HomeMy WebLinkAboutDWQ-2024-004536Project Orientation & Technical Support Document
Overview
Science Panel Meeting | May 18, 2023
Note: all information and data presented are considered draft, in-process
material
ULWQS Ongoing and Upcoming Activities
•Technical Support Document
▪Theme: link management goals to response conditions to nutrients
▪Reference analysis, stressor-response analysis, literature evidence
•SP NNC Recommendation
▪Theme: translate TSD into protective nutrient ranges, with uncertainty & report to SC
•SC Charge Questions Responses
▪Theme: describe the complex limnology of Utah Lake
▪Draft 1 completed previously ð look toward updating responses as new studies are
completed
•Implementation Planning
▪Theme: Determine how NNC will be achieved with the “levers” available
▪Takes into account source partitioning, cost & feasibility, timelines
Note: all information and data presented are considered draft, in-process
material
Available Lines of Evidence
Lines of evidence will be used differently across portions of the project
Note: all information and data presented are considered draft, in-process
material
Technical
Support
Document:
Reference
Technical
Support
Document: S-R
Analysis
Technical
Support
Document:
Literature
SC Charge
Questions
Responses
Implementation
Planning
ULWQS Studies
Empirical S-R
Modeling
Watershed
mechanistic model
Lake mechanistic
model
Literature
Pathway to Criteria (ULWQS Technical
Framework)
•SP recommendation ð SC recommendation ð Utah Lake Authority
endorsement
•Regulatory process ð multiple other groups involved with the process
▪DWQ Water Quality Standards workgroup
▪Legislature
▪EPA
Note: all information and data presented are considered draft, in-process
material
•Provide the technical basis for the
development of numeric nutrient criteria
(NNC) to protect designated uses
•Recreation
•Aquatic Life
•Others (Agriculture, Downstream)
•Conduct analyses to support multiple lines
of evidence in the NNC framework
Purpose of the Technical
Support Document
Note: all information and data presented are considered draft, in-process
material
Lines of Evidence
1.Reference-based
▪Results from paleolimnological studies
▪Utah Lake Nutrient Model prediction/extrapolation of reference conditions
2.Stressor-response analysis
▪Utah Lake Nutrient Model output
▪Statistical models
3.Scientific literature
▪Scientific studies of comparable/related lake ecosystems
▪Support/supplement other lines of evidence
Note: all information and data presented are considered draft, in-process
material
Weight of Evidence
•Ranges of nutrients deemed
protective of uses across lines
of evidence
•How to distill these lines of
evidence into a
recommendation?
▪Statistical distributions of endpoints
▪Interpret endpoints in the context of
their uncertainty ð weigh lines
against each other by their
relevance, strength, and reliability
Note: all information and data presented are considered draft, in-process
material
Questions and Discussion
Note: all information and data presented are considered draft, in-process
material
Stressor-Response Analysis Updates
Science Panel Meeting | May 18, 2023
Note: all information and data presented are considered draft, in-process
material
•Discuss details of more complex decisions
o Period of analysis
o Lake Regions
o Phytoplankton
o Zooplankton
•Present initial draft of S-R models
•Discuss Tt recommendations for S-R next
steps
Agenda
Note: all information and data presented are considered draft, in-process
material
Period of Analysis
•Recap: tradeoffs for using all available data vs. recent data
▪All available data: larger sample sizes and potentially larger gradient
▪Recent data: consistent analytical methods, more intensive sampling
▪Only relevant for constituents collected over a long period of time
•Goal for S-R analysis is not to characterize recent condition of
the lake but rather how the lake behaves
Note: all information and data presented are considered draft, in-process
material
Period of Analysis
•History of TP analytical methods
▪Pre-2000: method details unknown
▪2000: change to EPA 365.1, hot plate digestion,
Alchem instrument, MRL = 0.020 mg/L
▪2009: change to Lachat instrument
▪2011: change to autoclave digestion, MRL = 0.003
mg/L (2 calibration curves)
▪2016: J flag created to report values between MDL
and MRL
•Other constituents have had more consistent
methods over time, and we have fewer
issues with samples < detection limit in Utah
Lake
Note: all information and data presented are considered draft, in-process
material
Period of Analysis
•Distributions pre- and
post-2015 are not
substantially different
•Very few TP non-
detects in either period
(n = 3)
•TN was not sampled
extensively prior to
2015
Note: all information and data presented are considered draft, in-process
material
Period of Analysis
Exploratory
visualization of
bivariate relationships:
Note: all information and data presented are considered draft, in-process
material
•Tt recommendation: include entire period of
record
•SP discussion and support/non-support for this
recommendation?
Decision Point: Period of Analysis
Note: all information and data presented are considered draft, in-process
material
Aggregation Considerations
•Utah Lake assessed according to the
state’s CALM
▪Uses must be met in all sampled locations
▪Suggest analyzing S-R relationships at the
site scale ð identify most sensitive areas
of the lake
▪Could aggregate up to a waterbody-scale,
but would need to demonstrate the lake-
wide condition is protective of conditions
at every site. More complicated approach
Considering Multiple Regions
•Utah Lake currently has two
Assessment Units: main basin and Provo
Bay
•May consider Goshen Bay as a third
region by demonstrating unique
hydrology
•Suggested approach:
▪Define boundary between Goshen Bay/Main
basin by TDS differences
▪Run S-R analyses with the three regions,
and see if the relationships differ
▪If relationships do not differ ð combine and
set a single threshold
▪If relationships do differ ð analyze each
region separately and set multiple
thresholds
Lake Regions: Goshen Bay
•Sampling locations suggest a logical “break” to define Goshen Bay
•We don’t observe a difference in TDS in overall range (may not hold day-
to-day)
•Suggest keeping Goshen Bay included with main basin
Note: all information and data presented are considered draft, in-process
material
•Tt recommendations:
o Run analysis at the site scale
o Consider Goshen Bay a part of the main basin
•SP discussion and support/non-support for
these recommendations?
Decision Points: Lake Regions
Note: all information and data presented are considered draft, in-process
material
Zooplankton
•Intent of S-R relationship: quantify the level of primary production
associated with zooplankton community that supports aquatic life use
•Follow-up since last subgroup meeting:
▪Discussion of zooplankton national model with Lester Yuan (EPA)
▪Review of David Richard’s MIBI provisional report, follow-up email to Theron
▪Discussion of Zhang et al. 2022 with Mike Brett
▪Follow-up with June Sucker Recovery Program on zooplankton conditions that
support fish restoration efforts
Note: all information and data presented are considered draft, in-process
material
Zooplankton National Model (USEPA 2021,
2022)
•Metric of interest: relationship between
total zooplankton biomass and chlorophyll
▪Relationship has a steeper slope at low
chlorophyll
▪Trophic transfer becomes less efficient at
higher chlorophyll and slope “levels off”
▪Choose a slope threshold that protects against
decoupling
•Spoke with Lester Yuan about applying
this model for Utah Lake
•Could run the model at multiple slope
thresholds and generate a range of
chlorophyll targets
Note: all information and data presented are considered draft, in-process
material
Richards MIBI
•38 zooplankton metrics
suggested
•Diversity and biomass
suggested as primary
indicators, but rationale
is unclear
•Would need to identify
thresholds for metric(s)
chosen that represent
zooplankton conditions
supportive of aquatic
life use
Note: all information and data presented are considered draft, in-process
material
Zhang et al. 2022
•Zooplankton production:biomass ratio
(P/B)
▪Physiological control during cold periods
▪Resource control during warm periods
▪Underpinning: food web energy transfer
•Would need to convert zooplankton
counts to biomass via literature-derived
relationships
•Can compare P/B ratios in Utah Lake
against other lakes in the study to inform
thresholds
Note: all information and data presented are considered draft, in-process
material
Landom et al. June Sucker Report
•Larger-bodied Cladocera
(Daphnia, also Ceriodaphnia) are
indicators of zooplankton
community supportive of a
healthy fish community
•Could use cladoceran relative
biomass as a metric of interest
•Zooplankton community also
correlated with carp biomass –
potential confounding factor of
top-down controls
Note: all information and data presented are considered draft, in-process
material
Zooplankton
•Potential metrics and lines of evidence
▪Slope of zooplankton vs. phytoplankton biomass (EPA national model)
▪Metrics identified in Richards MIBI – which one(s), thresholds?
