 0 -
(0
S 4
2 -
0 -
O Obs - - DO Criteria DO Saturation
./I f( f>,
^t^v; 7^
^^ vw . X
""^\"/Ai" "jX"
V i. ^A/^
^.^^^v A^ IJiiiU ff\
^ V^^x f imti^^
\ 1 V i Jf
\ / V Jilllll
(j^j ^yff^^wj^y ^r (|^
JFMAMJJASOND
c;n
/in
Chlorophyl-a (ug/l)
-> M CO J
3 O O O C
u -i
2
1.8 -
1.6 -
1.4 -
^12-
ฃ 1 -
CO
^ 0.8 -
0.6 -
0.4 -
0.2 -
0 -
D_199S
Q Q O Q Q QQO Q Q O GDQGSOO'OQSOLGO O U O
JFMAMJJASOND
. . ... . . .
O
mi
~$\ 9Ski BJr ~jf
n-a^r0^ i\ AflK A ^1 R'w-*"S'^
__ฎ^ ^IT^fe^ o
JFMAMJJASOND
>-7 Klamath Straits Drain at Stateline Road (KSDSR) -1999
D 1999-20
-------
Model Configuration and Results
Q.
14
12 -
10 -
8 -
e"
4
2 -I
0
O
0.8
0.7 -
0.6 -
0.5 -
0.4 -
0.3 -
0.2 -
0.1 -
0
Max
Mean
-pH Criteria 2
-Min
Obs
- pH Criteria 1
M
M
O
D
Max
Min
Mean
Obs
M
M
O
D
M
M
D
D_l999-7 Klamath Straits Drain at Stateline Road (KSDSR) -1999 continued
D 1999-21
-------
Model Configuration and Results
-SOdayAvg
CCC if Early Stage Fish Present
0.0
M
M
Highest 4-day avg in 30 days
O
2.5xCCCEarly Stage Fish Present
0.0
M
CMCAcute
M
O
Mean
0.0
J FMAMJ JASOND
D_l999-7 Klamath Straits Drain at Stateline Road (KSDSR) -1999 continued
D 1999-22
-------
Model Configuration and Results
20
18
16
14 -
<= 12
CD
D)
x 10
o
"O o
(D O
^
O R
0) O
Max
Obs
Min
DO Criteria
Mean
DOSAT
4 -___ ..-^ ^.
2
50
40
ro
">s
Q.
|2(H
6
10
1
0.9 -
_ 0.8 -
5 0.6 -
.3 0.5 -
o
E 0.4 -
< 0.3 -
(ฐ 0.2 -
0.1 -
0
M
M
O
M
M
O N D
Max Min Mean o Obs
--O-
"O
J FMAMJ JASOND
D_l 999-8 Klamath Straits Drain at Township Road (pump E) (KSDTR) -1999
D 1999-23
-------
Model Configuration and Results
14
12
10
Q.
6 -
4
2 -I
0
Max
Mean
pH Criteria 2
Min
O Obs
. -pH Criteria 1
00
MAM
O N D
O
0.8
0.7 -
0.6 -
0.5 -
0.4 -
0.3 -
0.2 -
0.1 -
0
Max Min Mean o Obs
FMAMJ JASOND
4
3.75 -
3.5 -
3 -
2.7i
2.J
1>2.2J
ฃ 2\
g 1.7!
z i:
o.i
Max Min Mean o Obs
0
JFMAMJ JASOND
D_l 999-8 Klamath Straits Drain at Township Road (pump E) (KSDTR) -1999 continued
D 1999-24
-------
Model Configuration and Results
0.0
1.4
1.2
E
E
<
"ro
5
1.0
0.8 -
0.6 -
0.2-
0.0
1.4
_ 12 -
_i
E io -
i
0.8 -
55
0.6 -
0.2-
0.0
-SOdayAvg
CCC if Early Stage Fish Present
M
M
Highest 4-day avg in 30 days
O
2.5xCCCEarly Stage Fish Present
M
-CMCAcute
M
O
Mean
J FMAMJ JASOND
D_l 999-8 Klamath Straits Drain at Township Road (pump E) (KSDTR) -1999 continued
D 1999-25
-------
Model Configuration and Results
-in
Dissolved Oxygen (mg/l)
_i _i _i _i _i i.
DNj-&.a>oooNj.&.a>ooc
-
/Vy
-*-
J
'sn
/in
Chlorophyl-a (ug/l)
-^ NJ GO J
D O O O C
O Obs - - DO
'^^j\ ,;
F
xxy^vw w^
L TV"
-~~ S/AK
Criteria DOSAT
v.
/V
/^
f V
rv\/
iVW
^Vv^Nv g|
O-
MAMJ JASO
N
O Obs - - Chl-a Criteria
, ,
--
D
y>
J
1
0 9
0.8 -
^)
E, 0.7 -
I 0.6 -
| 0.5 -
o
| 0.4 -
* ฐ-3-
"5
i- 0.2 -
0.1 -
n
X jvAM
F
Q
M A M J J
O
-Jmj..jc.mmm,^_ p^^nvvn^v^ r\/\ ^ ^A^^,-.^^^
A S O
fw.
N
0
r
^^JV
D
Obs
O
m
a
i;
J~L_r_
O
A
/ \ r-A
Q \ /if
CP
/"S
"ToV "
c/\ / v/^
\r
J FMAMJ JASO
D_l 999-9 Klamath Straits Drain at Pump Station F (KSDPSF) -1999
r^
N
r
1
^
D
/
D 1999-26
-------
Model Configuration and Results
14
12
10
8
e'-
4
2 ^
0
o.
-Max
Mean
pH Criteria 2
Min
O Obs
- -pH Criteria 1
T7
o
0.8
M
M
O
0.7 -
0.6 -
_ 0.5 -
1)
-ง 0.4 -
O
0.3 -
0.2 -
0.1 -
0
Max
Min
Mean Obs
O
M
M
O
D
" I
3.5
3.25
3
2.75
2.5
2.25
I 1.75 -J
g 1.5
Z 1.25
1
0.75
0.5
0.25
0
JFMAMJJASOND
D_l999-9 Klamath Straits Drain at Pump Station F (KSDPSF) -1999 continued
D 1999-27
-------
Model Configuration and Results
0.0
1.4
1.2 i
1'0
r\ o
U.o -
0.6 \
E
< 0.4 -
0.2 -
0.0
1.4
1.2 -
1.0
0.8 -
0.6 ^
E
< 0.4 -
ro
"5
K 0.2 H
0.0
-SOdayAvg
CCC if Early Stage Fish Present
M
M
Highest 4-day avg in 30 days
O
2.5xCCCEarly Stage Fish Present
M
CMCAcute
M
O N
-Mean
J FMAMJ JASOND
D_l999-9 Klamath Straits Drain at Pump Station F (KSDPSF) -1999 continued
D 1999-28
-------
Model Configuration and Results
20
ฃ
"ฐ
-------
Model Configuration and Results
14
12
10
Q.
