DELTA INDEPENDENT SCIENCE BOARD: The California Environmental Flows Framework

California water is a seemingly unending battle of allocating an often scarce resource among cities, farms, and the environment.  In particular, the question of how much water should be left instream for the environment rather than diverted for human use is perhaps the most controversial.  Most (if not all) people would agree that some water should be left in the rivers and streams, but just how much?

California’s environment is incredibly diverse; it includes chaparral scrub, temperate coniferous forests, mountains, desert, and the coast, and the rivers and streams that flow through the landscape reflect that diversity.  This means that what works for environmental flows in one area will likely not work in another.  Further complicating matters, multiple state and local agencies share responsibility for determining the flows needed to protect freshwater ecosystems.  Still, the vast majority of streams and rivers in California do not have any instream flow prescriptions, and where such flow criteria have been developed, those efforts have been poorly coordinated and have not resulted in effective protection of the ecosystem.

Enter the California Environmental Flows Framework, which uses a functional flows approach to develop consistent science-based recommendations for setting ecological flow criteria that is flexible enough to apply statewide.  The Framework was recommended in Water Resilience Portfolio as a way to better protect the environmental flow patterns that sustain fish and wildlife.

At the March meeting of the Delta Independent Science Board, Dr. Sarah Yarnell, professor of research at the Center for Watershed Sciences at UC Davis, and Dr. Julie Zimmerman, lead scientist for the Nature Conservancy’s water program in California, gave a presentation on the state of the science of environmental flows, an overview of the California Environmental flows framework, and the opportunities and challenges for improving the science underpinning environmental flows.

Part 1: The science of environmental flows

Dr. Yarnell began the presentation with some basic definitions. 

Flow: The master variable

In all river systems, flow (or discharge) is often referred to as the master variable.  It is a quantifiable measure of how much water moves down a river past a particular location in time.

Flow is highly influential to the way rivers and ecosystems work.  The water moving over the landscape scours and erodes.  It carries the rocks and the sediment in the river channel.  It creates habitat in river channels and the structure and shape of the river channel that species occupy.  So, that interaction between water and sediment essentially controls the ecology of the river system.

Flow is highly correlated and controls water temperature within the river, an important variable for aquatic species that live within the river environment.  Nutrient transfer and carbon cycling from the riparian zone and within the river are also highly related to flow, she said.  So we tend to focus on flow as a primary driver within river ecosystems that then influences all of these other variables.

One of the key concepts in understanding river ecosystems and their function is the natural flow regime.  There are several components to understand how the flow moves down a river system, such as:

• Magnitude of flows: How big is a flow in the channel?  How much water is in a channel at a given point in time?
• Frequency of events: How often are high flows occurring?  How often are the periods of low flow occurring?
• Duration: What is the duration of those flows within the river channel? How long does the high flow occur?  How long is the low flow season?
• Timing: When do these events occur?
• Rate of change: How quickly are flows changing on a day-to-day basis or even an hourly basis?

These are all characteristics of flow that can be quantified and used to describe the natural flow regime as the water moves through the river system in both time and space.

The natural flow regime

The graphic is an example annual hydrograph for a mid-elevation stream in the Sierra Nevada for the water year from October 1 to September 30. 

On the hydrograph, river flows and discharge increase in early November following a storm event.  The rain falls and runs down the hills into the river channels, wetting up the channel, and creating a peak flow event. 

When winter storms come through at the middle elevations, they often quickly runoff and create flashy peak events, generally from October through February. But those same storms are also dropping snow at higher elevations.  When that snow melts in the springtime, it creates a spring snowmelt pulse followed by a recession or a gradual decline of flows down into the summertime.  The wet season is followed by low flows in the dry season when there is little or no precipitation. 

Dr. Yarnell pointed out that California’s native species have adapted to this flow regime.  For example, chinook salmon migrate at the beginning of the wet season, when those first storms come down and increase flows into the river channel, providing migratory corridors for the adult salmon to get up into the river channel.  They lay their eggs, the smolts come out, and eventually outmigrate following that spring recession. 

Even the willow and cottonwood’s seed dispersal is timed to take advantage of the spring recession.  They typically drop their seeds in the springtime; the seeds will land on the typically higher flows, and as those flows are gradually receding, the seeds disperse and settle out onto the gravel bars.  The seeds germinate, and the roots follow that gradual recession of flows, growing downward to maintain contact with the water.

“So our native species in California are very diverse,” she said.  “They are very highly adapted and evolved to these natural patterns.”

Humans also need water, which impacts the river system and its natural flows in numerous ways.  The dams create structural breaks that disconnect the upstream and downstream reaches, but they also change the nature of the flow being released in the river downstream of the dam.  

“Typically, we capture the high flows that come down in winter for water supply to deliver them for human uses for agriculture in the summer,” said Dr. Yarnell.  “This fundamentally changes the way that the water is traveling down through the system.  And it creates patterns and seasonality that are different from what many of our species are adapted to.  There are also water withdrawals; we often take water out of the river and don’t put it back in.  So the nature of the magnitude, the frequency, and the duration of how this water is moving down through the system are all impacted by the way that humans have impacted the rivers.”