▪Zooplankton P/B (Zhang et al. 2022)
▪Cladoceran relative abundance (June Sucker Report)
•Link metric(s) to chlorophyll concentrations
▪Strong relationship could indicate support for bottom-up controls linked to nutrients
▪Weak relationship could indicate support for top-down controls or other
confounding factors
Note: all information and data presented are considered draft, in-process
material
•What are the metrics of interest?
•How might we identify thresholds for the
metrics of interest?
•How do we weigh the value of zooplankton
analysis against other lines of evidence?
Decision Points: Zooplankton
Note: all information and data presented are considered draft, in-process
material
Phytoplankton
•Spatial components of sampling regime
▪Two depths: surface “scum” and surface composite (suggest
changing “scum” to surface grab for future iterations)
▪Sampling locations: open water, beach, and marina
•How to interpret spatial differences?
▪Assessment methodology: uses must be met in all areas of the lake ð
cannot ignore beach and marina sites
▪Should scum samples be interpreted the same way as composite
samples?
▪Suggest a “translator” approach for open water vs. nearshore
samples:
–To support the use in beaches/marinas, what must the open water
condition be?
–Can run other S-R models for open water sites to translate to nutrient
concentrationsNote: all information and data presented are considered draft, in-process
material
Phytoplankton: Cyanobacteria Cell Count
•“Scum” samples generally higher
than composite samples
•Distributions similar across site
types for composite samples
•Open water generally lower than
beach/marina for scum samples,
but smaller sample size
Note: all information and data presented are considered draft, in-process
material
Phytoplankton: Cyanobacteria Biovolume
•“Scum” samples generally higher
than composite samples
•Distributions similar across site
types for composite samples
•Open water generally lower than
beach/marina for scum samples
Note: all information and data presented are considered draft, in-process
material
Phytoplankton: Microcystin
•“Scum” samples generally higher
than composite samples
•Marina samples tend to have
higher toxin concentrations than
other site types
Note: all information and data presented are considered draft, in-process
material
Phytoplankton: Cyanobacteria Cell Count
•When paired samples are
available, scum sample always
exceeds composite sample (not
surprising)
•Samples almost exclusively
restricted to beach and marina
samples
Note: all information and data presented are considered draft, in-process
material
Phytoplankton: Cyanobacteria Biovolume
•When paired samples are
available, scum sample always
exceeds composite sample in
beach and marina samples (not
surprising)
•In main basin open water sites,
relationship straddles the 1:1 line
Note: all information and data presented are considered draft, in-process
material
Phytoplankton: Microcystin
•When paired samples are
available, scum sample always
exceeds composite sample (not
surprising)
•Samples almost exclusively
restricted to beach and marina
samples
Note: all information and data presented are considered draft, in-process
material
Open Water vs. Nearshore Translator
•To protect conditions in the nearshore, what
do open water conditions need to be?
•For cyano biovolume and cell count at
beaches and marinas, we don’t observe a
systematic bias between site types when
samples are matched in time
•For microcystin, beach and marina samples
tend to be higher than open water sites
•Recommendation: if translator shows a
systematic bias, use this to adjust open
water targets to protect nearshore. If no
systematic bias, treat sites as consistent
with one another
Note: all information and data presented are considered draft, in-process
material
•Should composite and scum samples be
treated similarly? If not, how should they be
handled?
•Tt recommends developing an open water
to nearshore translator. SP discussion and
support/non-support for this
recommendation?
Decision Points: Phytoplankton
Note: all information and data presented are considered draft, in-process
material
Stressor-Response Analysis
•Intent: establish the statistical relationship between stressors and
responses
▪Responses linked to management goals, designated use support
▪What stressor level would relate to a response level that supports the use?
▪Incorporate estimates of statistical uncertainty to provide protectiveness
▪Confounding factors can be incorporated as covariates or contribute to statistical
uncertainty
•S-R analyses pertain to water column conditions
▪Relationships chosen based on causal linkages
▪Relationships independent of e.g., nutrient sources
Note: all information and data presented are considered draft, in-process
material
General S-R Approach
•Dataset filtered for April-September, surface samples
•Dataset currently includes no time period restrictions, scum samples
included
•Each site-date observation is a single sample in the analysis
•Tested whether a linear regression was the appropriate fit
▪If analysis fit assumptions of linear regression, recommend proceeding
▪If analysis does not fit assumptions, recommend a different approach (e.g., quantile)
•Tested statistical significance:
▪Stressor variable
▪Main basin vs. Provo Bay slope
▪Main basin vs. Provo Bay intercept
Note: all information and data presented are considered draft, in-process
material
30-Day Mean DO vs. Chlorophyll
•Utah Lake rarely observed
exceedances, except in Provo Bay
•Chlorophyll was not a significant
predictor of 30-day mean DO
•Recommendation: this S-R model
is not a promising indicator for
generating chlorophyll target
Note: all information and data presented are considered draft, in-process
material
7-day mean DO vs. Chlorophyll
Note: all information and data presented are considered draft, in-process
material
•Utah Lake rarely observed
exceedances, except in Provo Bay
•Provo Bay had a negative
relationship, but sample size was
too small
•Recommendation: this S-R model
is not a promising indicator for
generating chlorophyll target
1-day minimum DO vs. Chlorophyll
Note: all information and data presented are considered draft, in-process
material
•Exceedances observed in Provo
Bay
•Significant negative relationship
•Provo Bay had a significantly
lower intercept but not slope
•Recommendation: generate
chlorophyll targets using this
linear regression, consider
separate intercept term for Provo
Bay
Daily max pH vs. Chlorophyll
Note: all information and data presented are considered draft, in-process
material
•Positive relationship observed in
main basin, negative relationship
observed in Provo Bay (but, n = 6)
•Overall R2 = 0.52
•Recommendation: generate
chlorophyll targets using this
linear regression, either for just
main basin or combine Provo Bay
given small sample size
Microcystin vs. Cyano. Biovolume
Note: all information and data presented are considered draft, in-process
material
•Microcystin significantly
positively related to
cyanobacteria biovolume
•Provo Bay did not have a
significantly different slope or
intercept
•Overall R2 = 0.26
•Recommendation: generate
cyano. biovolume targets using
this linear regression, combine
main basin and Provo Bay
Microcystin vs. Cyano. Cell Count
Note: all information and data presented are considered draft, in-process
material
•Microcystin significantly
positively related to
cyanobacteria cell count
•Provo Bay did not have a
significantly different slope or
intercept
•Overall R2 = 0.25
•Recommendation: generate
cyano. cell targets using this linear
regression and compare to
existing cell count threshold,
combine main basin and Provo
Bay
Microcystin vs. Chlorophyll
Note: all information and data presented are considered draft, in-process
material
•Microcystin not significantly
related to chlorophyll
•More limited sample size than
cyano metrics (n = 37), linkage is
less direct than for cyano. metrics
•Likely want to compare this
output to the EPA national model
for microcystin. Are these models
consistent?