6 -
4
2 -I
0
-Max
Mean
pH Criteria 2
-Min
O Obs
- -pH Criteria 1
M
M
O
N D
O
0.8
0.7 -
0.6 -
0.5 -
0.4 -
0.3 -
0.2 -
0.1 -
0
Max Min Mean o Obs
M
M
O
D
Max Min Mean o Obs
4
3.75
3.5
3.25
3
2.75
2.5
1)2.25
ฃ 2
g 1.75
^ 1.5
1.25
1
0.75
0.5
0.25
0
JFMAMJJASOND
D_1999-10 Klamath Straits Drain at Highway 97 (KSD97) -1999 continued
D 1999-30
-------
Model Configuration and Results
-SOdayAvg
CCC if Early Stage Fish Present
0.0
M
M
Highest 4-day avg in 30 days
O
'2.5xCCCEarly Stage Fish Present
0.0
1.4
M
CMCAcute
M
O
Mean
_. 12 -
a 1.0
*
z 0.8
55
0.6 H
0.2 -
0.0
J FMAMJ JASOND
D_1999-10 Klamath Straits Drain at Highway 97 (KSD97) -1999 continued
D 1999-31
-------
Model Configuration and Results
Appendix D_2004
Water Quality Calibration Results
D 2004-1
-------
Model Configuration and Results
DO Saturation - -DO Criteria
o
50
40 -
3. 30 -]
ro
Q.
O
20 -
O
10 -
1
0.9 -
.-. 0.8 -
_i
fo.7-1
5 o.e
z
.3 0.5 -
c
o
E 0.4 -I
E
5 0.3 ^
ro
ฃ 0.2-
0.1 -
0
M
M
O N
D
.J'wwfif*1^^
^r--^^**u-i*ซซ*ปMซYtMซซi / O
M
M
O N
D
Max Min Mean o Obs
JFMAMJJASOND
D_2004-l Lost River at Keller Bridge (LRKB) -2004
D 2004-2
-------
Model Configuration and Results
14
12 -
10 -
8v
Q-
6 -
4 -
2 -
n
J
0 Q
0.7 -
0.6 -
_ 0.5 -
1)
-ง 0.4 -
^r
O
D- 0.3 -
0.2 -
0.1 -
n _
V.
1.5
1.25
1
ง
-ง 0.75
X
o
0.5
0.25
0
D_2004
- pH criteria 2 - pH criteria 1
^i^^___^_. ^^^ปปiniiiปftiKi||
XVv^*V" "^A-^
FMAMJ JASOND
/\
/ ^\
,/ ^^^-@,***ซ*^
J FMAMJ JASOND
A
/ \
--/---v
/ \
IZ""""A
/ \
::/ X
7 "r^ZT
JFMAMJ JASOND
-1 Lost River at Keller Bridge (LRKB) - 2004 continued
D 2004-3
-------
Model Configuration and Results
SOdayAvg
CCC if Early Stage Fish Present
0.0
F M A M J
Highest 4-day avg in 30 days
A S O N D
- 2.5xCCCEarly Stage Fish Present
JFMAMJJASO
D_2004-l Lost River at Keller Bridge (LRKB) - 2004 continued
D 2004-4
-------
Model Configuration and Results
50
M
M
ON
D
40 -
30 -
Q.
12(H
6
10 -
Max - Min - Mean Obs Chl-a criteria
Mt
t^uW^^ ' " ' Vl^^,
M
M
OND
JFMAMJJAS
-2 Lost River at Harpold Dam (LRHD) - 2004
OND
D 2004-5
-------
Model Configuration and Results
14 -i
12 -
10 -
8-P
ฃ
6 -
4 -
2
n
J
0 8
0.7 -
0.6 -
_ 0.5 -
"5)
-ง 0.4 -
^r
O
D- 0.3 -
0.2 -
0.1 -
n
V
1.5
1.25
1
1
-ง 0.75
g
z
0.5
0.25
0
D_2004
O Obs - pH criteria 1 - - pH criteria 1
- -/\ i- - - JUT- - -
^vx"! .^^^^-^ L
v^ ^-^ \/*^^^ ^U>- " -jปป"M>*SBi
FMAMJ JASOND
XX
,!/ -Xj,
JFMAMJJASOND
_[ \
;:i
7 \
/ \
~ / \
/ \ o
" / \ -
/ \ O /*4r1M
^^ar^f^l1 5
JFMAMJJASOND
-2 Lost River at Harpold Dam (LRHD) -2004 continued
D 2004-6
-------
Model Configuration and Results
0.0
30.0
.-. 25.0
_i
20.0
I
.i 15.0 -
10.0 -
5.0 -
o
"ro
"o
0.0
SOday Avg
CCC if Early Stage Fish Present
M
M
O
D
Highest 4-day avg in 30 days
2.5xCCCEarly Stage Fish Present
FMAMJ JASOND
CMCAcute
Mean
JFMAMJJASO
D_2004-2 Lost River at Harpold Dam (LRHD) -2004 continued
D 2004-7
-------
Model Configuration and Results
90 _
18 -
16 -
|14-
c 12 -
CD
D)
X 10 -
0
a) 8 -
"o c
]^^ ^^^\pjฅ^r
J FMAMJ JASOND
D_2004-3 Poe Valley Bridge at RM 27 (HPDS2) 2004
D 2004-8
-------
Model Configuration and Results
14 T
12 -
10
8__
0.
6 -
4
2
n _
J
n s
0.7
0.6
_ 0.5 -
"3)
-ง- 0.4 -
M-
O
o- 0.3 -
0.2
0.1
n -
V.
1.5
1.25
1
^)
-ง 0.75
X
0
0.5
0.25
0
D_2004
- pH criteria 1 - pH criteria 1
o
"v u Uk^-_1J^\ii^M/' ^ ^^^^ f^&r* v=r\j*ir-
FMAMJ JASOND
/\T
y_ x o
/ ^^v^^^-^W^
y
JFMAMJJASOND
7\
7 \
/ \
"/ \
/ \ QUfr. A^^
JFMAMJ JASOND
-3 Poe Valley Bridge at RM 27 (HPDS2) 2004 continued
D 2004-9
-------
Model Configuration and Results
SOdayAvg
CCC if Early Stage Fish Present
0.0
M
M
O
D
Highest 4-day avg in 30 days
- -2.5xCCCEarly Stage Fish Present
M
CMCAcute
M
O
Mean
1.4
1.2 -
E 1.0 H
0.8 -
JFMAMJJASO
D_2004-3 Poe Valley Bridge at RM 27 (HPDS2) 2004 continued
D
D 2004-10
-------
Model Configuration and Results
JFMAMJJASOND
Obs - Chl-a criteria
M
M
O N D
JFMAMJJASO
D_2004-4 Lost River at Crystal Springs (WDUS/LRWRC) -2004
D
D 2004-11
-------
Model Configuration and Results
14
12
4
2 H
0
0.8
0.7 -
0.6 -
O
0.4 -
0.3 -
0.2 -
0.1 -
0
Min
Obs
- pH criteria 2 - pH criteria 1
Max
Mean
Max Min Mean r Obs
M
M
M
JFMAMJJASOND
D_2004-4 Lost River at Crystal Springs (WDUS/LRWRC) -2004 continued
O N D
Max - Mm Mean o Obs
O N D
D 2004-12
-------
Model Configuration and Results
1.4
1.2
lh-ฐ
z 0.8 -
ro
| 0.6 -
E
< 0.4
ro
5
H 0.2
0.0
SOdayAvg
CCC if Early Stage Fish Present
M
M
O
D
Highest 4-day avg in 30 days
2.5xCCCEarly Stage Fish Present
J FMAMJ JASOND
J FMAMJ JASOND
D_2004-4 Lost River at Crystal Springs (WDUS/LRWRC) -2004 continued
D 2004-13
-------
Model Configuration and Results
20
18
16
0
Obs
DOSAT
- -DO Criteria
M
M
O N
D
50
40 -
_
Q.
