Download as pdf or txt
Download as pdf or txt
You are on page 1of 10

Climate-Adaptive Water Year Typing for Instream Flow

Requirements in California’s Sierra Nevada


David E. Rheinheimer, A.M.ASCE 1; Sarah E. Null 2; and Joshua H. Viers 3

Abstract: Water year types (WYTs), whereby years are classified by river runoff quantity compared to historical runoff, are one tool to help
make major water management decisions. Increasingly, these decisions include instream flow requirements (IFRs) below dams for river
Downloaded from ascelibrary.org by David Rheinheimer on 07/12/16. Copyright ASCE. For personal use only; all rights reserved.

ecosystem management. However, WYTs are typically based on assumptions of stationarity, and are thus rendered less meaningful with
climate change. Hydrologic alteration resulting from climate change means that a WYT-based IFR scheme using stationary historical ob-
servations might inadvertently result in long-term river management outcomes inconsistent with original water management goals. This study
assesses the management implications of assuming hydrologic nonstationarity in a WYT-based IFR scheme in California’s upper Yuba River
and demonstrates a rolling period of record as a climate adaptation strategy. The existing, nonadaptive water management scheme leads to
vastly different possible water allocation outcomes than originally planned for. Results indicate that water year types, if regularly updated, can
help maintain historical instream flow distributions. However, gains toward maintaining desired IFRs are obfuscated by future increases in
unmanaged reservoir spill. These findings indicate that hydroclimatic uncertainty can partially be accounted for with simple modifications to
existing operating rules for reservoirs, though other, risk-based management approaches are also likely needed. DOI: 10.1061/(ASCE)WR
.1943-5452.0000693. © 2016 American Society of Civil Engineers.
Author keywords: Climate change; Water resources management; Adaptation; Hydropower; Environmental flows.

Introduction Poff et al. 2007; Strayer and Dudgeon 2010; Vörösmarty et al.
2010; Xenopoulos et al. 2005). These threats affect unregulated
Nonstationary hydrology resulting from climate change means that and regulated rivers alike (Null et al. 2013a; Palmer et al. 2008;
solely using the historical record for long-term water planning and Rheinheimer and Viers 2015; Xenopoulos et al. 2005). Efforts
management is increasingly inadequate (e.g., Jiménez Cisneros to curtail environmentally harmful effects of water development
et al. 2014; Milly et al. 2008; Stakhiv 1998). To account for this, include provisioning environmental flows (e-flows) to help main-
researchers and practitioners are identifying climate risks and de- tain downstream ecosystems (Petts 2009).
veloping climate-adaptive water management rules. Management In contrast to climate change adaptation studies for traditional
practices that adapt to climatic changes can help maintain a more water management for human use, an area that has been well-
predictable, more equitable, and more beneficial distribution of a studied, studies that assess climate change and e-flows jointly
wide range of water system conditions compared to nonadaptive are rare. Wilby et al. (2011) demonstrated that a smarter scheme
practices (Georgakakos 2012). Although there are positive signs for streamflow abstractions for water supply, whereby instantane-
that water management policy is moving toward broad scale inte- ous flow thresholds limit abstractions, can reduce the risk of not
gration of climate change adaptation, there remains a wide gap meeting e-flow targets under drier conditions. However, Wilby et al.
in many instances between the scientific consensus and manage- (2011) focused solely on meeting low season flow targets, without
ment practice (Null and Viers 2013; Viers 2011). Not including considering the flow regime as a whole. Fung et al. (2013) exam-
climate change in water planning might result in unintended con- ined the implications of nonstationary hydrology with specific
sequences, such as unmet management objectives. e-flows rules, noting that “abstraction rules established today
While climate adaptation studies have traditionally focused on may set the pattern for abstraction for several decades.” In address-
human water uses, adaptation is equally needed for managing water ing this, Fung et al. (2013) used ensemble climate projections to
for ecosystems. Both human water development and climate assess the potential effect of climate change and flow augmentation
change are major threats to global freshwater biodiversity, the for- on e-flow targets in a regulated river with three different abstraction
mer from globally aggregated local effects and the latter from schemes (natural flow, business-as-usual, and full abstractions) in
alteration of the Earth’s hydrologic regimes (Heino et al. 2009; southern England. The rules considered in these e-flows case stud-
ies, while responsive to environmental conditions, do not account
1
Postdoctoral Researcher, School of Engineering, Univ. of for changing information about hydrologic and socioeconomic in-
California, Merced, Merced, CA 95343 (corresponding author). E-mail: formation through time, an important characteristic, among others,
drheinheimer@ucmerced.edu of a truly adaptive water management scheme (Georgakakos 2012),
2
Assistant Professor, Dept. of Watershed Sciences, Utah State Univ., and hence are not truly adaptive.
Logan, UT 84322. E-mail: sarah.null@usu.edu Water year types (WYTs) are discretized indicators of water
3
Associate Professor, School of Engineering, Univ. of California,
availability and are often used for water allocation purposes, such
Merced, Merced, CA 95343. E-mail: jviers@ucmerced.edu
Note. This manuscript was submitted on September 22, 2015; approved as water delivery or environmental flow allocations. WYTs are typ-
on April 28, 2016; published online on July 12, 2016. Discussion period ically defined by comparing forecasted or observed total annual
open until December 12, 2016; separate discussions must be submitted for runoff from a river basin against specified flow thresholds, with
individual papers. This paper is part of the Journal of Water Resources thresholds defined based on the distribution of historical runoff ob-
Planning and Management, © ASCE, ISSN 0733-9496. servations. These comparisons are translated to descriptive WYTs