•Recommendation: this S-R model
is not a promising indicator for
generating chlorophyll target
Cyano. Biovolume vs. Chlorophyll
Note: all information and data presented are considered draft, in-process
material
•Significantly positive relationship
•Provo Bay did not have a significantly different slope or intercept
•Overall R2 = 0.25
•Recommendation: generate chlorophyll targets using this linear regression and biovolume targets from microcystin model, combine main basin and Provo Bay
Cyanobacteria Cell Count vs. Chlorophyll
Note: all information and data presented are considered draft, in-process
material
•Significantly positive relationship
•Provo Bay had a significantly
different intercept but not slope
•Overall R2 = 0.30
•Recommendation: generate
chlorophyll targets using this
linear regression, use different
intercept for Provo Bay
Cyano. Relative Biovolume vs. Chlorophyll
Note: all information and data presented are considered draft, in-process
material
•Very noisy relationship
•Y axis is 0-1, not continuous ð
linear regression not the best
approach, likely want something
like a logistic regression
•But, a different statistical
approach would likely not
illuminate a pattern. Likely other
drivers of cyano relative
abundance not captured here
•Recommendation: this S-R model
is not a promising indicator for
generating chlorophyll target
Secchi Depth vs. Chl, TN, TP
Note: all information and data presented are considered draft, in-process
material
•Significantly negative
relationship, but noisy (R2 = 0.05-
0.13)
▪Chlorophyll pictured here
▪TN and TP have similar relationships
•Likely have confounding influence
of suspended sediment
•Do we have a Secchi depth
threshold in mind that would
support aquatic life & recreation?
•Recommendation: discussion
needed
Cyano. Biovolume vs. TN
Note: all information and data presented are considered draft, in-process
material
•Relationship not significant
•Recommendation: this S-R model
is not a promising indicator for
generating a TN target
Cyano. Biovolume vs. TP
Note: all information and data presented are considered draft, in-process
material
•Significant positive relationship
•Provo Bay did not have a significantly
different slope or intercept
•N:P ratio may help explain variability
•Quantile regression may be considered
given wedge-shaped relationship
•Recommendation: Use this S-R relationship
to derive TP target, combine main basin
and Provo Bay, potentially switch to
quantile regression
Cyano. Cell Count vs. TN
Note: all information and data presented are considered draft, in-process
material
•Significant positive relationship,
but noisy (R2 = 0.03)
•Provo Bay did not have a
significantly different slope or
intercept
•Quantile regression may be
considered given wedge-shaped
relationship
•Recommendation: Use this S-R
relationship to derive TN target
but acknowledge uncertainty,
combine main basin and Provo
Bay, potentially switch to quantile
regression
Cyano. Cell Count vs. TP
Note: all information and data presented are considered draft, in-process
material
•Significant positive relationship, but
noisy (R2 = 0.11)
•Provo Bay had a significantly different
intercept but not slope
•Quantile regression may be considered
given wedge-shaped relationship
•Recommendation: Use this S-R
relationship to derive TP target but
acknowledge uncertainty, use different
intercept for Provo Bay, potentially
switch to quantile regression
Chlorophyll vs. TN
Note: all information and data presented are considered draft, in-process
material
•Significant positive relationship
•Provo Bay had a significantly different slope
and intercept
•Quantile regression may be considered
given wedge-shaped relationship
•N:P ratio may help explain variability
•Recommendation: Use this S-R relationship
to derive TN target, use different slope and
intercept for Provo Bay, use quantile
regression
Chlorophyll vs. TP
Note: all information and data presented are considered draft, in-process
material
•Significant positive relationship
•Provo Bay did not have a significantly
different slope or intercept
•Quantile regression may be considered
given wedge-shaped relationship
•N:P ratio may help explain variability
•Recommendation: Use this S-R relationship
to derive TP target, combine main basin
and Provo Bay, use quantile regression
Outstanding S-R analysis
•National Models
▪Microcystin: compare distributions in Utah Lake with distributions in national
model. If consistent, can run national model to generate potential chlorophyll
targets
▪Zooplankton: run for a range of input settings to generate potential
chlorophyll targets
▪TN: compare observations in Utah Lake with national relationship
▪TP: compare observations in Utah Lake with national relationship
•Lake Visitation
•Public Perception
Note: all information and data presented are considered draft, in-process
material
•Choose confidence/credible intervals and
prediction intervals/quantiles to represent
quantification of uncertainty and
protectiveness
General S-R Guidance Needed for Next Steps
Note: all information and data presented are considered draft, in-process
material
Questions and Discussion
Note: all information and data presented are considered draft, in-process
material
Reference Conditions Analysis
Science Panel Meeting | May 19, 2023
Note: all information and data presented are considered draft, in-process
material
•Provide the technical basis for the
development of numeric nutrient criteria
(NNC) to protect designated uses
•Conduct analyses to support multiple lines
of evidence in the NNC framework
Purpose of the Technical
Support Document
Note: all information and data presented are considered draft, in-process
material
Lines of Evidence
1.Reference-based
▪Results from paleolimnological studies
▪Utah Lake Nutrient Model prediction/extrapolation of reference conditions
2.Stressor-response analysis
▪Utah Lake Nutrient Model output
▪Statistical models
3.Scientific literature
▪Scientific studies of comparable/related lake ecosystems
▪Support/supplement other lines of evidence
Note: all information and data presented are considered draft, in-process
material
Reference-Based Analysis
•Typical approach elsewhere:
▪Identify comparable systems in locations with minimal human impact (e.g., land use,
shoreline)
▪Example application: Herlihy and Sobota 2013 for NLA lakes
▪Can set thresholds based on an upper percentile of reference condition
Note: all information and data presented are considered draft, in-process
material
Reference-Based Analysis
•BUT, Utah Lake has few, if any, comparable
systems
•Look to other methods to evaluate applicable
reference conditions
▪What did Utah Lake look like pre-EuroAmerican
settlement?
▪What would Utah Lake look like if the watershed was
minimally impacted by humans?
Note: all information and data presented are considered draft, in-process
material
Reference-Based Analysis
•Intended to set a “floor” & add context for how the lake has changed over
time
•Paleolimnological reconstruction of past conditions
▪Quantify pre-settlement nutrient conditions and how they have changed over time
▪Multiple studies, SC charge question responses
•Model-based prediction
▪Watershed model run under a “reference conditions” scenario ð watershed nutrient
loading
▪Watershed conditions then used as boundary conditions for the lake model
Note: all information and data presented are considered draft, in-process
material
Paleolimnological Studies
•ULWQS paleo study (PI Brahney, including King 2019, Devey 2021)
•Other recent theses/dissertations
▪Macharia 2012
▪Tate 2019
▪Williams 2021
•Historical studies
▪Brimhall 1972
▪Bolland 1974
▪Sonerholm 1974
▪Brotherson 1981
▪Javakul et al. 1983
Note: all information and data presented are considered draft, in-process
material
Using paleo evidence in the TSD
•Lines of evidence span qualitative, semi-quantitative, quantitative
▪Qualitative: community composition
▪Semi-quantitative: oligo/meso/eutrophic conditions
▪Quantitative: sediment nutrient concentrations, isotope values
•Some cores have reliable dating, some do not
•Combine paleo lines of evidence to form holistic picture of reference
conditions
Note: all information and data presented are considered draft, in-process
material
Benthic Diatoms (Bolland 1974, Jakuval et al. 1983, Brahney et al.