As for the science, we’ve done an incredible job understanding different aspects of river ecology, but it’s still complicated.  The hydrologic aspects of how water moves through the system and how surface water interacts with groundwater are key drivers in understanding food webs.   The impacts on one species can cascade to affect other species within the system.  And the geomorphology and geology of the landscape that the water flows over have a whole series of nested hierarchies and different scales in the way that it creates habitat from the watershed level down to the microhabitats.

“All of this together really makes a complex system that’s very complicated and often hard to quantify, but we have done a good job of starting to look at a lot of these processes to better understand how the systems work,” said Dr. Yarnell.

What water managers want

“The fundamental challenge in managed river systems is how do we take all of that complex ecology and complex understanding about how river systems work, and translate it into something that we can give to a water manager?” she continued.  “Managers just want a number – like how much water should I release from this dam?  Or how much water am I allowed to divert at a particular time?  They just really want a number, but it is quite difficult to condense all of that ecology down into a single number.”

She noted that most operators routinely work with flow schedules.  “So if we can effectively translate a lot of the complex ecology into things like a flow schedule, it can be more accessible and easy to manage and operate at a dam.  That’s the primary task that, as scientists, we were asked to do with environmental flows.”

Environmental flow methodologies

There are many methodologies for determining how much water to leave in the system to support the ecology in managed systems.

There are four basic categories of methodologies:

Hydrologic methodologies tend to focus on just the flow regime alone.  For example, a daily percent of flow, such as a rule that says no more than 10% of the daily flow can be diverted at any point in time.

Hydraulic methods look at the flow and the shape of the flow in the channel, considering the stage or depth of the water in the channel and the velocity that creates certain habitats for species.

Habitat-based methods consider the hydraulics between the relationship between the channel shape and the flow and then relate that to the physical habitat and the biology.  So an example would be looking at one species at one life stage, determining the flow and hydraulic needs of that species for that life stage, and then implementing environmental flows that provide those aspects.

Holistic methods:  The first three categories are data-driven, site-specific, and require a fair amount of knowledge about the system, but they only look at parts of the system.  So there’s been a movement in the last two decades to move towards more holistic environmental flow approaches that try to better address the entire complexity of the ecosystem.

“The problem isn’t that flow-ecology relationships are bad, or that these environmental methods are bad; they are not,” said Dr. Yarnell.  “They can be very useful in understanding how we might want to manage water in our systems for ecology.  But they are limiting in the sense that a lot of flow-ecology relationships are for a limited set of flow metrics.”

For example, they might focus on flow magnitude or the depth and velocity without considering things like water temperature, or nutrient cycling, she said.  They tend to be averaged over a flow record, so they’re perhaps less variable over time.  They are almost always single-species focused.  But it’s hard to understand whether or not that flow is also supporting the other related components within the ecosystem.  They tend to be based on observed empirical data rather than process-based data, and they may not account for shifting baselines or changes in climatic conditions.

“So it’s not that these methods aren’t useful; it’s just that they have a limited set of applicability that we need to be mindful of if we’re going to apply these types of methodologies in our river systems,” she said.

Dr. Yarnell emphasized that it’s not about flow volume or a single discharge that needs to go down the river; it’s about the pattern and the timing.  The hydrograph on the left shows a natural flow pattern; the hydrograph on the right shows a dam-altered flow pattern. 

The gray shaded area under those hydrograph curves are the same annual flow volume, but the hydrographs show how that same volume can provide fundamentally different conditions in the river system that may or may not be helpful for the aquatic species that are there.  The figures below the hydrograph illustrate what the low flows and high flows for that flow regime might look like in the channel.

“So it’s really important to move away from the discussions of annual flow volume alone without considering how we will be applying that flow volume in a river system and thinking about what that flow looks like when it’s in the channel,” she said.  “It behooves us to really think more holistically about the ecosystems needs and this variability of flow and geomorphology over time and space to try to do our best to understand and support the ecology within the system.”

Functional flows approach

One holistic method is the idea of functional flow, which considers what aspects or components within an annual hydrograph provide functionality to the ecosystem as a whole.  This doesn’t mean putting the entire natural flow regime back in the river, Dr. Yarnell noted, but rather, if we’re going to put flows back in, what are the aspects of the flow regime that we should target to have the greatest impact.   

For example, we might focus on components that support natural disturbances, such as high flows getting out on the floodplain that scour and erode, creating habitat.  Or getting moderately high flows to wet the gravel bars and allow cottonwood to germinate their seeds.  Or flows that provide particular ecosystem functions for life history strategies, such as high flows in springtime when salmon migrate.

“It’s thinking about the functioning of what those flows are doing more holistically, rather than thinking about a single species approach,” said Dr. Yarnell.  “And we have to take into account not only the flow, but that geomorphic setting and those channel flow dynamics, so the water can function because it’s connecting with the landscape in an appropriate way.  It doesn’t help to put a peak flow down a channel if the functionality we’re after for high flows can’t be achieved because there isn’t the space for that high flow to spread out.”

The hydrograph on the slide below is an example of a mid-elevation, mixed rain/snowmelt-driven hydrograph that might occur in the middle elevation of the Sierra Nevada.  The blue represents an unimpaired or natural flow regime in the mid-elevations of the Sierra; the red represents a functional flow regime that focuses on these four flow components that are known to be important.