S. 20 -]
_
O
10 -
- Chl-a criteria
O Obs
FMAMJ JASOND
1
0.9 -
.-. 0.8 -
_i
fo.7-1
5
z
ro 0.5 -
c
o
E 0.4 ^
0.3 ^
ro
0.2-
0.1 -
0
Max Min Mean o Obs
JFMAMJJASOND
D_2004-5 Upstream of Anderson Rose Dam (ARDMUS) -2004
D 2004-14
-------
Model Configuration and Results
14
12 -
8 -
0.
6 4
4
2 4
0
0.8
0.7 -
0.6 -
-ง- 0.4
O
o- 0.3 -
0.2 -
0.1
-Ji
1.5
1.25 -
1 -
D)
-ง 0.75 -
0.5 -
0.25 -
0
Max
Mean
pH criteria 2
Min
O Obs
- pH criteria 1
M
M
Max Min Mean r Obs
Max Min Mean o Obs
O N D
JFMAMJJASOND
JFMAMJJASOND
D_2004-5 Upstream of Anderson Rose Dam (ARDMUS) - 2004 continued
D 2004-15
-------
Model Configuration and Results
SOdayAvg
CCC if Early Stage Fish Present
0.0
0.0
1.4
1.2 -
E 1-0-1
no
U.o -
o 0.6 -
E
< 0.4 -
ro
"o
H 0.2 ]
0.0
M
M
Highest 4-day avg in 30 days
O
2.5xCCCEarly Stage Fish Present
M
CMCAcute
M
O
D
Mean
JFMAMJ JASON
D_2004-5 Upstream of Anderson Rose Dam (ARDMUS) - 2004 continued
D
D 2004-16
-------
Model Configuration and Results
20
18 -
16 -
|> 14
c 12 -
(D '^ 1
x 10
O
T3 p
-------
Model Configuration and Results
14
12 -j
10
4
2 -
0
0.?
-Max
Mean
pH criteria 2
Min
O Obs
- pH criteria 1
0.7 -
0.6 -
_. 0.5 -
B)
-ง- 0.4
M-
O
o- 0.3 -
0.2 -
0.1 -
0
C\n
1.5
1.25 -
1 --
-ง 0.75 -
0.5 -
0.25 -
0
M
M
O
Max Min Mean o Obs
M
M
Max Min Mean o Obs
N D
ON
JFMAMJJASOND
D_2004-6 Lost River at East West Road (LREW) - 2004 continued
D 2004-18
-------
Model Configuration and Results
SOdayAvg
CCC if Early Stage Fish Present
0.0
0.0
M
M
Highest 4-day avg in 30 days
O
2.5xCCCEarly Stage Fish Present
M A
CMCAcute
M J
Mean
O
J FMAMJ JASON
D_2004-6 Lost River at East West Road (LREW) - 2004 continued
D 2004-19
-------
Model Configuration and Results
on
Dissolved Oxygen (mg/l)
^CDOOOMJ^CDOOC
2 -
n -
0
^/t
Obs
Win
DOSAT - -DO Criteria
|WCx
/ ^\ *
--W
^V
\l
f
k/
i v \
\7
V
o
m *
i\ry
O
%p
j!
/
J F M
50 -i
40 -
I 30-
ro
i
Q.
220-
-C
O
10 -
n
V.
1 -
0.9 -
_ 0.8 -
1 0.7-
J 0.6 -
.| 0.5 -
o
E 0.4 -
< 0.3-
CD
-ง->
(5 0.2 -
0.1 -
0 -
D_2004
A
O Obs
Chl-a criteria
M
F M
A
\
V
\
M
J J A
A
j j
fi
\
\
A
SON
D
fl
-
S O N D
/y^^
/ \
J F M
\-l P-Canal (PC) -
\
V
\
A
2004
M
;-Aaปu
ifctota
A
A
S O N D
D 2004-20
-------
Model Configuration and Results
14
12 -
8 -
Q.
6 -
4 -
2
n
J
n s
0.7 -
0.6 -
0.5 -
13)
-ง 0.4 -
S
ฐ- 0.3 -
0.2 -
0.1 -
n
1.5
1.25
I 0.75
|
0.5
0.25
0
D_2004
Mean o Obs
pH criteria 2 - pH criteria 1
"^II -J. ' ' ฎrj~'r' Q ^'V
"^ >^-^-- ^K^J
FMAMJ JASOND
^^^^ ^n_*43L. @
^^--: ^
JFMAMJJASOND
^^^
"X" "T
ฉ/^
JFMAMJJASOND
-7 P-Canal (PC) - 2004 continued
D 2004-21
-------
Model Configuration and Results
SOdayAvg
CCC if Early Stage Fish Present
0.0
1.4
1.2
I 1.0
0.8 -
0.6 -I
E
< 0.4 -
ro
H 0.2
0.0
M
M
O
D
Highest 4-day avg in 30 days
2.5xCCCEarly Stage Fish Present
M
CMCAcute
M
O
Mean
J F M A M J
D_2004-7 P-Canal (PC) - 2004 continued
D
D 2004-22
-------
Model Configuration and Results
90 -,
18
16 -
|> 14 _
ง12-
x 10 -
O
-------
Model Configuration and Results
14
12 -
10
e
4
2 4
0
0.8
0.7 -
0.6 -
~ ฐ-5"
B)
-ง- 0.4
0.3 -
0.2 -
0.1
1.5
1.25 -
1 -
D)
-ง 0.75 -
0.5 -
0.25 -
0
Max
Mean
pH criteria 2
Min
O Obs
- pH criteria 1
M
M
O
Max Min Mean o Obs
O
Max Min Mean o Obs
JFMAMJJASOND
JFMAMJJASOND
D_2004-8 Klamath Straits Drain at Stateline Road (KSDSR) - 2004 continued
D 2004-24
-------
Model Configuration and Results
SOdayAvg
CCC if Early Stage Fish Present
0.0
1.4
FMAMJ JASOND
Highest 4-day avg in 30 days .... 2.5xCCCEarly Stage Fish Present
1.2 -
no
U.o -
0.6 \
E
< 0.4 -
ro
"o
0.0
1.4
M
CMCAcute
M
OND
Mean
1.2 -
o 0.6 -
E
< 0.4 -
ro
5
H 0.2 -I
0.0
JFMAMJJASOND
D_2004-8 Klamath Straits Drain at Stateline Road (KSDSR) - 2004 continued
D 2004-25
-------
Model Configuration and Results
90 T
18 -
o> 14 _
o)
x 10 -
O
T3 p
0 O -
"o K
CO D -
CO
b 4-
2
n
O Obs DOSAT --DO Criteria
"A"
rtV^ vwr V
/ v^ ^Xj/*^
v M u
Vnf A p]fs*ฅ~*^y/|fM^%ปปiF11*'
A^^H^
JFMAMJJASOND
^n
40 -
ง 30
CD
Q.
o 9f)
6
10 -
n
V.