2021)
•Benthic & epiphytic diatoms dominated in pre-industrial conditions
•Increasing prevalence of planktonic diatoms approaching present day
•Supports evidence for clearer water conditions pre-settlement
Note: all information and data presented are considered draft, in-process
material
Bolland 1974
Planktonic Diatom Community
•Conditions shift from oligo-mesotrophic to eutrophic diatom taxa from
deep to shallow core (Bolland 1974)
•Pollution-tolerant species increase moving up-core (Brahney et al. 2021)
•Supports evidence for a shift from oligo-mesotrophic to eutrophic
conditions
Note: all information and data presented are considered draft, in-process
material
Bolland 1974
Phytopigments (King 2019)
•Increase in diatom production
around 1890
•Transition from diatoms to
cyanobacteria and green algae
production in recent decades
•Support evidence for a shift from
oligo-mesotrophic conditions to
eutrophic conditions
Note: all information and data presented are considered draft, in-process
material
King 2019
Phytopigments (King 2019)
•Greater chlorophyll a degradation
rates post-1890
•Highest chlorophyll a concentrations
near the surface (subject to
degradation)
•Decreases in production around 1950,
consistent with timeline of
wastewater treatment plants in 1950s
•Support evidence for a shift from
oligo-mesotrophic conditions to
eutrophic conditions
Note: all information and data presented are considered draft, in-process
material
King 2019
eDNA (King 2019)
•eDNA records show an increasing
abundance of cyanobacteria post-
1900 and representation of
hardstem bulrush pre-1900
•Support evidence for a shift from
oligo-mesotrophic conditions to
eutrophic conditions
Note: all information and data presented are considered draft, in-process
material
King 2019
Other taxa
•Macrophytes
▪Paleo evidence points to a historical macrophyte-dominated clear-water state
▪Present-day macrophyte restoration dependent on several factors, not just nutrient
regime (carp, non-algal turbidity, algal production, hysteresis)
▪King et al. 2023 found a self-stabilizing submerged macrophyte community would
require water clarity consistent with chlorophyll < 18 µg/L and Secchi depth ~ 1 m
•Zooplankton
▪Increase in cladocerans post-1890
▪Decrease in ostracods post-1890
Note: all information and data presented are considered draft, in-process
material
Nutrient Concentrations in Sediments
•P concentrations increase in
cores from 20 cm to present
(Brimhall 1972, Bolland 1974)
•Exchangeable and Ca-bound P
increase through time – fractions
indicative of water column P
(Devey 2021)
•Fe-bound P and Al-bound P
relatively stable through time
(Devey 2021)
•Support a shift to more eutrophic
conditions from preindustrial
times to present
Note: all information and data presented are considered draft, in-process
material
Devey
2021
Nutrient isotopes
•δ15N isotope values increase through time post-
1900 (Macharia 2012, King 2019)
•δ15N positively correlated with %N in sediment
bulk organic matter (R2 = 0.79, Macharia 2012)
•Suggest greater prevalence of wastewater-
derived N to present (consistent with pop.
growth)
•Wastewater delivery and treatment timeline:
▪Provo sewers were built in 1908, treatment plant in 1956
▪Orem sewers were built in 1945, treatment plant in 1959
Note: all information and data presented are considered draft, in-process
material
Goshen Bay core, King
2019
Inferring water column nutrients from
sediments
•Sediments store integrated information about system conditions
•Sediment nutrient concentrations are proportional to water column
concentrations, but not a direct 1:1 relationship
•Converting sediment concentrations to water column concentrations
could be difficult in Utah Lake
▪Alkalinity, P speciation, equilibration, water levels, etc.
▪Many calibration efforts needed to arrive at a reliable water column conversion
▪An approach more likely to be successful would be to calculate relative differences
over time
•Subgroup may choose to pursue a mass balance approach (use Mike
Brett’s approach or similar?)
Note: all information and data presented are considered draft, in-process
material
Limitations
•Additional changes have co-occurred alongside nutrient regime changes
▪Carp introduction
▪Hydrologic changes
▪Climate change
•Emphasizes the need to evaluate lines of evidence holistically
▪Do paleo lines of evidence agree (direction, magnitude)?
▪Reference scenario for mechanistic model will mimic present-day, minimally disturbed
condition
Note: all information and data presented are considered draft, in-process
material
•Paleo subgroup planning to meet to discuss
ULWQS paleo study
•June SP meeting: get into analysis specifics,
synthesizing info sources
Plans for Finalizing Paleo Analysis of
Reference Conditions
Note: all information and data presented are considered draft, in-process
material
Questions and Discussion
Watershed Model Reference Scenario
•“Reference” could be interpreted in different ways
▪Remove human-influenced land cover
▪Remove human-influenced hydrologic changes
▪Remove all traces of human impact
▪Extreme example: simulate Lake Bonneville! ð need to draw the line somewhere
•What is the goal of this scenario?
▪Utah Lake doesn’t have comparable systems to consider as reference
▪Develop simulation of comparable system with minimal anthropogenic nutrient inputs
▪What would Utah Lake look like under minimal human influence?
▪Recall: serves as a “floor,” not as a realistic expectation
Note: all information and data presented are considered draft, in-process
material
Proposed Approach
1.Run watershed model in “reference mode”
2.Extract watershed nutrient loading & hydrology to be passed to lake
model
3.Run lake model with reference watershed inputs (i.e., new boundary
condition)
4.Extract water column nutrient and primary production data
Note: all information and data presented are considered draft, in-process
material
Proposed Watershed Model Reference Scenario
•Changed elements
▪Land cover based on pre-EuroAmerican settlement vegetation from LANDFIRE, accounting for natural wildfires
▪Water withdrawals (i.e., for irrigation, public water supply) and releases removed
▪Irrigation removed (no agricultural land)
▪Point source discharges removed (no developed land)
▪Septic systems removed (no developed land)
•Maintained elements
▪Dams (e.g., Deer Creek Reservoir, Jordan River outflow) and stream hydraulics (i.e., channel geometry)
▪Subbasin boundaries, drainage divides, and stream routing
▪Weather conditions
▪Calibrated parameter values for natural land covers (e.g., forest)Note: all information and data presented are considered draft, in-process
material
Questions and Discussion
Utah Lake WQ Model Update
May 19, 2023
Topics
•Background on Individual Models
▪EFDC
▪SWAN
▪WASP
•How the Models Work Together
•Model Hydrodynamic Performance
•Model Water Quality
•Hydrodynamic model
•EFDC solves the equations of mass and momentum transport
•EFDC is a 2-D/3-D orthogonal curvilinear grid
•EFDC also provides solutions for salinity, temperature, and conservative
tracers with full density feedback to handle stratified conditions
Environmental Fluid Dynamics Model (EFDC)
Hydrodynamics
Dynamics
(E, u, v, w, mixing)Dye Temperature Salinity Near Field
Plume Drifter
•Wave Model
•Wind-generated waves in water bodies (coastal areas, lakes).
Simulating WAves Nearshores (SWAN)
Wind
Wind
z
x
Periphyton Biomass
D : C : N : P : Chl
IP
IN
Phytoplankton Biomass
Group 3
D : C : N : P : Si: Chl DOGroup 2
D : C : N : P : Si: ChlGroup 1
D : C : N : P : Si : Chl
TIC
H2CO3 –HCO3-–CO32-
Total
Alkalinity
Particulate Detrital OM
SiPNCD
Dissolved OM
Si
P
N
CBOD1
CBOD2
CBOD3
Inorganic Nutrients
NO3PO4SiO2NH4
pH
atmosphere
uptake
excretion
Inorganic Solids
S3S1S2
ox
i
d
a
t
i
o
n
ox
i
d
a
t
i
o
n
ni
t
r
i
f
i
c
a
t
i
o
n
photosynthesis and respiration
death
dissolution
mineralization
sorption
Water Quality Analysis Simulation Program (WASP)
EFDC
SWAN
•Grid cell volumes
•Velocities
•Temperatures
•ISS WASP
Hydrodynamics Hydro Linkage
Water Quality
Outputs
•Grid cell volumes
•Velocities
•Temperatures
•ISS
•Shear stress
Outputs
•Nutrient concentrations
•Algae biomass
•BOD
•Dissolved oxygen
Modeling Framework (How Models Work Together)
Linkage Between EFDC-SWAN
Model Setup Overview
Computational grid
Model Bathymetry
Meteorological Data
•NCDC NOAA WBAN 24174 (Provo
Municipal Airport, UT) was used as a
primary station for:
▪Air Temperature
▪Relative Humidity
▪Altimeter Pressure
▪Wind Speed and Direction
•KPVU (MesoWest), Buoy Stations used
as supplemental Stations
•PRISM data used for precipitation.