“So the wet season initiation flows begin, that first pulse wetting up the river that comes down is often captured behind the dam because people say, ‘finally, water is coming down,’” said Dr. Yarnell.  “But that has such a huge functioning in rewetting the system.  These are the migration pulses that the salmon use as a cue to start to move upstream.  And they tend to be relatively small, so we might choose to release that flow downstream to really drive some of that ecosystem functioning. … Even lower dry season base flows can be important to help provide that natural seasonality between the wet season and the dry season that our native species are adapted for.”

Peak magnitude flows are quite important within our Mediterranean climate systems because these are the primary disturbances.  These are the large floods that come through, connect to the floodplain, move sediment, create habitat, destroy habitat, and create that dynamic nature that leads to more and greater habitats available for biodiversity.  These floods reset the natural processes, redistribute sediment, and reduce vegetation encroachment, so they need the space to function.

“So a high peak flow at full magnitude can provide an incredible amount of functions if we have access to the floodplain, or we have our levees set back, where there is space for the river to move, and create that dynamic diversity of habitats that support the functioning of our ecosystems,” said Dr. Yarnell.  “We would potentially then advocate for not necessarily releasing every peak flow at the full magnitude, but rather releasing multiple floods at a lower magnitude.  It might be more beneficial to release one flow at a peak magnitude to get that full bang for your buck.  And maybe some of the other peaks get stored in the reservoir for human uses.”

The spring pulse recession is also an important ecological trigger.  It’s the high to low flow transition once precipitation has stopped, and the snowmelt from the upper elevations that extends the cold water and water volume in the rivers well into summer.  As a result, almost all native species have evolved to take advantage of the predictable recession of transitioning from the wet to dry season.  

The spring pulse recession resorts sediment, helps to provide habitat, limits vegetation encroachment, and provides timing cues for outmigration.  And research has found that the longer the floodwaters are on the floodplain, gradually receding into the summer, the more ecological benefit there is.  So it’s an important aspect to include in managed flow regimes.

Another important aspect is the interannual flow variability.  California has incredibly wet years and incredibly dry years, which adds to the diversity of habitat types that are out there, so it’s important to vary the type of flow regime with the climate.

“If we have a really wet year, we want to put more water down the channel, we want to make our spring recession bigger, we want to have a longer duration, and we want to take advantage of the water that’s there to be able to support the ecosystem,” said Dr. Yarnell.  “Then conversely, in drier years, we would want to retain those flow components that are there, but they might be of a smaller volume or a shorter duration just because there’s less water available within the system.  But over time, wetter years will be more beneficial for some species and drier years might be more beneficial for some species.  So that interannual variability is just as important as that seasonal variability.”

Lastly, there is spatial flow variability.  The graphic on the slide is an illustrative schematic of a watershed in the Sierra Nevada; tributaries flow down out of the mountains onto the Central Valley floor and eventually out to the Delta.  The triangles represent dams that release water.

For example, the purple triangle might represent a dam in the higher elevations of the watershed.  The purple line on the graph shows the functional flow regime of the dam, which is releasing a small snowmelt pulse as the snow is melting.

Further downstream, a smaller reservoir and a smaller dam, shown in pink, might release a winter peak flow from rain-driven runoff and might also have a subsequent snowmelt recession flow.  The yellow line on the hydrograph might be the next watershed with a snowmelt release hydrograph.

“If we coordinate the timing of these releases, they start to have a cumulative effect,” she said.  “There might be some slight shifts in the timing of when these pulses are coming down naturally, but they can be coordinated such that by the time we get to the larger dam, shown in red, we’ve been able to take advantage of those cumulative releases to be able to provide a more effective, potentially larger type of functional flow regime.  So we need to think about the spatial variability of flow and how we might be able to coordinate the management of our rivers to be as effective as possible with the environmental water that we have to create the desired conditions in ecology that we want.”

Functional flows in California

Five primary functional flows have been delineated that are important for California: the fall-pulse flow, the peak magnitude in the wintertime, the spring recession flow, the dry season base flow, and the wet season base flow, which acknowledges that the wet season should have higher magnitude flows than in the dry season.  

Functional flows need flow to function.  “I get asked this question a lot: Is it more about the flow?  Or is it more about the habitat?  What if I just go do a bunch of habitat restoration?” said Dr. Yarnell.  “They are intimately linked, and you cannot think about one or the other: it must be flow plus habitat.  So in some cases, to make the flows functional, we might need to think about doing physical habitat restoration coupled with a more functional flow regime.  We might need to think about connecting to the floodplain or setting back our levees so that those flows can function as we want them to.”

Implementing a functional flows approach in California is hard.  “We have a huge state; it’s incredibly diverse and complex.  The North Coast watersheds are very different from the watersheds in Southern California.  And it can be very hard to understand the environmental flow needs across the state and figure out how to balance these with the broad range of demands that can range from hydropower to water supply, to wastewater treatment to wastewater discharge.”

And on top of that, a paper by Julie Zimmerman and co-authors evaluated stream gauge data and determined that flow is altered pretty much everywhere in the state.  The red dots on the map indicate flow depletions at the gauges, whereas the blue dots indicate flow inflation.  

“The summary of this is that we tend to hold back all of the water in the wet season behind the dams and release only minimal flows during the wet season, and then we tend to augment our dry season flows to release more flow in the summer.  So we’ve taken a natural hydrograph that would be more seasonal with high flows in the winter and low flows in the summer, and we’ve essentially evened it out; we’ve flatlined it.  And that’s essentially what this data shows.  So trying to work back from that and think about how to change that can be quite a challenge.”