1
0.9
_ 0.8
| 0.7
J 0.6
.| 0.5
o
E 0.4
< 0.3
CD
i? 0.2
0.1
0
D_2004
/-\ Ohc
- -Chl-a
criteria
O o
FMAMJ JASOND
ry
^<^
(I O
h ฐ ฐ y
n x^7^ /kv7^*^^
i
/ y ii
uปy\liM"
J FMAMJ JASOND
-9 Klamath Straits Drain at Railroad (KSDM) - 2004
D 2004-26
-------
Model Configuration and Results
14
12 -j
10**'
0.
8 -
6 -
4
2 -
0
O.E
-Max
Mean
pH criteria 2
O
-Min
Obs
- pH criteria 1
0.7 -
0.6 -
_. 0.5 -
B)
-ง- 0.4
M-
O
o- 0.3 -
0.2 -
0.1 -
0
1.5
1.25 -
1 -
-ง 0.75 -
x
o
Z
0.5 -
0.25 -
0
M
M
O
Max Min Mean o Obs
0
o
M
M
Max Min Mean o Obs
JFMAMJ JASON
D_2004-9 Klamath Straits Drain at Railroad (KSDM) - 2004 continued
0 N D
D 2004-27
-------
Model Configuration and Results
SOdayAvg
CCC if Early Stage Fish Present
no
U. O -
M
M
O
D
Highest 4-day avg in 30 days
2.5xCCCEarly Stage Fish Present
J FMAMJ JASON
D_2004-9 Klamath Straits Drain at Railroad (KSDM) - 2004 continued
D 2004-28
-------
Model Configuration and Results
Appendix E_2004
Evaluation of Macrophyte Mass and Diel DO Comparison
E 2004-1
-------
Model Configuration and Results
10000
~ 1000 --
.E
a
Q.
O
100 -
10 --
1 --
0.1
a:
O
o:
tn
o
LLJ
a:
o
D.
Monitoring Data Simulated Minimum . . . . Simulated Maximum
E_2004-l Comparison of macrophyte mass (g/m2) along the Lost River
E 2004-2
-------
Model Configuration and Results
20
18
16
14
12
10
8
6
4
D)
E,
c
0)
D)
>>
X
O
D
0)
"5
V)
V)
2
LROG
LRDR
LREW
KSD97
Monitored Minimum
Simulated Minimum
A Monitored Maximum
- - Simulated Maximum
E_2004-l Comparison of diel DO along the Lost River
E 2004-3
-------
Model Configuration and Results
Appendix F_1999
Sensitivity Analysis
F 1999-1
-------
Model Configuration and Results
-------
Model Configuration and Results
1.0
0.8
0.6
0.4
0.2
0.0
-0.2 -
-0.4 -
-0.6 -
-0.8 -
-1.0
ro
m
.2
O
Q
O
-0.8x BOD decay rate (k)
-1.2 x BOD decay rate (k)
F_l 999-3 BOD decay rate sensitivity simulation - Lost River at Harpold Dam (LRHD)
7.0
6.0
5.0
4.0
3.0
2.0
1.0
0.0
-1.0
-2.0 --
-3.0 -
-4.0 -
-5.0 -
-6.0 -
-7.0
O
Q
c
ro
.c
O
0.8 x Macrophyte grow th rate
1.2 x Macrophyte grow th rate
F_l 999-4 Macrophyte growth rate sensitivity simulation - Lost River at Harpold Dam
(LRHD)
F 1999-3
-------
Model Configuration and Results
ro
m
.2
O
Q
O
1.0
0.8
0.6
0.4
0.2
0.0
-0.2 -
-0.4 -
-0.6 -
-0.8 -
-1.0
F_l 999-5 SOD sensitivity simulation - Lost River at Keller Bridge (LRKB)
1.0
0.8
0.6
0.4
0.2
0.0
-0.2 -
-0.4 -
-0.6 -
-0.8 -
-1.0
O
Q
ro
.c
O
-0.8x Algal growth rate
-1.2 x Algal growth rate
O
F_l 999-6 Algal growth rate sensitivity simulation - Lost River at Keller Bridge (LRKB)
F 1999-4
-------
Model Configuration and Results
1.0
0.8
0.6 -
0.4 -
ro
CD
o
Q
o
-0.2 -
-0.4 -
-0.6 -
-0.8 -
-1.0
-0.8x BOD decay rate (k)
-1.2x BOD decay rate (k)
M A M
O N D
F_l 999-7 BOD decay rate sensitivity simulation - Lost River at Keller Bridge (LRKB)
2.0
1.5
1.0 -
-2.0
0.8x Macrophyte growth rate
-1.2x Macrophyte growth rate
F_l 999-8 Macrophyte growth rate sensitivity simulation - Lost River at Keller Bridge
(LRKB)
F 1999-5
-------
Model Configuration and Results
20
18 -
16 -
14
12 -
lz- 1
s
(0
2 -^
0
Existing Condition
Nitrogen Reduction Only
Nitrogen and Phosphorus Reduction
JFMAMJ JASOND
F_l 999-9 Nutrient load reduction simulation - Lost River at Harpold Dam (LRHD)
20
18 -
16 -
Nitrogen and Phosphorus Reduction
Existing Condition
JFMAMJ JASOND
F_1999-10 Nutrient load reduction simulation - Lost River at Crystal Springs (LRWRC)
F 1999-6
-------
Model Configuration and Results
20
18 -
16 -
a) 14
c 12 -
lz- 1
T3 c
O5
18 -
16
14
12 -
'^ i
U3
15
-DO-Existing Condition
DO-lncreased Shade
JFMAMJ JASON
F_1999-12 Increased shade simulation - Lost River at Crystal Springs (LRWRC)
F 1999-7
-------
Model Configuration and Results
20
18 -
16
-DO-Existing Condition
JFMAMJ JAS
F_1999-13 Increased shade simulation - Anderson-Rose Dam (LRAR)
0
30 -
-DO-Existing Condition
DO-lncreased Shade
25
05
E
20 -
O5
O 15
0 10
05 IU
U5
XX^A^^W*
5 -
j^^
J F M A M J J A S 0
F_1999-14 Increased shade simulation - Lost River at East-West Road (LREW)
F 1999-8
-------
Model Configuration and Results
16
14
e 12
O5
f 1ฐ
05
^*^ o
o
6 -
o
C/5
^ 4
2 -
0
A/
-DO-Existing Condition
V
V
DO-lncreased Shade
f\
JFMAMJ JA
F_1999-15 Increased shade simulation - Tule Lake (TLTO)
0
16
14
^ 12 -
05
E
r 10-1
-------
Model Configuration and Results
14
12
10
8
x
6 -
-DO-Existing Condition
DO-lncreased Shade
M
M
0
F_1999-17 Increased shade simulation - Klamath Straits Drain at Township Road
(KSDTR)
16
14
^ 12 -
O)
E
IT 10 -]
05
S "
O
U3
5 4
2 -
-DO-Existing Condition
DO-lncreased Shade
M
M
0
F_1999-18 Increased shade simulation - Klamath Straits Drain at Pump Station F
(KSDPSF)
F 1999-10
-------
Model Configuration and Results
16
14
e 12
O5
-------
Model Configuration and Results
Appendix G
TMDL Scenario
G-1
-------
Model Configuration and Results
MAMJ JASOND
20
18 -
16 -
14
12 -
"8 8 -
o
ta
en
b
6 -I
4
2 -I
0
DO - TMDL Scenario_30day
- -DO Criteria
20
18 -
16 -
1 14
-!o
"O o
-------
Model Configuration and Results
0.14
0.45
0.4
0.35
0.3
0.25
3
Q_
0.15 -
0.1 -
0.05
0.75
0.5 -
=5)
0.25 -
0
MAM
O N D
PO4-TMDL Scenario
\.