NASA’s NLDAS data was used for
Solar Radiation and NARR was used
for Cloud Cover.
11
12UTAH Lake Hydrodynamic and Water Quality Modeling
KPVU
UTLAK
Provo Airport
(KPUV)
Utah Lake Near
Mosida (UTLAK)
Wind
Buoy
4917390
Buoy
4917450
4917390
4917450
Hydrodynamic Calibration
WSE and Current Velocity
2019 ADCP Deployment
Significant Wave Height
WQX-4917390
Calibration Statistics WSE and Significant Wave
Height
Mean Median 5 %tile 95 %tile Mean Median 5 %tile 95 %tile
Hs 0.08 0.04 0.01 0.40 0.07 0.03 0.00 0.38 0.75 0.04 0.06 0.47 0.93
Mean Abs
Err RMS Err Norm RMS
Err
Index of
AgrmtParameterMeasuredSimulatedR2
Mean Median 5 %tile 95 %tile Mean Median 5 %tile 95 %tile
WSE -1.09 -1.10 -2.00 -0.14 -1.05 -1.07 -1.95 -0.11 0.99 0.05 0.07 0.03 1.00
Parameter Measured Simulated R2 Mean Abs
Err RMS Err Norm RMS
Err
Index of
Agrmt
Temperature
Temperature
Temperature
Temperature
Temperature
Temperature
Calibration Statistics Temperature
Number
Obs Mean Median 5 %tile 95 %tile Mean Median 5 %tile 95 %tile
WQX-4917520 WTEMP 42 16.94 19.36 4.52 25.58 18.03 19.36 6.01 25.64 0.96 1.31 1.71 0.07 0.98
WQX-4917310 WTEMP 43 17.35 19.41 4.22 24.69 17.60 18.59 5.76 25.65 0.95 1.33 1.59 0.07 0.98
WQX-4917320 WTEMP 33 17.16 18.94 3.98 26.26 17.16 18.59 5.60 25.14 0.97 1.26 1.40 0.06 0.99
WQX-4917370 WTEMP 41 16.73 18.13 5.32 25.06 17.78 18.95 5.79 25.54 0.97 1.15 1.50 0.07 0.99
WQX-4917390 WTEMP 43 16.75 17.75 5.77 24.54 17.70 18.66 5.55 25.79 0.96 1.26 1.59 0.07 0.98
WQX-4917770 WTEMP 42 17.33 18.70 6.31 24.91 17.48 18.36 6.03 24.68 0.93 1.24 1.65 0.07 0.98
WQX-4917450 WTEMP 36 17.26 18.67 6.03 24.16 16.92 17.32 6.26 24.33 0.92 1.27 1.69 0.08 0.98
WQX-4917500 WTEMP 42 16.95 18.48 5.50 25.10 17.50 18.68 5.98 24.65 0.97 1.06 1.32 0.06 0.99
WQX-4917600 WTEMP 36 16.30 17.78 4.72 25.70 16.79 17.28 6.08 24.70 0.95 1.35 1.54 0.07 0.98
RMS Err Norm RMS
Err
Index of
AgrmtStationParameter
Measured Simulated
R2 Mean Abs
Err
Draft Water Quality
Sediment Transport (inorganics)
•Sediment size classes (hydrometer
tests UDWQ)
•2 Cohesive
•1 Non-Cohesive
•Transport parameters incipient motion
•Critical Shear estimates (UDWQ)
[0.38 –2.81]Pa
•Wave Shear estimates (Utah Lake WQ
Study, 2020)
[~0.17] Pa
Clay Silt Sand
0.005 mm 0.062 mm 2 mm
South Lake 0-5 85.4 14.6 0.0
South Lake 5-15 79.4 20.6 0.0
South Lake 15-30 90.6 9.4 0.0
Provo buoy 0-5 58.2 41.8 0.0
Provo buoy 5-15 87.0 13.0 0.0
Provo Buoy 15-30 73.0 27.0 0.0
North Lake 0-5 73.0 27.0 0.0
North Lake 5-15 72.7 27.3 0.0
North Lake 15-30 77.1 22.9 0.0
Provo Bay 0-5 69.2 10.2 20.6
Provo Bay 5-15 48.3 13.5 38.2
Provo Bay 15-30 45.9 26.5 27.6
Goshen 0-5 82.2 17.8 0.0
Goshen 5-15 78.3 21.7 0.0
Goshen 15-30 77.1 22.9 0.0
Averaged 73.2 21.1
Calculated
Percent Content
Sediment Transport: Shear Stress
Sediment Transport: Shear Stress & TSS
TSS (draft)
TSS (draft)
TSS (draft)
TSS (draft)
TSS (draft)
Sediment Diagenesis and P-Binding
•Sediment CNP contents and nutrient fluxes
•Littoral Sediment Study (2022). CNP content & fluxes
•Hogsett et al. (2018). Fluxes
•P-Binding
•Partition coefficients from P-Binding study Kd [20 –130] L/Kg
Chlorophyll a (draft)
Chlorophyll a (draft)
Chlorophyll a (draft)
Chlorophyll a (draft)
Chlorophyll a (draft)
DISSOLVED OXYGEN
Dissolved Oxygen (draft)
Dissolved Oxygen (draft)
Dissolved Oxygen (draft)
Dissolved Oxygen (draft)
Dissolved Oxygen (draft)
TOTAL NITROGEN
Total Nitrogen (draft)
Total Nitrogen (draft)
Total Nitrogen (draft)
Total Nitrogen (draft)
Total Nitrogen (draft)
TOTAL PHOSPHORUS
Total Phosphorus (draft)
Total Phosphorus (draft)
Total Phosphorus (draft)
Total Phosphorus (draft)
Total Phosphorus (draft)
May 19, 2023
Maddie Keefer, Afshin Shabani, Cole Blasko, Jon
Butcher, Kevin Kratt (Tetra Tech)
Utah Lake Watershed Model Update
Note: all information and data presented are considered draft, in-process material
2
Agenda
Note: all information and data presented are considered draft, in-process material
Watershed Model Overview
Hydrology Calibration Recap
Water Balance Analysis
Water Quality Calibration Process/Progress Update
Watershed Model Overview
3Note: all information and data presented are considered draft, in-process material
Note: all information and data presented are considered draft, in-process material
Watershed Model Applications
•Quantifying nutrient load
contributions to Utah Lake
by source
•Simulating impact of
management actions (e.g.,
permit limits, BMPs) on
nutrient loading to Utah
Lake
•Evaluating alternative
watershed conditions (e.g.,
climate, land use)
4
Example TP relative source contribution chart (Bear
Creek watershed, Colorado)
Example reservoir chlorophyll‐a response to nutrient
load reductions (Bear Creek watershed, Colorado)
Note: all information and data presented are considered draft, in-process material
Watershed Model Applications
•Not currently set up to:
▪Evaluate other pollutants (e.g., bacteria, metals)
▪Simulate individual or field-scale BMPs
–BMPs must be simulated in aggregate
▪Identifying certain pollution issues (e.g., locations of failing septic
systems)
5
Watershed Model Selection
•Specific criteria defined related to watershed characteristics and
simulation capabilities, source representation, usability, and
general platform criteria
•Quantitatively ranked 11 modeling platforms
•Top rated: Hydrologic Simulation Program - FORTRAN
Modeling Process
Gather
Data
Build
Model
Calibrate
Model
Assess
Current
Conditions
Run
Scenarios
7Note: all information and data presented are considered draft, in-process material
8
Watershed Modeling QAPP
•Quality objectives for
measured and modeled
data
•Model framework to support
the project goals and
objectives
•Data collection and
acquisition to support model
build and calibration
•Specification of quality
assurance/quality control
(QA/QC) activities to assess
model performance
•Model usability assessment
Note: all information and data presented are considered draft, in-process material
9
Model Extent
Note: all information and data presented are considered draft, in-process material
10
Model Delineations
Note: all information and data presented are considered draft, in-process material
Started with
HUC12s and
modified
boundaries to
account for gaging
stations and water
diversions
11
Hydrologic Response Units
•Climate, geology, topography, and land use/cover influence runoff and
stream water quality; combined features to form model HRUs