© ASCE 04016049-1 J. Water Resour. Plann. Manage.

J. Water Resour. Plann. Manage., 04016049


(e.g., “dry” through “wet”), which are then used in allocation rules. • As WYT distributions change, are environmental and human
Water managers typically assume stationary hydrology when defin- water allocation objectives maintained?
ing and applying WYTs by relying on fixed runoff thresholds. • More broadly, how does water allocation to different water uses
However, distributions of WYT frequencies will necessarily shift with changing conditions? and
change as climate change alters runoff volume and timing if thresh- • How might alternative WYT adaptation strategies explicitly
olds remain static. maintain allocations to different water uses under changing
The use and misuse of WYTs, including exclusion of climate conditions?
change in their design and application, is exemplified in California, To answer these questions, a hydrology model forced with
where WYTs are used for major human and environmental water representative downscaled climate data was developed. Then, a
management decisions. For example, the California State Water monthly hydropower system optimization model was developed
Resources Control Board (SWRCB) (1999) Water Right Decision and applied with different climate-adaptive management schemes
1641 (Revised) (D-1641; SWRCB 1999) and the later U.S. and hydroclimate scenarios. The study focused on California’s
Fish and Wildlife Service (2008) Biological Opinion (U.S. Fish Yuba River basin, where WYT classes and water allocations are
and Wildlife Service 2008) both use WYTs for specifying well-defined for upper basin operations. The case study is first de-
Downloaded from ascelibrary.org by David Rheinheimer on 07/12/16. Copyright ASCE. For personal use only; all rights reserved.