Many regulatory programs are working on setting environmental flows throughout the state that involve different agencies, groups, community members, and stakeholders working in different systems and watersheds.  They might have different endpoints, management goals, ecological goals, and priorities.

“That’s a challenge,” said Dr. Yarnell.  “It makes it difficult to coordinate throughout the state to share knowledge, share understanding, and think about solutions that might be working in one place that might work in a different place.  So several years ago, the need was recognized for a more coordinated approach to developing environmental flows throughout California.  So a large group has come together over the last three years to develop the California Environmental Flows Framework that is an effort to put this into place.”

Part 2: The California Environmental Flows Framework

Next, Dr. Julie Zimmerman gave a presentation on the California Environmental Flows Framework.

Environmental flows started as fairly simplistic hydrologic methods of keeping minimum flows instream.  However, over time, it was realized that aquatic ecology is more complex, so site-specific studies were done to determine the local relationship between flow and ecological outcomes.

The problem was that the minimum flows approach could be applied everywhere, but it wasn’t very protective.  On the other hand, site-specific studies were more protective but could only be implemented in a few places because of the time and resources needed for the studies. 

“We really need something that balances the two,” said Dr. Zimmerman.  “So we came together with a group of partners who all worked on environmental flows in California and thought if we could all bring our common knowledge together and come up with a framework that to bring these two things together, we could develop something that was complex in its conceptual underpinnings, but relatively simple to apply in a management context.”

Overview of the California Environmental Flows Framework

“So that’s what really started our work on the California Environmental Flows Framework.  The Framework provides technical guidance for managers to develop scientifically defensible environmental flow recommendations, following the functional flows approach.”

The Framework utilizes a multi-step process to define two different sets of flow criteria: ecological flow criteria and environmental flow criteria.  Those are two different things, she said.  People tend to use those terms interchangeably, but how they’re used in managing flows is different.   

Ecological flow criteria are metrics that describe the range of flows that must be maintained within a stream and its margins to support the natural functions of healthy ecosystems.  They can be thought of as the ecological only scenario; what flows would you put in streams that maybe allow some water diverted for human use but are meant to maximize or support ecological outcomes.

Environmental flow recommendations are metrics that consider human uses and other management objectives along with ecological flow criteria.  These are ‘balanced flows.’  They are set by evaluating trade-offs among a broader set of management objectives, like human uses for water.  They consider environmental needs but are not intended to maximize ecological outcomes.

This figure presents an overview of the CEFF framework.  It is a twelve-step process organized into three sections:

Section A determines the natural ranges of flow metrics for each functional flow component for the location of interest. 

Section B considers other aspects such as the habitat context, geomorphology, water quality, and the biological interactions in the stream.  Is site-specific information needed, given any physical or biological constraints?

The outcome of Section A and B is a set of ecological flow criteria.  This could also be used to prepare a scientific basis report to support the ecological flow needs, such as what might be requested by the State Water Board.

Section C is the balancing.  This involves reconciling ecological flow needs with non-ecological management objectives to create balanced environmental flow recommendations.

Section A

Section A is intended to be a straightforward process for any user.  The steps identify ecological flow criteria based on the reference ranges of functional flows.  Tools developed to support this process include a web app for functional flow metric predictions, which includes a calculator to input a time series and calculate functional flow metrics and a repository of research papers.  

“From these tools, you can actually click on a stream reach of interest and download the modeled reference base functional flow metrics for your place of interest,” said Dr. Zimmerman.  “So you need to go through the steps of thinking through, what are the ecological management objectives for the location of interest?  Are there any site-specific needs where maybe I need to go into a more in-depth process?  But you can still come out of Section A with a set of quantitative flow predictions that you can use for implementation.”

The slide to the left shows the conceptual model that forms the basis for the Framework.  

“In Section A, the model says that if your physical habitat, water quality, and biological interactions are within reference ranges, and you think your reference flow ranges would support ecological outcomes, then you can use the flow criteria just from Section A,” said Dr. Zimmerman. 

“That’s complicated, but what it’s really saying is there might be some alteration, but if you put down reference ranges of your flow metrics, you would get the ecological outcomes you care about.  So there’s not a huge dam that’s elevating water temperatures, or the channel is so incised that flood pulses wouldn’t flow out into the floodplain.  If you put down reference ranges of flows, you would expect them to be able to support the ecosystem.  And in that case, you can use flow criteria from Section A.”

The slide below shows the five functional flow components: the fall pulse, peak magnitude flows in the wet season, wet season baseflow, spring recession, and dry season baseflow.  The table on the right lists the flow metrics used to quantify the flow components, such as magnitude, duration, timing, and frequency.

There are 24 metrics, which include three different peak flows and two magnitudes for the wet and dry seasons, which were modeled using a machine learning approach based on data from reference gauges across California.

The technical team used flow data from about 250 reference gauges in California and PRISM data on rainfall and temperature to derive watershed variables.  They next developed a model to predict functional flow metrics from 1950 to 2015 at the reference gauges.  They then applied those models to ungauged locations.  The result is estimates of the reference range functional flow metrics for every stream segment in the state for three different water year types: dry, moderate, and wet water years.  Hydrologic model predictions were used for sixteen metrics; observed reference gauge data was used for eight metrics.