•Land use: combined NLCD 2016, Utah’s water related land use coverage,
post-2016 fire perimeters
Note: all information and data presented are considered draft, in-process material
12
Hydrologic Response Units
•Impervious HRUs applied coverages for buildings and roads
Note: all information and data presented are considered draft, in-process material
13
Hydrologic Response Units
•USDA Soil Survey Geographic
Database (SSURGO)
▪Hydrologic Soil Group
–A/B: higher infiltration
–C/D: lower infiltration
•USGS 10-meter digital
elevation model (DEM)
Model Slope Percent of Total Area
Low (<10 degrees)41.22%
Medium (10‐30 degrees)43.58%
High (>30 degrees)15.20%
Note: all information and data presented are considered draft, in-process material
14
Irrigation
•Irrigation of agricultural lands and
lawns/landscapes represented
▪Reference evapotranspiration
(ETo) from Utah Climate Center
(e.g., AgWeather sites)
▪Crop coefficients to estimate
water demand for crop type
▪Irrigation demand = crop water
demand - precipitation
Note: all information and data presented are considered draft, in-process material
Note: all information and data presented are considered draft, in-process material 15
Diversions and releases
•Data obtained from
▪Central Utah Water
Conservancy District
▪Provo River Water Users
Association
▪Utah Division of Water
Rights
▪Annual distribution
system reports
•Diversion time series
represented as withdrawals
•Releases represented as
external water imports
16
Weather
•Hourly time series
•13 weather zones
•Derived from gridded
weather datasets
Variable Source
Precipitation PRISM
Potential
evapotranspiration
NLDAS
Air temperature NLDAS
Wind speed NLDAS
Solar radiation NLDAS
Dew point temperature NLDAS
Cloud cover NARR
High: 61 inches
Low: 11 inches
Note: all information and data presented are considered draft, in-process material
17
Permitted Point Source Discharges
Facility Type
Payson City WWTP Municipal
Salem City WRF Municipal
Salem City Corporation Municipal
Provo City Corporation Municipal
Mona City WWTP Municipal
Springville WWTP Municipal
Spanish Fork WWTP Municipal
Jordanelle WRF Municipal
Nephi Rubber Products Industrial
Ensign Bickford Company Industrial
Payson Power Plant Industrial
McWane Ductile – Utah Industrial
PacifiCorp Lake Side Power Plant Industrial
Midway Fish Hatchery Aquatic Animal Production Facility
Springville Fish Hatchery Aquatic Animal Production Facility
•Time series based on
facility data from:
▪Discharge
monitoring report
(DMR) data
▪Monthly operating
reports (MOR)
Note: all information and data presented are considered draft, in-process material
Hydrology Calibration Recap
18Note: all information and data presented are considered draft, in-process material
Note: all information and data presented are considered draft, in-process material
Hydrologic Calibration (complete)
•Calibrated to multiple endpoints to ensure robust model:
▪Remotely-sensed snow depth and water storage
▪Actual evapotranspiration
▪Daily, monthly, and cumulative gaged flow
•Guided by multiple visuals and statistical metrics related to:
▪Total flow
▪Seasonal/monthly flows
▪High/low flow distribution
▪Nash Sutcliffe efficiency (NSE) coefficients
•Daily streamflow records obtained from USGS monitoring sites
•Calibration sought to obtain the best overall fit at multiple locations, with
priority on larger tributaries to Utah Lake (Spanish Fork, Provo River)
Note: all information and data presented are considered draft, in-process material
Hydrologic Calibration Sites
20
Site
Total Flow
Relative
Error
Monthly
NSE
Provo River 10% 0.756
Spanish Fork 7% 0.814
American Fork 32% 0.737
Sixth Water Creek ‐2% 0.597
Salt Creek at Nephi ‐4% 0.994
Diamond Fork ‐4% 0.699
Currant Creek 42% 0.707
Hobble Creek 31%‐0.054
Dry Creek ‐14% 0.582
Summit Creek ‐56%‐0.017
Provo River (USGS Gage ID 10163000)
Water Balance Analysis
21Note: all information and data presented are considered draft, in-process material
Note: all information and data presented are considered draft, in-process material
•Streams in the Utah Lake watershed have significant interactions with deep groundwater ▪Recharge of deep aquifer occurs at foot of mountains, discharges back into streams near lake or direct to the lake
•Difficult to simulate in HSPF, which simulates only local shallow groundwater pathways
•Focused on this issue as one of the final steps of calibrating hydrology
•Compared output from HSPF to USGS groundwater modeling
22
Water Balance QA –
Groundwater/Surface Water Interactions
Note: all information and data presented are considered draft, in-process material
Conceptual Model of Groundwater Recharge and Discharge, Northern
Utah Lake (from Cederberg et al., 2009)
23
Water Balance QA – Consistency of HSPF Losses
and USGS Groundwater Balances
Note: all information and data presented are considered draft, in-process material
•Conclusion: HSPF model provides a close match to the USGS groundwater
studies – although the physical mechanisms represented in HSPF do not
fully match the USGS conceptual model
Approximate Annual Average (acre-feet/year) Discharge of
Groundwater in the Utah Lake Watershed
USGS Groundwater
Modeling Reports
HSPF Watershed Model
(v58)
205,000 to 340,000 323,957
Note: all information and data presented are considered draft, in-process material
•Comparison of the Utah Lake
inflows implied by the recorded
lake levels and water balance
compared to the flows predicted
by HSPF at a monthly time scale
•The overall water balance of Utah
Lake can be written as:
ΔV = Qin – Qout + GW - E + P + W
•Rearranged to solve for Qin:
Qin = ΔV + Qout – GW + E – P – W
25
Model Comparison to Utah Lake Water Balance
•EFDC model runs over calendar
years 2015 – 2020
▪Water balance analysis
suggests that Qin is
approximately 515,840
AF/yr
•HSPF model has almost
identical output of flow to the
lake at 515,751 AF/yr
Note: all information and data presented are considered draft, in-process material
Utah Lake Water Balance
26
Component AF/yr Notes
ΔV 15,590 ULDB_Stage_Vol_SA workbooks provided by Scott Daly, based on 1/1/05 elevation
of 1366.6 m and 12/31/20 elevation of 1367.4 m, resulting in net volumetric gain
of 1.15E8 m3 converted to AF/yr
Qout 137,785 From daily flows at Jordan River Narrows in Jordan_River_Outflow, provided by
Scott Daly. Likely has a small high bias due to groundwater inputs above Narrows.