NOX-TMDL Scenario
JFMAMJJASOND
JFMAMJJASOND
LRGR continued
G-3
-------
Model Configuration and Results
1.4
s 1.0 -
i 0.8 -
TO
o 0.6 H
< 0.4 -
30dayAvg-TMDL Scenario
CCC if Early Stage Fish Present
J F
LRGR continued
MAMJJASO
Highest 4-day avg in 30 days-TMDL Scenario
... -2.5xCCCEarly Stage Fish Present
N
D
MAM
- - -CMCAcute
J A S O
TMDL Scenario
N
D
M
M
O
D
G-4
-------
Model Configuration and Results
50
40 -
30 -
Q.
2 20
_
O
10 -
14
Chl-a-TMDL Scenario
Chl-a Criteria
M
O N D
Q.
12 -
10 -
8
6 -I
0
pH - TMDL Scenario - -pH Criteria 2 - -pH Criteria 1
JFMAMJJAS
LRGR continued
O N D
G-5
-------
Model Configuration and Results
DO -TMDL Scenario
AMJ JASOND
DO - TMDL Scenario_30day
- -DO Criteria
AMJ JASOND
DO - TMDL Scenario 7day
- -DO Criteria
M
AMJJASOND
LRKB
G-6
-------
Model Configuration and Results
0.14
M
M
OND
PO4-TMDL Scenario
JFMAMJJASOND
0.75
0.5
"5)
x
o
0.25
0
NOX-TMDL Scenario
JFMAMJJASOND
LRKB continued
G-7
-------
Model Configuration and Results
30dayAvg-TMDL Scenario
CCC if Early Stage Fish Present
0.6
0.5 -.
0.4 -
|
.3 0.3 -
0.2 -
0.1 -
O)
o
E
0.0
M
M
O N
D
Highest 4-day avg in 30 days-TMDL Scenario
. 2.5xCCCEarly Stage Fish Present
1.4
^ 1.2
15>
E, 1.0
z 0.8 -
ro
o 0.6
< 0.4
"ro
ฃ 0.2
0.0
1.4
1.2
^ 0.8 -
ro
o 0.6 -
E
< 0.4 -
ro
"5
H 0.2 4
J F
LRKB continued
M
M
JFMAMJJASOND
... -CMCAcute TMDL Scenario
O N
G-8
-------
Model Configuration and Results
50
40 -
.30 -
Q.
2 20
^
6
10 -
14
Q.
12 -
10 -
8 -I
6 -
4
2 -I
0
Chl-a-TMDL Scenario
Chl-a Criteria
M
M
O N D
pH-TMDL Scenario - -pH Criteria 2 - -pH Criteria 1
MAM
O N D
LRKB continued
G-9
-------
Model Configuration and Results
20
18 -
14
ฃ10
w 6 -
(0
DO- TMDL
- -DO Criteria
2 -
M
M
O
20
18 -
10 -
T3
-------
Model Configuration and Results
ฃ
0.14
0.12 -
0.1 -
0.08 -
0.06
0.04 -
0.02 -
0
NH4-TMDL Scenario
JFMAMJJASOND
0.45
0.4 -
0.35 -
0.3 -
0.25 -
0.15 -
0.1 -
0.05 -
JFMAMJJASOND
0.75
0.5 -
x
o
0.25 -
0
NOX-TMDL Scenario
JFMAMJJASOND
LRHD continued
G-11
-------
Model Configuration and Results
30dayAvg-TMDL Scenario
CCC if Early Stage Fish Present
0.6
0.5 -
0.4 -
|
.i 0.3 -
0.2
O)
o
E
o.H
0.0
M
M
O
D
Highest 4-day avg in 30 days-TMDL Scenario
2.5xCCCEarly Stage Fish Present
1.2 -
r\ o
U. O -
0.6 -
E
< 0.4 -
ro
"5
H 0.2 -I
J F
LRHD continued
M
M
O
G-12
-------
Model Configuration and Results
50
40 -
30 -
Q.
e 20
_
O
10 -
Chl-a-TMDL Scenario
Chl-a Criteria
14
MAM
ASOND
12 -
10 -
0
pH-TMDL Scenario - -pH Criteria 2 - -pH Criteria 1
JFMAM
LRHD continued
OND
G-13
-------
Model Configuration and Results
FMAMJJASOND
20
18 -
16 -
14
ง 12 H
D)
x 10 -j
O
T3 o I
CD O ~
D)
x
O
"O
CD
^
O
w
en
b
6 -
4
2
0
J
20
18
16 ^
14
10 -
8 -
6 -
4 -
2 -
- ^_
^W ^ ^^B ซ ^^B
DO- TMDL Scensrio 30d3y * DO Criteris
^^^ ^~-
"^^\^ /^^
._,_._._._, _ 7^7^^-^-*=i^^rT ,_-._-._
FMAMJ JASOND
^~^x_^~- -__^__ ^_
DO TMRI ^rpn^rio 7rl^\/ -DO Pritprm
^^^^--^ j~~~^~
^'^ ,^/
\^-^^-^-^jC~^^-
FMAMJ JASOND
LRSP
G-14
-------
Model Configuration and Results
0.16
0.14 -
0.12 -
0.1 -
0.08 -
0.06
0.04 -
0.02 -
0
0.45
0.4 -
0.35
0.3 -
0.25 -
s
D.
0.15 -
0.1 -
0.05 -
0.75
0.5 -
x
o
0.25 -
0
J F
LRSP continued
M
NH4-TMDL Scenario
M
MAM
NOX-TMDL Scenario
M
M
O N
OND
O N
G-15
-------
Model Configuration and Results
0.6
30dayAvg-TMDL Scenario
CCC if Early Stage Fish Present
5~ 0-5 -i
15)
*OA-\
i
.5 0.3 -
o
E
E 0.2 ^
"ro
o 0.1 H
0.0
1.4
MAMJ JASO
Highest 4-day avg in 30 days-TMDL Scenario
... -2.5xCCCEarly Stage Fish Present
N
D
1.2 -
O)
E
1.0 -
0.8 -
0.6 -
0.4 -
TO
o 0.2 H
0.0
M
M
O
D
J F
LRSP continued
CMCAcute
TMDL Scenario
M
M
O
N
D
G-16
-------
Model Configuration and Results
50
40 -
30 -
Q.
2 20 -]
_
O
10 -
Chl-a-TMDL Scenario
Chl-a Criteria
14
F M A M J
A S O N D
12 -
10 -
8 -
6 -
4 -
2
0
pH-TMDL Scenario - -pH Criteria 2 - -pH Criteria 1
JFM
LRSP continued
M
O N
G-17
-------
Model Configuration and Results
FMAMJ JASOND
LROG
G-18
-------
Model Configuration and Results
0.45
0.4 -
0.35 -
0.3 -
0.25 -
5
CL
0.15 -
0.1 -=
0.05 -
0.75
0.5 -
"Si
E
0.25 -
0
M
M
O N D
J F M
LROG continued
M
JFMAMJJASOND
O N D
G-19
-------
Model Configuration and Results
30dayAvg-TM DL Scenario
CCC if Early Stage Fish Present
0.6
0.5 -
O)
0.4 -
o
E
0.2 H
1.4
1.2 -
O)
1.0 -
0.8 -
0.6 -
0.4 -
0.2 -
0.0
1.4
IlO-
i 0.8 -
TO
o 0.6 H
I 0.4 H
CO
H 0.2 -j
J F
LROG continued
MAMJJASO
Highest 4-day avg in 30 days-TMDL Scenario
- - - -2.5xCCCEarly Stage Fish Present
M
M
O
-CMCAcute
TMDL Scenario
M
M
O
D
D
N
D
G-20
-------
Model Configuration and Results
50
40 -
30 -
Q.