All the data is available on a public web app maintained by the Nature Conservancy called the Natural Flows Web Tool at rivers.codefornature.org.  There are two datasets available: a set of monthly flow predictions from 1950 to the present and a set of functional flow metrics.  Users can click on any stream reach and get predictions for both monthly flow and functional flow metrics for that reach.

The example on the slide shows the dry season baseflow metric predictions for the Tuolumne River.  The user can choose different water year types to get the different flow values,  the median values, and a prediction interval around those values derived from the machine learning approach. 

The bottom window shows the monthly flow prediction hydrographs for dry, moderate, and wet year types.  The information is publicly available and can be downloaded.

At the end of Section A, users then assess their site to consider if the flow data needs to be refined to reflect local conditions, such as being downstream of a dam or highly incised channels.  If so, which of the flow criteria can be used and which need to be refined?  This is done in Section B.

Section B

Section B is only needed when physical habitat, temperature, water quality, and/or invasive species are such that reference base functional flows are not expected to achieve their intended ecological functions.

It also could be used when there is local data available that could enhance the predictions from the statewide model, such as a local model that might better integrate groundwater.  It can also be used to develop specific flow ecology relationships to support trade-off analyses in Section C.

“Sometimes people want flow ecology curves so that they can look at different management alternatives and see the effect of implementing something different on the ecology,” said Dr. Zimmerman.  “Those trade-off curves can be done using developed in section B.  So we don’t expect Section B to need to be applied everywhere, or for every functional flow component.”

This is a conceptual model for a hypothetical North Coast stream channel that had been widened through land use change and logging practices.  The reference range of the dry season baseflow was not expected to create sufficient habitat conditions for coho salmon; it would result in flows that were too low and too slow because the channel was too wide and had higher temperatures.

“In this case, you wouldn’t need to go through Section B for your other functional flow components; you would just focus on the dry season baseflow – the specific things you care about are water quality, the temperature component, and depth and velocity for salmon habitat,” said Dr. Zimmerman.  “You might do a temperature model and look at temperature suitability; you might do a physical habitat model, looking at depth and velocity requirements.  As you go through this, you will likely find is that you need flows higher than the natural reference ranges to achieve those outcomes because you’d need more flow to get cooler temperatures and more flow to get the depth and velocities within the ranges that you’d need to support salmon.”

“Then you can start to think, can I do that?  In this place, if you don’t have a dam, it’s really difficult to increase flows in the summertime to achieve those conditions.  So that might lead you to look at habitat restoration or some other way to alleviate those issues.  This is something that you’d investigate in a Section B context.”

Once the Section B analyses are finished, those criteria are integrated with the Section A criteria, giving a full set of ecological criteria representing what would be implemented if ecological outcomes were the priority.

Section C

Section C is determining environmental flow criteria.  This section is similar to a structured decision-making approach or decision science where a full suite of management objectives is first defined.  Then, ecological objectives and other uses are considered, a set of management alternatives is determined, the consequences of implementing different alternatives are analyzed, trade-offs are evaluated, choices are made, and an implementation plan is developed.

Section C includes an assessment of flow alteration, which compares the ecological flow criteria developed in Sections A and B to the functional flow metrics calculated using observed flow data.  The method is outlined in an appendix on how to do an alteration analysis to determine if the flows are considered to be altered.

“The outcome of Section C is environmental flow criteria for all five functional flow components and an implementation plan,” said Dr. Zimmerman.  “Again, you could use the ecological flow components if you don’t care about other management objectives.  Or maybe for certain flow criteria, you only have issues in the summer, so you would only develop separate flow criteria for the dry season and keep the Section A or Section B flows for all the other functional flow components.”

The slide shows another way to look at the steps in Section C that is similar to an adaptive management wheel. 

“So once you clarify your decision context, define your objectives, assess your alteration, propose and evaluate alternatives, define recommendations and develop your implementation plan, you might then decide that you need to do something different, and you could be iterative about it and go through the cycle again.”

Dr. Zimmerman said they don’t know yet of any case study that has gone through all the Section C steps; the Framework is still fairly new.  However, case studies have utilized Section A and B; most have done the alteration analysis from Section C. 

“So rather than doing that whole suite of steps, they focused on the alteration, which is helpful for planning flow projects and looking at flow studies.”

Flow alteration analysis

The technical team did an alteration analysis for all the USGS gauge data with at least 20 years of post-1990 data.  Those gauges with alteration data available are shown in purple on the web tool.

The slide shows an example for the Tuolumne River.  The top box shows the wet season baseflow functional flow component, and the bottom line indicates the two-year flood functional flow component.  

“At the top line, what it shows is that the observed median wet season base flow is 226 CFS, but the predicted reference range would be between about 650 and 1760s,” said Dr. Zimmerman.  “So because the observed value is outside that prediction interval for the reference ranges, we consider that to be altered and depleted.”

“The same thing for the wet season median baseflow, shown on the lower graph.  The first column is the 10th percentile baseflow; the second is the median.  For the two-year flood magnitude, the observed value was 2980 CFS which is outside the prediction interval for the reference range, which would be about 9400 to 24,000.  So this data is available for several gauges across the state.”