minimum flows for water quantity and/or quality objectives in the scribed, including modeling methods, climate change scenario data
Sacramento–San Joaquin Delta, and both implicitly assume hydro- and selection, and management schemes, including the current non-
logic stationarity. Many similar examples exist. In examining these adaptive scheme and two climate-adaptive schemes. Second, the
issues for the Sacramento–San Joaquin Delta, Null and Viers hydroclimatic effects on WYTs and IFR allocations in the upper
(2013) found that WYT distributions are expected to change with Yuba River are assessed, assuming the current, nonadaptive IFR
warming and that environmental water uses are disproportionately scheme is used. Incongruities in regional WYT schemes are also
vulnerable to climate change if WYT definitions remain fixed by described. Third, the two climate-adaptive schemes are applied
static numerical runoff thresholds, potentially harming river eco- using the optimization model and water allocation outcomes are
systems. Adapting to climate change by regularly updating WYT described. Finally, implications of the findings are discussed and
definitions to maintain historical WYT distributions can more future management and research needs are suggested.
equitably allocate flows between water users. This paper extends
Null and Viers (2013) for environmental flow distributions and
high-elevation hydropower. Case Study: Yuba River, Sierra Nevada
Climate change has also been categorically and explicitly omit-
California’s Yuba River basin (Fig. 1) has complex hydropower
ted from hydropower licenses in the United States, which usually
operations and ongoing salmon recovery efforts, providing a useful
include instream flow requirements. This has implications beyond case study to explore WYTs, IFRs, and climate change adaptation.
the potential environmental consequences discussed above. Hydro- This study focuses on the upper Yuba basin, where water is diverted
power licenses in the United States are issued by the U.S. Federal via a series of medium-sized reservoirs (storage:inflow ratios of
Regulatory Commission (FERC) for 30–50 years, after which they 0.3–1.3) to the adjacent Bear River basin for hydropower genera-
must be renewed. Due to pressure from environmental interest tion and water supply (Rheinheimer et al. 2013). Historically, envi-
groups and government mandates, instream flow requirements have ronmental flows consisted of constant instream flow requirements
become increasingly more ecologically beneficial, by design. How- of 0.09–0.14 m3 =s, depending on location, below each reservoir.
ever, the lack of climate change in licensing policy potentially Recent negotiations for better instream flow requirements in the
undermines both intended environmental flow improvements and South Fork Yuba River below Spaulding Dam (SFY) (Fig. 1) have
even the central purpose of a license, which is long-term opera- resulted in an IFR scheme whereby minimum instream flows de-
tional certainty (Viers 2011). pend on WYTs that are defined with static annual runoff thresholds,
The objectives of this study were to understand the implications based on an estimate of unimpaired runoff in the greater Yuba River
of assuming hydrologic stationarity in a WYT-based water alloca- called the Yuba River Index (YRI) (U.S. Forest Service 2012)
tion scheme for a high-elevation hydropower system in the Sierra (Fig. 2). The YRI is calculated from estimated unimpaired annual
Nevada, California, and to better understand the potential for adapt- runoff on the mainstem Yuba River about 80 river-km below the
ing management rules to account for hydrologic nonstationarity. SFY flow requirement location and below the confluence of all
As stakeholder negotiations are the main mechanism for defining major branches of the Yuba River. The YRI and, therefore, the
water allocation schemes during the FERC relicensing process, WYT, is updated on the 15th of each month. In October–February,
final allocation schemes reflect some trade-off between environ- the YRI is based on the previous water year’s estimated unimpaired
mental, economic, and other interests. In this context, several runoff. In March–April, it is calculated from year-to-date
assumptions are made. First, environmental stakeholders have in- (October 1 through estimation date) plus forecasted year-to-go
fluence over water allocations. Second, environmental stakeholders (estimation date through September 30) unimpaired runoff. For
consider hydrological characteristics, such as frequency of WYTs, May–September, it is calculated from year-to-date plus year-to-
when developing environmental allocation rules, and implicitly or go unimpaired runoff as estimated in May. Because the WYT
explicitly assume hydrologic stationarity. Third, environmental al- can change through the year as the YRI is updated, IFR releases
locations such as IFRs are legally prioritized over other uses when can too.
implementing the FERC license. Finally, all stakeholders seek to
maintain or improve conditions that affect their respective manage-
ment domains, implying a long-term environmental goal of limiting Hydropower and Hydrology Modeling
uncertain and potentially ecologically harmful hydrologic change. The Bear River hydropower complex, which comprises 10 power-
Thus, historical IFRs were used, including the distribution of the plants, was represented as a single composite plant with a head
WYTs on which they depend, to represent desired future condi- of 957 m and efficiency of 100% (Rheinheimer et al. 2013). Hydro-
tions, even if these may not be ecologically ideal and could change power operations were represented with a monthly version of the
in future negotiations. Given these assumptions and implications, weekly time step, revenue-maximizing, mixed integer linear pro-
this paper asks the following questions: gramming (MILP) optimization model of the upper Yuba River

© ASCE 04016049-2 J. Water Resour. Plann. Manage.