GW 38,708 Direct‐to‐lake GW estimate developed for U. Utah model and adopted for Tt
model, derived from USGS groundwater reports. May have a high bias if USGS
estimates are from 1990s data.
E 588,013 Daily rate (inches) applied to timeseries of surface area from ULDB_Stage_vol_SA
workbook.
P 157,999 Daily rate (inches) applied to timeseries of surface area from ULDB_Stage_vol_SA
workbook.
WWTP 30,090 Timpanogos and Powell Slough WWTP discharges direct to lake as specified in
EFDC model
Qin = ΔV + Qout - GW + E – P - WWTP
27
Comparison of Monthly Inflows
Note: all information and data presented are considered draft, in-process material
There are some seasonal discrepancies, but these are likely mainly due to the lack
of seasonal specification of groundwater inflow and wastewater discharges,
coupled with the uncertainty in estimating evaporative losses
28
Comparison of Cumulative Inflow
Note: all information and data presented are considered draft, in-process material
Suggest that the current version of the watershed model is an unbiased estimate
of flow to Utah Lake
+ Progress Update
Water Quality Calibration Process
Note: all information and data presented are considered draft, in-process material
Note: all information and data presented are considered draft, in-process material
•Sediment calibration follows
the hydrologic calibration
and precedes nutrient
calibration
•Water temperature, DO, and
nutrient calibration follows
sediment
30
Approach for Watershed Model Calibration
Hydrology
Calibration
Sediment
Calibration
Nutrient
Calibration
Note: all information and data presented are considered draft, in-process material
•Inputs and observed sediment concentrations have various sources of uncertainty
•Calibrate to multiple endpoints to ensure robust model:▪Daily, monthly, and cumulative monitored sediment concentration▪Sediment Source Assessment▪Reach/Sediment Balance
•Objective is to represent the overall sediment behavior of the watershed, with knowledge of the morphological characteristics of the stream (i.e., aggrading or degrading behavior)▪Using sediment loading rates that are consistent with available values and providing a reasonable match with instream sediment data
31
Multi-Objective Sediment Calibration
Note: all information and data presented are considered draft, in-process material
Example of Debris Flows after 2016 Wildfires
Note: all information and data presented are considered draft, in-process material
Sediment Calibration
•Guided by multiple visuals and statistical metrics related to:
▪Sediment loading rates from the landscape
▪Seasonal/monthly TSS concentrations
▪High/low flow distribution
▪Average and median concentration and load error
•Observed TSS records obtained from Utah stream monitoring
sites
•Calibration seeks to obtain the best overall fit at multiple
locations, with priority on larger tributaries to Utah Lake (Spanish
Fork, Provo River)
33
Note: all information and data presented are considered draft, in-process material 34
Sediment/WQ Calibration Sites
Sediment calibration is underway and progressing well. Once sediment calibration is
complete and final statistics are calculated, water quality (e.g., water temperature,
nutrients) calibration will begin
Site TSS
Samples
Provo River at Murdock and
Olmstead 79
Provo River at Wildwood 44
Hobble Creek 132
Spanish Fork at Utah Lake 104
Spanish Fork at Moark 71
Diamond Fork 89
Starvation Creek/Upper Soldier
Creek 11
Thistle Creek 21
Currant Creek 46
Note: all information and data presented are considered draft, in-process material
Main objective is to obtain
acceptable agreement of observed
and simulated concentrations, while
maintaining the instream water
quality parameters within physically
realistic bounds and the nonpoint
loading rates within the expected
ranges from literature
35
Nutrient Calibration
Concentration Metrics
TN TP
Average
Observed
Concentration
(mg/L)
0.531 0.035
Average
Simulated
Concentration
(mg/L)
0.538 0.039
Relative Error 2.1% 12.8%
Example Nutrient Calibration Results - Bear Creek, Colorado
36
Steps in Nutrient Calibration
Note: all information and data presented are considered draft, in-process material
1.Estimate all model parameters, including land use specific accumulation and depletion/removal rates, washoff rates, and subsurface concentrations
2.Compare simulated nonpoint loading rates with expected range of nonpoint loadings from each land use and adjust loading parameters when necessary to improve agreement and consistency
3.Calibrate instream water temperature
4.Compare simulated and observed instream concentrations at each of the calibration stations
5.Analyze the results of comparisons in steps 3 and 4 to determine appropriate instream and/or nonpoint parameter adjustments, and repeat those steps as needed until calibration targets are achieved;
▪Watershed loadings are adjusted when the instream simulated and observed concentrations are not in full agreement
▪Instream parameters are adjusted within the range determined reasonable
37
Prospective Timeline
Note: all information and data presented are considered draft, in-process material
Sediment
Calibration
Complete
May 2023
Water Quality
Calibration
Complete
June 2023
Sensitivity
Analysis + Model
Documentation
June 2023
Questions and Discussion
Note: all information and data presented are considered draft, in-process material
Quantification of nutrient inputs, sediment storage and release,
and long-term recover in Utah Lake
Michael Brett
Professor of Limnology
Department of Civil & Environmental Engineering
University of Washington
https://health.utah.gov/enviroepi/appletree/HAB/Utah_Lake_2016_1.JPG
•Quantifying nutrient fluxes in Utah Lake is important
because this lake frequently has toxic cyanobacteria
blooms
•Cyanobacteria blooms are strongly associated with
high phosphorus and nitrogen concentrations
•These nutrients can come from natural watershed
weathering and biological processes
•However, high nutrient concentrations are usually
associated with anthropogenic inputs (e.g., from
wastewater discharges, agricultural runoff,
stormwater, etc.)
https://www.inaturalist.org/guide_taxa/707293
Quantification of nutrient inputs, sediment storage and release,
and long-term recover in Utah Lake
Outline
1.The general mass balance model
2. The first-order TP mass balance for Utah Lake
3. The effect of external nutrient sources on Tplake
4. Transition to a new steady-state
5. Is internal loading included in this model?
6. Conclusions
The nutrient mass balance in lakes
Change in concentration = Inputs – Outputs – Removal
The nutrient mass balance in lakes
Change in concentration = Inputs – Outputs – Removal
It is conventional practice to assume “steady-state”
to characterize long-term average conditions
The nutrient mass balance in lakes
Change in concentration = Inputs – Outputs – Removal
Or
Inputs = Outputs + Removal
0
Inputs = Outputs + Removal
Inputs = Outputs + Removal
Inputs = Outputs + Removal
Outputs
Inputs = Outputs + Removal
Mass balance models quantify Net Removal
(sediment deposition - sediment release)
Sediment accumulation
Phosphorus vs. Phytoplankton Biomass
0.1
1
10
100
1000
Ch
l
o
r
o
p
h
y
l
l
(µg
L
-1
)
1 10 100 1000
Total Phosphorus (µg L-1)
y = 0.08x1.5
r2 = 0.91
Jones and Bachmann (1976)
•Phosphorus and Nitrogen
concentrations are the main
determinants of phytoplankton
biomass
•Cyanobacteria blooms are also
strongly associated with high
phosphorus and nitrogen
concentrations
Phosphorus vs. Phytoplankton Biomass
0.1
1
10
100
1000
Ch
l
o
r
o
p
h
y
l
l
(µg
L
-1
)
1 10 100 1000
Total Phosphorus (µg L-1)
y = 0.08x1.5
r2 = 0.91
Jones and Bachmann (1976)
•Mass balance tells us that
nutrients are either in the water
or in the sediments!