Outcomes of the California Environmental Flows Framework

If a stakeholder completes the entire process, using the guidance documents and going through the steps, they will end up with:

Ecological flow criteria for areas of interest.  “When we talk about flow needs in California, it’s actually pretty rare that we start with a scenario that just looks at the ecology; we default to balance flows without being very explicit about the trade-offs between all of our management objectives.  So it gives you the ecological flow criteria that you can then use to discuss the trade-offs.”

Environmental flow recommendations.  “We recommend that this is done with a stakeholder process because those management objectives are really values-based; they’re not scientifically derived.” 

A set of recommended mitigation measures and an implementation, monitoring, and adaptive management plan.

“All of this is supported through these online tools that we’ve developed,” she said.  “There’s the web tool that I’ve been showing you.  A functional flows calculator is available, where you can run a daily time series of flows, and it’ll calculate your functional flow metrics.  This is what you would do if you had modeled or observed data, and then an information repository in the website, which has the framework document, a work plan, and links to all of the publications and some presentations on the approach.”

The research papers have been published in a special issue on environmental flows in the journal, Frontiers in Environmental Science.  Six of the papers in the special issue are from members of the technical team.

Read the special issue here: https://www.frontiersin.org/research-topics/16816/environmental-flows-in-an-uncertain-future#articles

Environmental flows in the Delta

Dr. Zimmerman noted that they did not do predictions for the Delta in the web tool.  While the model generated predictions, the technical team did not feel confident in those predictions.  However, she noted that there is a paper in that same special issue authored by other researchers that discusses ecological flows for estuaries. 

“I like this conceptual figure; on the left, it is a functioning estuary, and on the right is one that’s been impacted by drought,” she said.   “When we talk about ecological flows for the estuaries, we often think that there are different processes that need to be considered – it’s a focus on salinity or a focus on hydrodynamics, for example.  But it’s the outcome of having supportive ecological flows for the rest of the watershed.  So if we manage ecological flows well for the rest of the watershed, it will provide the flows for the estuaries, especially if it’s done so that they’re coordinated both spatially and temporally.”

“A lot of the processes are the same.  We need to have high flows in the winter to inundate habitat, we need to be able to flush salinities, and we need to be able to keep temperatures in the right range – all these things are really very similar.  You’re just adding that added dynamic of having the tidal action.”

Next steps for CEFF

Dr. Zimmerman concluded by talking about the next steps for California Environmental Flows Framework.  The Framework is a living document and a product of the Water Quality Monitoring Council’s eflows workgroup, jointly chaired by Dan Schultz with the State Water Board and Robert Holmes with the Department of Fish and Wildlife.  The workgroup reviewed the draft, and changes were made in response to comments received.  The version on the CEFF website is considered draft-final; it still needs to be approved by the State Water Board.

There are FAQs, an overview paper, and other supporting papers on the website.  Multiple case studies are underway in Southern California, the Little Shasta River, and Cosumnes River, and some papers have already been published.  There’s also a case study on the Eel River, but it’s not published yet.

The funding for the work done so far has been used up, so the team has put together a work plan for continued work on the Framework for soliciting funding and support proposals for the new efforts.  The work plan includes improvements to the technical tools, a proposal to model actual flow at all stream segments, improve the functional flows calculator, track and document case studies, develop mechanisms for ongoing data and information sharing, and training and outreach.

Part 3: Questions and Answers

DR. STEVEN BRANDT: I’m a fish guy, so I’m going to ask a question that’s probably very naïve … When I see flows as CFS or cubic feet per second – I understand why hydrologists and people talk that way. But it’s meaningless to a fish. The fish cannot detect the change in CFS; they cannot measure CFS … It has no physiological meaning to a fish, per se.  And frankly, I’m not even sure humans can measure it … So when you say 20 cubic feet per second over here is good and 25 cubic feet per second here is not good for a chinook salmon, that’s meaningless.  Because at any one location, changing cubic feet per second is correlated with velocity, depth, residence time, and temperature – things that fish can detect and have direct meaning to them.  If I have a change of 100 cubic feet per second in a small stream, it might impact fish through the actual things that are changing.  If I have 10,000 cubic feet per second change in the Mississippi River, it’s undetectable by a fish.”

“So I think that if you want to start talking about functional flows, you need to get into that Section B and pin down that what you’re changing is temperature, you’re changing velocity, you’re changing flows, but flows is just an indicator.  It’s meaningless, except in that one location … Am I missing something here?”

Dr. Yarnell:  “You’re absolutely right, and it works both ways.  So if we are going to have a conversation with a dam operator, they don’t give a crap about what the depth and velocity are five feet downstream.  So we rely on those relationships exactly like you said, between what the fish feel, what the fish care about, and how that relates to discharge so that we can implement it and put it into management.  So I completely agree with you.

“But if we’re thinking about what we do and how we can do it in a uniform way everywhere, we need to look for those more transferable and more applicable relationships.  And that’s exactly why we’ve done the modeling where you can go to all of the different stream gauges in the state and look at those ranges of discharge that are then scaled to and more indicative of the size of the creek.  Because you’re right, just putting down 25 CFS in one place will be meaningless and not necessarily comparable to 25 CFS and the other.  But the focus is not on the 25 CFS per se but on what that 25 CFS represents in that location.  And that’s the piece that we use to relate to the functionality because that’s the piece that we measure, implement, and regulate.”