J. Water Resour. Plann. Manage., 04016049


Downloaded from ascelibrary.org by David Rheinheimer on 07/12/16. Copyright ASCE. For personal use only; all rights reserved.

Fig. 1. Upper Yuba basin and its water diversion scheme (reprinted from Rheinheimer et al. 2013, Copyright © 2012 John Wiley & Sons, Ltd.)

system described by Rheinheimer et al. (2013). The objective of the District 2006; Pacific Gas and Electric Company 2007). The MILP
MILP model was to maximize revenue from hydropower model was implemented using the General Algebraic Modeling
generation. The MILP model was modified to run using a rolling System (GAMS) and the CBC optimization solver.
12-month period at the beginning of each month, allowing the pos- To mimic imperfect foresight, in each 12-month optimization
sibility of conditional updating of operational rules and imperfect period future inflows were estimated from modeled monthly snow
foresight. accumulation and runoff data. Snow accumulation and runoff were
A deregulated, market-driven electricity system was assumed, from a hydrologic model developed for the western Sierra Nevada
whereby the hydropower operator responds to wholesale electric- with the Water Evaluation and Planning System (WEAP) (Yates
ity prices to make release decisions, with greater releases when et al. 2005; Young et al. 2009), modified and recalibrated for the
electricity prices are higher. To maximize hydropower generation, daily time step with PEST (Doherty 2010) using reconstructed his-
the MILP model used linearized electricity price nonexceedance torical flows from the system licensee (DTA Sacramento 2007).
curves (Madani and Lund 2009; Olivares and Lund 2012; Calibration was from 1984 to 1988 and validation was from
Rheinheimer et al. 2013). Historical hourly day ahead market 1989 to 1999, both of which included a range of year types. Cal-
prices for a single year from the California Independent System ibration resulted in monthly Nash-Sutcliffe efficiency (Nash and
Operator were used, with year 2007 selected for its relative stability Sutcliffe 1970) of 0.67 and percent bias (Moriasi et al. 2007) of
compared to other years with available price data (Rheinheimer −4.7% for the Yuba River. From October through January, year-
et al. 2013). to-go (through September) monthly inflows were from median
System infrastructure characteristics, such as reservoir and diver- monthly runoff for the period-of-record (which may be updated,
sion channel capacities, were from publicly available documents, depending on the updating scheme described later). From February
particularly FERC relicensing documents (Nevada Irrigation through June, year-to-go median monthly inflows were scaled

© ASCE 04016049-3 J. Water Resour. Plann. Manage.

J. Water Resour. Plann. Manage., 04016049


Instream flow requirement (m 3/s) 7000

Upper threshold (mcm)


3.0
6000 >3997
2.5 3997
2.0 5000
4000 2701
1.5
1.0 3000 1801
1110
0.5 2000 759
0.0 1000
0
ECD CD D BN AN W
(a) E ECD CD D BN AN W (b) Water Year Type

Fig. 2. Current (a) water year type-based instream flow requirements in the South Fork Yuba River; (b) upper annual runoff thresholds (million cubic
Downloaded from ascelibrary.org by David Rheinheimer on 07/12/16. Copyright ASCE. For personal use only; all rights reserved.

meters, mcm) used to define water year types; water year types (WYTs) include emergency (E), extreme critically dry (ECD), critically dry (CD), dry
(D), below normal (BN), above normal (AN), and wet (W); emergency conditions are based on more than WYT, so are omitted from panel (b)

based on a regression between modeled snow accumulation and distribution of WYTs. Representative observed historical distribu-
modeled year-to-go inflow. In August and September, perfect tions were from the hydrology model forced with the meteorologi-
year-to-go foresight was assumed. In all cases, median monthly cal dataset from Livneh et al. (2015) to span a longer period of
runoff was used for the following water year (i.e., after September). record (1950–2009) than the dataset used in calibration.