2 20
^
6
10 -
Chl-a-TMDL Scenario
- - Chl-a Criteria
14
M
M
OND
12 -
10 -
Q.
0
pH-TMDL Scenario - -pH Criteria 2 - -pH Criteria 1
J FMAM
LROG continued
OND
G-21
-------
Model Configuration and Results
JFMAMJJASOND
to
to
20
18
16 -
14 -
12 -
y
DO TMDI ^rpnario ^Drlav _ - ซ-nO Pritpria
^ ^^^ ^^ '
"^^^^ ^^
\ s^
V -J
^^XIZII/^"
M
M
OND
cs
x
O
20
18
16 ^
14
10 -
to
Q
2 -]
0
LRWRC
r-^^
1
V
.__.__
DO Tl\/ini ^rpnario 7Ha\/ -DO Pritpria
^~~^--_ ^-- ~~ -~
^^V-^\ /^^
A- A 7^
\ r ~\ ~ w
"V \ / T^A"7
^ VA/^
~\/~
M
M
O
D
G-22
-------
Model Configuration and Results
0.16
0.14 -
0.12 -
0.1 -
"Si
0.08 - -
0.06 -
0.04 -
0.02 -
0.45
0.4 -
0.35 -
0.3 -
0.25 -
5
Q_
0.15 -I
0.1 -"
0.05 -
0.75
0.5 -
0.25 -
0
M
M
NOX-TMDL Scenario
J F M A M
LRWRC continued
O N D
JFMAMJJASOND
O N D
G-23
-------
Model Configuration and Results
30dayAvg-TMDL Scenario
CCC if Early Stage Fish Present
1.4
1.2 -
.. 1.0 ^
0.8 -
o
<
ro
0.6 -
0.4 -j
0.2 -
0.0
1.4
1.2 -
0.8 -
o 0.6 -
E
< 0.4 -
ro
5
H 0.2 4
J F
LRWRC continued
MAMJJASO
Highest 4-day avg in 30 days-TMDL Scenario
... -2.5xCCCEarly Stage Fish Present
D
MAM
. -CMCAcute
J A S O
TMDL Scenario
M
M
O
D
G-24
-------
Model Configuration and Results
50
40
D)
.3- 30 -
ro
Q.
ง 20 ^
6
10 -
Chl-a-TMDL Scenario
- - Chl-a Criteria
JFMAMJJASOND
14
Q.
12 -
10 -
8 -L
6 -
4 -
2 -
0
pH-TMDL Scenario - -pH Criteria 2 - -pH Criteria 1
JFMAMJJASOND
LRWRC continued
G-25
-------
Model Configuration and Results
20
18 -
16 - -
14
O)
10 -
T3
CD
"O R I
(/) D -
5
4
2 ^
0
DO-TMDL Scenario
. -DO Criteria
MAMJJASOND
20
16 -
1 14
<= 12 -
CD '
D)
x 10 H
o
T3
CD
| 6^
S 4
2
0
20
__^^
DO-TMDL Scenario_30day - -DO Criteria
"~ "^ - S~
^^_ __ ^~"
^- -
MAMJ JASOND
16 -
14
ง 12 H
D)
x 10 ^
o
T3 o I
CD O ~
6 -
4
2 H
0
\J^^
DO- TMDL Scensrio 7d3y _ ,. DO Criteris
,s^
^-^^^ ^^\^^'^
^^^W7^
MAMJ JASOND
LRDR
G-26
-------
Model Configuration and Results
0.7
0.6 -
0.5 -
0.3 -
0.2 -
0.1 -
0.7
0.6 -
0.5 -
fo.4
O 0.3
D.
0.2
0.1 -j
0
1.25
=5)
E
1 -
0.75 -
0.5 -
0.25 -
0
M
M
O N D
JFMAMJJASOND
JFMAMJJASOND
LRDR continued
G-27
-------
Model Configuration and Results
30dayAvg-TMDL Scenario
CCC if Early Stage Fish Present
MAMJJASO
Highest 4-day avg in 30 days-TMDL Scenario
... .2.5xCCCEarly Stage Fish Present
N
D
M
M
O
D
-CMCAcute
TMDL Scenario
J F
LRDR continued
M
M
O
D
G-28
-------
Model Configuration and Results
50
40 -
30 -
Q.
2 20
^
6
10 -
Chl-a-TMDL Scenario
Chl-a Criteria
MAM
A S O N D
14
pH-TMDL Scenario - -pH Criteria 2 - -pH Criteria 1
0
J FMAMJ JASOND
LRDR continued
G-29
-------
Model Configuration and Results
20
18 -{
14 -
O)
x 10
O
T3 p
CD o
"O K
co D
CO
2
0
20
18 -{
x
O
T3
CU
j 6^
1 4
2
0
20
18
'14 -
O)
X
O
T3
CU
"O
CO
CO
6 -I
4
2 -I
0
LR39
DO-TMDL Scenario
- -DO Criteria
FMAMJJASOND
^-
DO- TMDL Scenario SOday DO Criteria
-^^___^ ^___X
~^^--_ _^ ^^
FMAMJJASOND
~
/ -^_
DO- TMDL Scenario 7day DO Criteria
, /s
^^_ fT
^^^W-^^^^^^
FMAMJ JASOND
G-30
-------
Model Configuration and Results
JFMAMJJASOND
JFMAMJJASOND
1.25
J F
LR39 continued
MAM
S O N D
G-31
-------
Model Configuration and Results
30dayAvg-TM DL Scenario
CCC if Early Stage Fish Present
JFMAMJJASOND
Highest 4-day avg in 30 days-TMDL Scenario .... 2.5xCCCEarly Stage Fish Present
FMAMJJASOND
CMCAcute
TMDL Scenario
J F
LR39 continued
M
M
O
G-32
-------
Model Configuration and Results
50
40
O)
5-30 -
ro
Q.
ง 2ฐ
6
10 -
Chl-a-TMDL Scenario
- - Chl-a Criteria
14
M
M
O N D
0
pH-TMDL Scenario - -pH Criteria 2 - -pH Criteria 1
J F M
LR39 continued
M
O N
G-33
-------
Model Configuration and Results
90 -I
18 -
1R
^) 14
c 19
5 ^ "
O)
51 m
x 10-
T3 o
>
0 fi
O
(0
b 4
9
f)
90
18 -
1R
O) 14
c 19
W l2 "
o)
51 m
O
T3 p
^
ฐ R
(0
S 4
0
n
90
18 -
1R
O) 14
c 19
5 n^ "
cs
51 10
x 10-
~ฐ ft
>
o K
0)
Q 4
o
n -
DO- TMDL Scenario --DO Criteria
J^^* JW\
^V - ^^^f,^~~^ f]\^>r\ ./v
\f \r^>% /V**ih^ i Ak lAlA (f
"vJ" "v^ T ^r^mi~r\\
t \ it- i ii
""M """ V "T"
........... |,_J|1LJ - . .