“So I do not disagree with you at all.  I’m just talking about when we go to implementation and the languages and how we communicate with each other, that has been the most successful way to communicate.”

DR. BRANDT:  “I would suggest when you’re putting up a plot of something in a particular location of CFS, you might just put down velocity as a secondary scale.  It’s not always the same … But I think that people reading these charts believe that flow is the magic thing we’re trying to control.  lt is not; we’re trying to control temperature, velocity, depth, and those things.  And somehow, that reverse message needs to be made.”

Dr. Julie Zimmerman:  “My background is in fisheries and stream ecology, so I come from the ecology side.  We use CFS and think about it in terms of reference ranges or the percent of unimpaired because it is a proxy for these ecological conditions that we’re trying to achieve, which is having the right depth and velocity.  But apart from the depth and velocity that fish are experiencing, it’s also supporting food production, floodplain inundation, riparian vegetation, and all other things you need to have a functioning system.”

“On one side, you get to the very site-specific studies that only look at depth and velocity and ignore those other important ecological pieces, or you’re just looking at the hydrology side and not thinking about the ecology at all.  So we’re trying to get an approach that you can implement in a lot of places because it’s based on hydrology, but the underlying conceptual Framework is ecological in theory, so we’re trying to support multiple species and processes.”

DR. LISA WAINGER: I’m an economist, so what I’m trying to understand is what you’re providing in terms of evaluating those trade-offs.  So I see that you have this sort of ideal historic realm that we would like to recreate, but we probably won’t in most cases.  And then I see that you have recommendations, such as maximizing fish habitat or something like that.  But what happens in the middle ground?  Do we have any information about if you miss these targets by 20%, we will lose a species, cottonwood range will contract?  Do we have any real understanding of those relationships in any way?  

Dr. Zimmerman:  “I think we do in certain places.  We don’t have them that can be applied to all the streams statewide, which is why we developed this in the first place.  The science has been trying to develop these flow ecology curves and relationships that you can then apply broadly to look at trade-offs: If I were to take this much water, what would I get as far as ecological outcomes?  But as a field, we haven’t been able to take curves that we can generalize and broadly apply.  So this is meant to fill the gap.”

“We don’t have ecological flow criteria hardly anywhere in the state.  There are site-specific studies that have been done in a few places.  There are environmental flow regulations that have been implemented in places, but they’re not meant to be supportive of resilient ecological systems.  They’re meant to avoid jeopardy; they’re usually within the context of the Endangered Species Act, which is just to avoid extinction.  So we don’t have a broadly applicable framework that lets you look at those trade-offs.  So we’re providing the Framework to get the curves; you still have to go through Section B because we haven’t figured out a way to broadly make it generalizable across space.  So that’s what we’re trying to fill.”

DR. LISA WAINGER: Could this work be applied to understand priorities at a more landscape scale or regional scale?  Because we’re going to be allocating scarce resources, what if we could only restore some streams?  Is this a tool that would enable that conversation?

Dr. Zimmerman:  “We’re doing a lot of that in the Nature Conservancy.  We’re trying to figure out ways to do that.  You have to figure out your philosophy.  Do we take places that are really altered and try to work on those?  Do we take places that are just a little bit altered?  What’s your strategy?  It helps you plan and ask those questions and start to categorize places across the landscape.

Dr. Yarnell:  “What we have found that’s been an advantage is that by laying out and quantifying the pieces we know, and laying out the pieces we don’t know, it allows for those conversations to happen.  And so, for example, do I want to prioritize certain species because they’re listed?  Do I want to prioritize certain functions?  Do I think that this function is more important than this other function?  So we’re trying to be more transparent and more specific about those trade-offs.  Do we have the answers to all those?  No, but you do have the ability, I think, to have a more open, transparent conversation about what you do and don’t know as you move through that essentially sociopolitical process of trying to determine and make those trade-offs.  So that’s also one of the goals is to try to open up that conversation in that dialogue more transparently.”

DR. TANYA HEIKKILA: Given the complexity of the institutions and policies governing water in California, from water rights to endangered species to dam operating requirements and various other things, what do you think are some of the biggest challenges to implementing this type of Framework?  And conversely, what are there any kind of policy levers or incentives that could get people motivated to implement this?  Where do you see the long-term implementation challenges and incentives here?

Dr. Zimmerman:  “I think that there has been pressure to have flow criteria that have been defined and implemented in a lot more places.  I think our agencies have struggled to do those site-specific studies.  

“We get asked, especially in a year like this, what flow should I have in place so that I’m protective, knowing that it’s one of the driest years on record?  We can use our tools to help answer that.  We’re doing some work with the Salmon and Steelhead Coalition, which are partners with Trout Unlimited and Cal Trout on the North Coast, where we’re taking the monthly natural flow predictions, and looking at the distribution of percentiles over the period of record that we’ve modeled 1950 to present and saying, okay, where does the value we’re predicting for this year fall in that distribution?  And in a lot of places, it’s the lowest on record, or below the 10th percentile.  What does that mean, as far as trying to do an action to maintain flows in those streams and trying to get through the drought?  I think we have a lot of policy levers in drought years that we don’t have in other years.”

“We do a lot of flow management under ESA and the biological opinions, but most of the time, that’s just to avoid jeopardy.  I’d like to see us push that more to have recovery flows.  And I think this can help with that conversation.”