Management Schemes
Climate Projections and Time Periods
Three WYT-based IFR schemes were considered. The first scheme
The hydrology model was forced with meteorological data (air
entailed no adaptation. The second two adaptively redefined the
temperature, precipitation, and relative humidity) downscaled
WYT definition thresholds to maintain the historical WYT distri-
from general circulation models (GCMs) from the Coupled Model
bution to satisfy environmental objectives, as discussed earlier,
Intercomparison Project Phase 5 (CMIP5) (Knutti and Sedláček
based on changing hydroclimatic knowledge. The schemes were
as follows: 2013; Taylor et al. 2012). Downscaled meteorological data based
1. Base-case scheme (BCS): This consisted of the existing WYT on these GCMs was obtained from the CMIP5 hydroclimate a
scheme, calibrated to ensure equal WYT distributions across hy- rchive (Brekke et al. 2013), and prepared using bias-corrected con-
drologic datasets during the historical period (1950–2010). structed analogues (Maurer et al. 2010). Four GCMs forced with
WYT definitions are never updated; representative concentration pathways of þ8.5 W=m2 (RCP8.5)
2. Adaptive, with historical WYT distribution (AHD): In this were selected for representation of regional meteorology and range
scheme, WYT thresholds are updated every 5 years with a roll- of conditions (Table 1). All GCMs except the drier-future (MIROC-
ing, retrospective 50-year period of record to maintain the his- ESM) were selected specifically for El Niño-Southern Oscillation
torical distribution of WYTs. This scheme thereby incorporates (ENSO) representation (Rupp et al. 2013), a dominant hydrocli-
the latest available hydrologic information into the IFR scheme matic driver in California. Climate change projections through
by regularly re-defining a retrospective new normal; and 2099 anticipate a steady increase in regional temperatures, yet
3. Adaptive, with historical WYT distribution and future hydrol- reveal great uncertainty about future precipitation; projected
ogy (AHDF): This scheme is the same as AHD, yet with the 2050–2099 mean temperatures and precipitation are about þ2 to
50-year rolling period of record spanning both past (30 years) 3°C and −8.5 to þ35.5%, respectively, compared to the projected
and future (20 years) periods to better represent current climate historical (1950–2009) means (Table 1).
conditions. The latter time frame is suggested by Wilby et al. Four multidecadal time periods are considered for analysis:
(2009) for both the greater certainty in climatic trends and historical (1950–2009), near term (2010–2039), medium term
the length of typical planning horizons. (2040–2069), and far term (2070–2099). As the most optimistic
Because the hydrologic time series for each climate scenario adaptation management scheme relies on future data, the last
is unique, even during the historical period, the historical time period (far term) is omitted from analyses of management
WYT thresholds were adjusted to match the observed historical schemes.

Table 1. World Climate Research Programme (CMIP5) Projections Used in This Study with Mean 2050–2099 Change in Yuba River Basin Precipitation and
Temperature, Compared to 1950–2009
WCRP CMIP5 climate 2050–2099 temperature 2050–2099 precipitation
model (GCM) identifier CMIP5 projection change (°C) change (%)
MIROC-ESM.1 MIROC-ESM.1.RCP85 þ2.0 −8.6 (drier)
MIROC5.1 MIROC5.1.RCP85 þ2.8 −1.1 (slightly drier)
CCSM4.1 CCSM4.1.RCP85 þ2.8 þ3.4 (slightly wetter)
CNRM-CM5.1 CNRM-CM5.1.RCP85 þ3.0 þ35.8 (much wetter)
Note: GCM version numbers are omitted in this study; WCRP stands for World Climate Research Program.

© ASCE 04016049-4 J. Water Resour. Plann. Manage.

J. Water Resour. Plann. Manage., 04016049


60%

1950−
40

2009
ΔRunoff (%) 45% 20
0

Percent of years (%)


30% WYT

2010−
40

2039
ECD
15% 20
CD
0 D
0% BN

2040−
40

2069
AN
-15% 20
(a) W
MIROC-ESM MIROC5 CCSM4 CNRM-CM5 0

2070−
40

2099
20% 20
15% 0
MIROC−ESM MIROC5 CCSM4 CNRM−CM5
Downloaded from ascelibrary.org by David Rheinheimer on 07/12/16. Copyright ASCE. For personal use only; all rights reserved.