•Due to Stoichiometric
constraints only the nutrients in
the water at any given time
contribute to phytoplankton
biomass!
•The nutrients in the water
column on average equal the
output/advected concentration
that regulates phytoplankton
biomass
Change = Input – Output – Removal
V = lake volume
Q = water flow into/out of lake
C = concentration into/out of lake
s = first order loss rate of nutrients
1. The general mass balance model
Where the time rate of change is due to the mass flux into the system, mass flux out of the system, plus or
minus any mass changes within the system, where V represents the volume of the reactor (or lake in this
example, L3), Cout represents the constituent (or nutrient) concentration in the outflow (M/L3), Qin represents
the total inflow to the reactor (L3/T), Cin represents the flow-weighted average input concentration (M/L3), Qout
represents the total outflow (L3/T), and 𝜎represents a first-order reaction rate for the removal or production of
the constituent in the system. In this model, Cout also equals the concentration in the reactor or Clake because
the reactor (or natural system) is assumed to be perfectly mixed. L above represents length, and volume or L3 is
commonly represented as m3. M represents mass, and T presents time.
In this case, we focus on first-order kinetics (e.g., exponential decay and growth) for the sake of simplicity, but these
reactions can also be zero-order, second-order or even intermediate-order.
Change = Input – Output – Removal
If we assume steady-state or long term
average conditions
0
Input = Output + Removal
V = lake volume
Q = water flow into/out of lake
C = concentration into/out of lake
s = first order loss rate of nutrients
where = the lake’s water residence time
relative to inflows (i.e., V/Qin)
=
where = the lake’s water residence time
relative to inflows (i.e., V/Qin)
=
Due to the high
evaporative losses
in Utah Lake
where = the lake’s water residence time
relative to inflows (i.e., V/Qin)
=
Without these high
evaporative losses
the model predicts a
TPlake concentration
of 22 µg L-1
where = the lake’s water residence time
relative to inflows (i.e., V/Qin)
or in term’s of Total Phosphorus (TP)
=
Lake TP concentrations are determined by the flow weighted
input concentration (TPin)
The first-order net loss rate to the sediments (s)
The amount of time available to be lost to the sediments ()
2. The first-order TP mass balance for Utah Lake
Lake TP concentrations are determined by the flow weighted
input concentration (TPin)
The first-order net loss rate to the sediments (s)
The amount of time available to be lost to the sediments ()
2. The first-order TP mass balance for Utah Lake Without the high
evaporative losses
the model predicts a
TPlake concentration
of 22 µg L-1
•Phosphorus removal depends on the water
residence time () and the loss rate (s)
From: Brett and Benjamin (2008)
Freshwater Biology 53: 194–211
N = 305
Based on a tremendous effort originally by LaVere Merritt (who laid a very
strong foundation) and subsequently by Kateri SalkGundersen, Scott Daly
and various sub-committee members to quantify nutrient inputs from
tributaries, WWTPs, Atmospheric Deposition, etc. (and accounting for
attenuation between WWTPs and Utah Lake)
TPlake = 68 µg L-1
TPin = 343 µg L-1
Qout/Qin = 0.29
= V/Qin = 1.09 yr
s = 4.31 yr-1
TP Removal = 1 - (Q*TPout/Q*TPin) = 0.937
•Phosphorus removal depends on the water
residence time () and the loss rate (s)
•Utah Lake is a very effective sink for
phosphorus
•The 96th percentile for TP removal relative to
these data
•This is probably due to reactions with CaCO3in the lake and formation of calcium-P
mineral complexes in the sediments
From: Brett and Benjamin (2008)
Freshwater Biology 53: 194–211
N = 305
ೝ್ೠೌೝೞ ೝೌೞ ೝೌ ೈೈು
ೝ್ೠೌೝೞ ೝೌೞ ೝೌ ೈೈು
3. The effect of external nutrient sources on TPlake
ೝ್ೠೌೝೞ ೝೌೞ ೝೌ ೈೈು
ೝ್ೠೌೝೞ ೝೌೞ ೝೌ ೈೈು
Vary WWTP effluent
TP concentrations (TPeff)
We can use this mass balance approach to predict how the lake’s
phosphorus concentration will depend on the input
concentrations for the major point sources
We can use this mass balance approach to predict how the lake’s
phosphorus concentration will depend on the input
concentrations for the major point sources
Increasing capital and O&M
costs, energy use, and
greenhouse gas emissions at
lower WWTP effluent
concentrations
Example from
Spokane Basin
of WA state
0
20,000
40,000
60,000
80,000
100,000
CO
2
Re
l
e
a
s
e
d
(
k
g
d
-1
)
0
20
40
60
80
100
120
In
p
u
t
T
P
(
µ
g
L
-1
)
1001000
WWTP effluent TP (µg L-1)
TPRIVER
CO2
Re
s
e
r
v
o
i
r
A few calculus steps
= concentration at time t
= new steady-state concentration
= initial concentration
r = flushing rate = 1/
4. Transition to a new steady-state
Transition to a new steady-state in Utah Lake
s = 4.31 yr-1
r = 0.29 yr-1
This model predicts transition to a new steady-state
in Utah Lake will be mainly governed by the
removal term (s) and will be rapid
Transition to a new steady-state in Utah Lake
Transition to a new steady-state in Utah Lake
5. Is internal loading included in this model?
5. Is internal loading included in this model?
How to estimate internal loading at a lake wide
scale?
a.Upper Klamath Lake example of internal
loading
b.Calculate Utah Lake internal loading
How to estimate internal loading at a lake wide
scale?
a.Upper Klamath Lake example of internal
loading
b.Calculate Utah Lake internal loading
c.If internal loading in Utah Lake was 1,500
tonnes P/yr, that would be equivalent to
adding ≈ 2,200 μg P/L to the lake water
each year above and beyond all external
inputs.
How to estimate internal loading at a lake wide
scale?
a.Upper Klamath Lake example of internal
loading
b.Calculate Utah Lake internal loading
c.If internal loading in Utah Lake was 1,500
tonnes P/yr, that would be equivalent to
adding ≈ 2,200 μg P/L to the lake water
each year above and beyond all external
inputs.
d.Recall, only the nutrients actually in the
water column at any given time support
phytoplankton blooms
Internal loading comparison with Upper Klamath Lake
(in south-central Oregon)
TPIN adjusted for
evaporation
We can use this approach to quantify summer releases of phosphorus
from the sediment back to the water column (i.e., “internal loading) for
Utah Lake based on long-term monthly average concentrations
•This internal loading
averages + 24 µg/L
•This is equivalent to about
30 tonnes P/yr of internal
loading
•If a more flexible approach
is used this estimate
increases by about 50% to
45 tonnes P/yr
Interestingly, this internal loading tends to follow the summer
phytoplankton bloom, and not vice-versa
•This internal loading
averages + 24 µg/L
•This is equivalent to about
30 tonnes P/yr of internal
loading
•If a more flexible approach
is used this estimate
increases by about 50% to
45 tonnes P/yr
Interestingly, this internal loading tends to follow the summer
phytoplankton bloom, and not vice-versa
•A similar pattern is seen
in Upper Klamath Lake
External Loading Tonnes TP/yr
WWTPs 133.4
Tributaries 49.6
Atmos Dep 32.0
Internal Loading 45.0
62%
23%
15%
Conclusions:
Utah Lake has quite high phosphorus inputs mostly from WWTPs
Conversely, the lake is exceptionally effective at removing phosphorus
from the water column and sequestering it in the sediments
However, lake TP concentrations are still directly linearly related to input
concentrations
Recovery once external inputs are reduced is predict to be very rapid
Internal loading increases summer TP concentrations by about 33 µg L-1