“The Sustainable Groundwater Management Act requires surface flows to be kept in good condition.  But there’s no data in a lot of places about what those flows should be, so I think it can help in that application.  So we’re starting to explore the list.  But right now, there’s not one policy that says, implement environmental flows in lots of places; it’s piecemeal looking at these different applications.”

Dr. Sarah Yarnell:  “I would just add, there are two challenges that the technical team and the environmental flows working group have been coming across.  One is that thinking about flows more holistically and not in a single species way towards the ESA is a challenge.  We’re asking a scientific community that spent 30 years thinking about flow ecology relationships and thinking about what do salmon need to rethink a little bit about the larger ecosystem and the functionality of that ecosystem.  So there’s a challenge in getting people to think about those ideas that maybe feel new.

“Then this idea of implementing a pulse that can have lots of benefits seems frustrating to some people.  Initially, we tend to think about what we care about.  Are you sure that pulse will benefit the fish that I care about?  And the answer often is yes, so let’s look at it, we need to show you, but then also getting folks to see that.  So this shift in thinking scientifically is a bit of a challenge that I think will require more time and outreach and discussion about this.”

“Part of that is the CEFF framework itself; it’s a very flexible framework that could be applicable to a lot of different regulatory levers.  The benefit is there are a lot of regulatory levers: SGMA, FERC, Endangered Species Act, water quality certs, wastewater discharge – all of them require knowing an environmental flow regime, or they need ecological flow criteria.  So this is a need that comes across all of these different regulatory agencies.  But folks need to know about it; they need to be able to talk to each other and share that information.  So those are the types of challenges that I think will require some time and some effort and some communication to help try to build that.”

“Then working through these case studies with folks testing this out, putting some of these types of flow regimes down, implementing them, monitoring them, and seeing if all of our hypotheses work out.  There are places in the state where we have done more of these functional flow regimes, such as the Trinity River and Putah Creek, and we’ve seen great success, but we want to try to be bolder about implementing these and get a lot more places to try it.  But on the other hand, the challenge is people start to nitpick, and they want to pull it apart, and it’s a lot easier to argue and go to your lawyer …

DISB MEMEBER: My question comes from the perspective of someone who works on ecological flows in low gradient landscapes.  One of the things I think about a lot is hysteresis.  Once flows become altered by things like vegetation encroachment, that would require a bigger kick in the opposite direction to restore the former ecological functionality.  And I’m wondering if the extent of alteration in some of these streams induces similar concerns?  One example might be stream bed armoring that would require even higher flows to mobilize sediment to form bars for fish spawning.  I’m just wondering whether this comes into your analyses or if it has been observed in practice?

Dr. Yarnell:  “I know you don’t have all the steps, and we don’t have the slides in front of us.  But step three of Section A is to ask those questions and think about if there could be considerations where your natural range of reference flows will not provide the functionality that you’re looking for.  So if you have a highly armored system with no sediment supply, it’s like having a highly incised channel that if I put a lot of water down it, I’m not going to get the function.  So that requires you to then shift into an analysis where you would be looking at those relationships and thinking about what to do to try to increase that functionality. 

“That’s where restoration might need to come into play, augmentation of sediment might need to come into play, and you might need to get out there with a bulldozer and take out the willow trees because there’s never been a scouring flow that’s come down your river.  So many rivers in the lower Sierra – the Middle Fork of the American River is a perfect example.  It’s dry rock 20 feet away from the channel, and willows are a foot wide at the channel edge because the flow literally never changes, so putting down a high flow is not going to rip those trees out.  You may have to go in and kick start it.  That’s where physical habitat restoration or some of these other methods are going to be needed to be coupled with that flow regime to make that functionality.”

“It’s never just physical habitat restoration alone.  For example, I’m going to put down 10 CFS all year, and I want to do a bunch of physical habitat restoration and make it perfect – that’s not going to work.  But putting down a flushing flow through a canal or through a boulder field is also not going to work.  So we have to be thinking about both pieces together, and the idea with the Framework is to set that up in that way.  And it becomes very easy when we talk about flows to forget about that side.  And that’s definitely not something we want to do, and we’ve hopefully explicitly created places for you to think about as you walk through the steps in the Framework.”

SR. STEVEN BRANDT: A temperature standard has been adopted in certain areas like Oregon, where decisions, for example, on restoration are based on trying to get the right temperatures. Flow is a component of that, but not the only one.  Building trees and things like that are also considered when you consider temperature as the endpoint rather than flow.  Is that even doable in the Delta?”

Dr. Zimmerman:  “I think most of our temperature management is reservoir management.  But I think there are places upstream where you could do habitat restoration and get some temperature outcomes.  It gets really hot here, and it’s harder because we’re on the edge of what is survivable.  So it makes it a little bit different than Oregon.  And we have these big dams.  I live on the American River, and Folsom is always on the edge of killing everything.  So I think it’s more challenging because we are just on the southern end of the species range.

Often, it does default to flow, but I think there are some headwater areas – we’ve lost a lot of spawning reaches in the headwaters.  We’ve lost a lot of passage up to those cooler temperature streams.  So I think that that’s going to be something that we have to consider is getting fish above dams where it is cooler, and maybe doing some habitat restoration would be part of that.  But I think moving fish into colder places is a big one.”