FMAMJJASOND
DO- TMDL Scenario SOday DO Criteria
\. ^
/ ^ ^-~-^ ^ ~
i^^^-. . ^-^L-
JFMAMJJASOND
DO Tl\/ini ^rpnario 7Ha\/ -DO Pritpria
_/ ^~~X ^"^~
\T AT^, /^^ Z!
x/ AT^TY n/
~ \ '
" y\r - - - -
\r
M
AMJ JASOND
LRAR
G-34
-------
Model Configuration and Results
0.8
M
O
N D
0.7 -
0.6 -
0.5 -
O)
E
0.4 -
0.3 -
0.2 -
0.1 --
JFMAMJJASOND
1.25
1 -
~ 0.75 -
O)
E
0.5 -
0.25 -
0
NOX-TMDL Scenario
J F M A M
LRAR continued
O N D
G-35
-------
Model Configuration and Results
30dayAvg-TMDL Scenario
CCC if Early Stage Fish Present
0.0
1.4
1.2 -
10
no
U.o -
o 0.6
E
< 0.4
ro
"o
H 0.2
J F
LRAR continued
M
M
O
D
Highest 4-day avg in 30 days-TMDL Scenario
2.5xCCCEarly Stage Fish Present
MAM
- - -CMCAcute
J A S O
TMDL Scenario
D
M
M
O
D
G-36
-------
Model Configuration and Results
50
40 -
30 -
ro
">s
20 -
6
10 -
Chl-a-TMDL Scenario
Chl-a Criteria
JFMAMJJASOND
14
Q.
12 -
10 -
8 -
6 -
4
2 -I
0
pH-TMDL Scenario - -pH Criteria 2 - -pH Criteria 1
JFM
LRAR continued
M
ON
D
G-37
-------
Model Configuration and Results
90 -,
18 -
1fi
O) 1A
d 19
S n^ "
O)
51 m
O
T3 p
>
0 fi
CO
S 4
0
0 -
i
90
18 -
1fi
^) 14
c 19
(D IZ
CS
51 m
O
T3 p
CO D
CO
2 -
n
90 -,
18 -
1R
^) 14
ฃ
c 19
(5 l2 "
cs
51 m
O
T3 p
"O K
CO
Q .
9
n -
DO TMDI ^ronnrin _- PA DO Pritorin _- OR DO Pritorin
I
^A^^^^^M \}> i In || I |j| 1 1 || 1 I |ILiii M^"
\T^%\m^ jl| j i i || | f I f
^'^ll|fW|||!'^f^
IFMAMJ JASOND
DO- TMDL Scenario SOday OR DO Criteria
""~^ "^x"^- ^~^-^ ^^~/
^^^-^ ^~^^~"
JFMAMJ JASOND
DO- TMDL Scenario 7day OR DO Criteria
"^^^x^^ y^
^\^- ,^^
^^xAv-^-^^^^^^
MAMJ JASOND
LRSR
G-38
-------
Model Configuration and Results
MAM
S O N D
0.7
0.6 -
0.5 -
io.4-
O 0.3 -j
Q_
0.2 -
0.1
0
M
M
1.25
J F
LRSR continued
M
M
O N D
N D
G-39
-------
Model Configuration and Results
0.0
1.4
1.2 -
E
0.8 -
30dayAvg-TMDL Scenario
. CCC if Early Stage Fish Present
J F
LRSR continued
M
M
O
Highest 4-day avg in 30 days-TMDL Scenario
2.5xCCCEarly Stage Fish Present
M
O
CMCAcute
TMDL Scenario
M
O
D
G-40
-------
Model Configuration and Results
50
40 -
30 -
Q.
2 20
^
6
10 -
Chl-a-TMDL Scenario
- - Chl-a Criteria
J F M A M J
14
A S O N D
4
2 -{
0
pH-TMDL Scenario - -pH Criteria 2 -pH Criteria 1
J F M
LRSR continued
M
O N
D
G-41
-------
Model Configuration and Results
M
M
LREW
G-42
-------
Model Configuration and Results
1.25
1 -
~ 0.75 -
-ง
x
i 0.5-
0.25 -
0
M
M
O N
JFMAMJJASOND
NOX-TMDL Scenario
J F
LREW continued
M
M
^ i i
J J A
O N
G-43
-------
Model Configuration and Results
50
40 -
-30 -
Q.
2 20
^
6
10 -
Chl-a-TMDL Scenario
14
M
M
O N D
Q.
12 -
10 -
8 -
6 -
0
pH-TMDL Scenario - -pH Criteria 2 - -pH Criteria 1
J F M
LREW continued
M
O N
G-44
-------
Model Configuration and Results
20
CD
CO
10-1
T3
CD
"5
CO
3 5
DO- TMDL Scenario
- -DO Criteria
M
M
TLTO
G-45
-------
Model Configuration and Results
0.4
0.35
0.3
0.25
B)
-ง- 0.2
*
z 0.15
0.1
0.05
0
0.3
0.25 -
0.2 -
0.15 -
o
Q_
0.1 -
0.05 -
0.5
0.25 4
0
J F
TLTO continued
M
M
NH4-TMDL Scenario
^^^^^_^^^^^
M
O N
JFMAMJJASOND
NOX-TMDL Scenario
M
O
G-46
-------
Model Configuration and Results
90
l^
ii
^
?
= in
Chlorophy
n c
0
14 -,
12 -
in
-
x
Q.
A
9
n -
Chl-a-TMDL Scenario
V
JFMAMJJASOND
pH-TMDL Scenario - -pH Criteria 2 - -pH Criteria 1
~*~*-*>~*^^^ ^^s-*,**; "
J F M
TLTO continued
M
O N
D
G-47
-------
Model Configuration and Results
20
DO-TMDL
- -DO Criteria
0)
O)
X
O
T3
0)
10 -
M
M
LKL
G-48
-------
Model Configuration and Results
0.4
0.35 -
0.3 -
0.25
-ง- 0.2 -
I
z 0.15
0.1 -
0.05 -
0.35
0.3 -
0.25 -
0.2 -
O 0.15 -
Q_
0.1 -
0.05 -
0
0.5
0.25 4
0
J F
LKL continued
NH4-TMDL Scenario
iiilii^
F M A M
NOX-TMDL Scenario
M
M
S O N D
JFMAMJJASOND
OND
G-49
-------
Model Configuration and Results
10
Q
1)
=s 6
ro
i
">s
_c
Q.
O A
O
_c
O
9
n
14 -i
12 -
m
Q
x
Q.