10% GCM
ΔIFR (%)

5% Fig. 4. Distribution of water year types (ECD = extreme critically


0% dry; CD = critical dry; D = dry; BN = below normal; AN = above
normal; W = wet) under the environmental flow allocation scheme
-5% by the U.S. Forest Service for different periods and climate
-10% scenarios; dashed lines represent the observed historical period
(b) MIROC-ESM MIROC5 CCSM4 CNRM-CM5 (1950–2009)

2.0
not account for changes in actual instream flows, which includes
1.0 spill (potentially undesirable).
ΔIFR / ΔRunoff

Climate also affects the distribution of WYTs, shown in Fig. 4


0.0
for different time periods and climate scenarios along with an ob-
-1.0
served historical dataset. The historical period shows WYT fre-
quency distributions that are generally consistent between
-2.0 GCMs and the observed historical dataset. In contrast, future peri-
ods show WYT frequency distributions that are much different
-3.0 from historical. Any major changes in WYT frequency distribution
(c) MIROC-ESM MIROC5 CCSM4 CNRM-CM5
will alter annual IFRs (Fig. 5), with potential cascading ecological
effects downstream. Thus, static WYT definitions may have eco-
Fig. 3. (a) Relative change in Yuba River runoff; (b) relative change in
logical consequences inconsistent with the original intent of the
annual IFR in the SF Yuba River; (c) relative change in annual IFR per
IFR scheme if used with a nonstationary hydroclimate. Further-
change in runoff compared to historical (1950–2009) for selected
GCMs; y-axis scale changes between figures
more, future periods show significant differences between GCMs
compared to the historical period, indicating great uncertainty of
future IFR distributions. The combination of greater potential
for ecologically harmful outcomes and greater uncertainty implies
a need to assess whether current management regimes lead to po-
Hydroclimate Effects on WYT and Base-Case tential changes that are within ecologically acceptable limits, assess
IFR Scheme how the potential burden of climate change is shared with compet-
ing water users, and identify potential adaptation mechanisms if the
The range of climate projections selected (Table 1) result in changing exposure to ecological and economic risk is found to be
uncertain and variable multidecadal changes in Yuba River unacceptable. The following section focuses on burden-sharing and
runoff [Fig. 3(a)]. End-of-century runoff changes range from the potential of adaptation.
about −11 to þ50% compared to 1950–2009 averages within WYT-based schemes in the region are also spatially nonadap-
individual projections, revealing future runoff uncertainty consis- tive, despite interdependence with other water management oper-
tent with precipitation uncertainty between the selected climate ations. The current WYT scheme in the Yuba River is defined
scenarios. independently from the downstream four-river Sacramento Valley
Changes in Yuba River runoff result in somewhat commensurate Index (SVI), a key water allocation tool for water deliveries through
changes to average annual dedicated IFR allocations in the SFY the Sacramento–San Joaquin Delta that incorporates estimated
[Fig. 3(b)]. Changes to IFRs are a function of the IFR values monthly unimpaired Yuba River outflows, resulting in discontinu-
and WYT thresholds (Fig. 2). As WYT distributions change (dis- ities in key management indicators (anticipated annual runoff) de-
cussed later), environmental allocations do not change proportion- spite the hydraulic upstream/downstream relationship between the
ally. This is reflected in Fig. 3(c), which shows the change in Yuba Yuba River and the Sacramento-San Joaquin Delta. This is shown
River runoff (modeled unimpaired flow) per change in dedicated in Fig. 6, which maps historical (1906–2013) WYTs between YRI
IFR allocation. Changes in dedicated IFR allocations are always and SVI and shows that designations between water year types are
less than one, indicating that the current IFR scheme provides some incongruent if not arbitrary. This categorical mismatch will likely
buffer against climate change on a mean annual basis. However, the change as the regional hydroclimate changes. As with hydrologic
interest in this study was the statistical distribution of IFR alloca- changes within the Yuba River itself, this type of incongruous
tions, not simply mean annual allocations. Furthermore, this does WYT formulation and designation also implies a need to assess

© ASCE 04016049-5 J. Water Resour. Plann. Manage.