R
A
O
n -
Qlil-3-j|\/IQI_ Scenario
I
JFMAMJ JASOND
pH TMDL Scenario pH Criteria ฐ pH Criteria 1
-^^ t_rfซlf-WJ>ซ>^l^.ซ^(^^-^*V_1^|./l
J FMAMJ JASOND
LKL continued
G-50
-------
Model Configuration and Results
DO-TMDL Scenario - -CA DO Criteria - -OR DO Criteria
JFMAMJ JASOND
DO- TMDL Scenario_30day
OR DO Criteria
JFMAMJJASOND
DO- TMDL Scenario_7day
- -OR DO Criteria
KSDSR
G-51
-------
Model Configuration and Results
0.4
0.35 -
0.3 -
0.25 -
O
Q-
0.2 -
0.15 -
0.1 -
0.05 -
1.75
1.5 -
1.25 -
=3, '
g 0.75 -
M
M
M
O N D
M
NOX -TMDL Scenario
J F M A M
KSDSR continued
O N D
O N D
G-52
-------
Model Configuration and Results
30dayAvg-TMDL Scenario
. CCC if Early Stage Fish Present
1.4
1.2 -
E
0.8 -
o 0.6
E
< 0.4
ro
"o
H 0.2
0.0
M
M
O
Highest 4-day avg in 30 days-TMDL Scenario
... -2.5xCCCEarly Stage Fish Present
MAM
- -CMCAcute
J J A S O
TMDL Scenario
JSk.
J F M
KSDSR continued
M
O
D
N D
N D
G-53
-------
Model Configuration and Results
50
40 -
.30 -
Q.
2 20 -]
^
6
10 -
14
Chl-a-TMDL Scenario
Chl-a Criteria
FMAMJ JASOND
12 -
10 -
Q.
6 -
4
2 -{
0
pH-TMDL Scenario - -pH Criteria 2 -pH Criteria 1
J F M
KSDSR continued
M
O N
D
G-54
-------
Model Configuration and Results
on _,
18 -
1fi
O) 1A
d 19
S n^ "
o)
51 m
O
"S a
>
o fi
CO
S 4
0
n -
i
on
18 -
1R
^) ^\A
c 19
(D IZ
cs
51 in
O
T3 o
CO D
CO
2 -
n
9n _,
18 -
1R
^) 14
ฃ
c 19
(5 l2 "
cs
51 m
O
T3 p
"O o
(/)
9
n -
J\ A K
II " ~ v\ 7i " ^^J
W^\ ix*^
VSflft Art A**rflr
W^^^i^
IFMAMJ JASOND
DO- TMDL Scenario SOday DO Criteria
^^^-- --TA
\^_ ^X^
JFMAMJ JASOND
DO- TMDL Scenario 7day DO Criteria
/x A-
"A" "/ vi ~/\~ --/-
Vx^A ^^s
VA^^J^^^A
J F
KSDTR
M
AMJ JASOND
G-55
-------
Model Configuration and Results
0.4
0.35
0.3
0.25
B)
-ง- 0.2
*
z 0.15
0.1
0.05
0
0.45
0.4
0.35
0.3
"& 0.25
3
CL
0.15
0.1
0.05
0
JFMAMJJASOND
0
J F M
KSDTR continued
JFMAMJJASOND
O N D
G-56
-------
Model Configuration and Results
30dayAvg-TMDL Scenario
CCC if Early Stage Fish Present
0.0
1.4
1.2 -
ttcH
0.8 -
(0
o 0.6
E
< 0.4
ro
5
H 0.2 -
0.0
J F
KSDTR continued
M
M
O
Highest 4-day avg in 30 days-TMDL Scenario
2.5xCCCEarly Stage Fish Present
MAM
-CMCAcute
J A S O
TMDL Scenario
M
O
D
G-57
-------
Model Configuration and Results
30
li20
Q.
O
I 10
o
14
12 -
10 -
Q.
6 -
4 -
2 -
0
Chl-a-TMDL Scenario
Chl-a Criteria
pH-TMDL Scenario - -pH Criteria 2 - -pH Criteria 1
JFMAMJJASOND
JFMAMJJASOND
KSDTR continued
G-58
-------
Model Configuration and Results
90 -,
18 -
1R
O) 1A
d 19
a ^z ~
O)
51 10
O
"S a
>
o fi
0)
S 4
0
n -
i
90
18 -
1fi
^) 14
c 19
W l2 "
cs
51 m
O
T3 o
>
O c
(0 D -
(0
Q 4
2 -
n
90 -,
18 -
1R
O) 14
c 19
5 l2 "
cs
51 m
O
T3 p
>
0 fi
CO
b 4
9
n
i
KSD97
DO TMDI ^rpnarin _^ - -.DO Pritorin
ZL Av iM
r V X^ "K 7^
;v;;;vJ^\; ^s^ ::
\ /S ^^^^yW|
S^^^^ \jj
IFMAMJ JASOND
DO- TMDL Scenario SOday DO Criteria
^\^__^ ,y\
-. ,-^"
JFMAMJ JASOND
DO- TMDL Scenario 7day DO Criteria
-T-\ r\ r
\^J \-^ ~~r\~
\ /
^-^__\ -s"^'
\ /\ /
\^/ \ s^"-^
IFMAMJ JASOND
G-59
-------
Model Configuration and Results
0.4
0.35 -
0.3 -
0.25 -
-ง- 0.2 -
I
z 0.15 -
0.1 -
0.05 -
0.45
3
CL
0.4 -
0.35 -
0.3 -
0.25 -
0.15 -
0.1 -
0.05 -
M
M
PO4-TMDL Scenario
^VM
M
M
J F M A M
KSD97 continued
ON
O N D
O N D
G-60
-------
Model Configuration and Results
30dayAvg-TMDL Scenario
. CCC if Early Stage Fish Present
0.0
J F
KSD97 continued
M
M
O
Highest 4-day avg in 30 days-TMDL Scenario
... -2.5xCCCEarly Stage Fish Present
MAM
- -CMCAcute
J A S O
TMDL Scenario
M
O
D
G-61
-------
Model Configuration and Results
20
15 - -
-------
Model Configuration and Results
90 -,
18 -
1R
O) 1A
d 19
a 1z ~
O)
51 10
O
"S a
>
o fi
0)
S 4
0
n -
i
90
18 -
1fi
^) 14
c 19
o fi
Q 4
2 -
n
90 -,
18 -
1R
O) 14
c 19
5 l2 "
cs
51 m
O
T3 p
>
0 fi
b 4
9
n
i
KSDM
DO TMDI ^rpnarin _^ - -.DO Pritorin
AL A, A^
A "^ f ^ n " ft
r ox v^ KA Mr
V i h, j, A / u
W^\ /^CT^
i /S ^r
\J M^H, /S^^*-rf*f
V %ff!5h(*^VlX^iAJ
IFMAMJ JASOND
DO- TMDL Scenario SOday DO Criteria
"^v^^^^ x^
"<^x^ ^^
^V^^^- ^^^1.
JFMAMJ JASOND
DO- TMDL Scenario 7day DO Criteria
/\ f\ r
\_^J \J~ ~/^J
\ /^
v/\__ __^_ -^^^
^^^
JFMAMJ JASOND
G-63
-------
Model Configuration and Results
0.45
3
CL
0.4 -
0.35
0.3 -
0.25 -
0.15 -
0.1 -
0.05 -
2
1.75 -
1.5 -
1.25 -
1 -
0.75 -
0.5 -
0.25 -
0
M
M
OND
PO4-TMDL Scenario
M
M
NOX-TMDL Scenario
J F M A M
KSDM continued
ON
O N D
G-64
-------
Model Configuration and Results
30dayAvg-TMDL Scenario
. CCC if Early Stage Fish Present
0.0
J F
KSDM continued
M
M
O
D
Highest 4-day avg in 30 days-TMDL Scenario
... -2.5xCCCEarly Stage Fish Present
MAM
. -CMCAcute
J A S O
TMDL Scenario
M
O
G-65
-------
Model Configuration and Results
20
15 - -
------- |