J. Water Resour. Plann. Manage., 04016049


1950−2009 2010−2039 2040−2069 2070−2099
1.00 GCM

Non−exceedance
MIROC-ESM

probability
0.75
MIROC5
0.50
CCSM4
0.25
CNRM−CM5
0.00
1950−2009 mean
0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50
IFR (mcm/year)

Fig. 5. Nonexceedance probability for annual IFRs in the South Fork Yuba River for different periods and GCMs; dashed line represents the mean of
the historical period (1950–2009)
Downloaded from ascelibrary.org by David Rheinheimer on 07/12/16. Copyright ASCE. For personal use only; all rights reserved.

distributions of WYTs, whereas Fig. 8 shows the nonexceedance


W
distributions of actual instream releases. The base-case, nonadap-
YRI Water Year Type

SVI Water Year Type


W
tive scheme (BCS) shows distinct differences in both WYT changes
AN and the distributions of actual releases between GCMs in future
AN
periods. Better adaptation (AHDF) shows a substantial decrease
BN BN in difference in specific WYT occurrence from the historical,
base-case WYT distributions. Under the less-adaptive scheme
D D (AHD), this difference in WYT distribution is not readily discern-
CD CD ible. Effects of adaptation are more apparent with actual
instream flows. As the IFR scheme becomes more adaptive
Fig. 6. Sankey plot mapping historical occurrences of Sacramento (AHD and AHDF), the difference between GCMs becomes less
Valley Index (SVI) and Yuba River Index (YRI) water year types pronounced in future periods. However, by end-of-century (2079–
(CD = critical dry; D = dry; BN = below normal; AN = above normal; 2099), adaptation can only partially accommodate hydroclimatic
and W = wet); extreme critical dry YRI water year types have been change, as indicated by the larger distribution differences between
reclassified as critical dry years to match SVI water year types GCMs.
The greater similarity between GCMs for future time periods
with greater adaptation (Fig. 8) is more pronounced in low-flow
whether or not static approaches to water management, either lo- years and less wet GCMs. This is due to algorithmic operating rules
cally or regionally, are ecologically and economically acceptable that ensure minimum flows are met under any instream flow re-
and, if not, to identify adaptation measures that take into account quirement scheme in drier years. This IFR allocation is particularly
regional hydrologic connectivity. apparent during the driest years in the medium-term period, when
both modest and aggressive adaptation (AHD and AHDF, respec-
tively) nearly eliminate differences in flows for the same nonexcee-
Water Allocations with Adaptive Management dance probabilities. In contrast, infrastructure and limited seasonal
Schemes foresight can limit the ability to reduce high-flow releases in wetter
years, resulting in greater spill and potentially ecologically unde-
To quantify water allocations, the difference between controlled sirable changes to annual environmental allocations. This spill is
and uncontrolled instream flow is noted. Prior to the 21st Century, apparent with CNRM-CM5, the wettest of the GCMs, in most fu-
any flow in the river was considered environmentally beneficial. ture periods. Adjustments in WYT definitions by the most adaptive
More recently however, instream flow targets have been designed management scheme (AHDF) can accommodate some, but not all,
for a particular flow regime, mimicking the timing, duration, of the higher annual flows expected. The limitations of adaptation,
magnitude, frequency and rate of change of flows observed in at least under existing infrastructure constraints (medium-size res-
unregulated hydrology. In the Sierra Nevada, uncontrolled, high ervoir with no flood control capacity), is generally apparent in all of
magnitude flows (i.e., spill) can be detrimental to ecosystems the highest-flow years, where even the most anticipatory manage-
(Yarnell et al. 2010), even if more water overall is available to ment scheme can do little to maintain a historical distribution of
the ecosystem. This is accounted for by examining both managed environmental allocations. However, it also indicates that improved
IFRs and actual instream flows (AIFs), where the latter includes foresight can provide for better overall management of water for
spill, such that IFRs are often much less than AIFs. A “functional multiple, and often competing, water demands.
flow” regime would account for this discrepancy and include a
range of beneficial flow types, including large magnitude peak
Water Allocations between Competing Water Users
flows (Yarnell et al. 2015).
Relative changes in IFR allocations in the SFY and hydropower
releases to the Bear River basin compared to the historical time
Adaptive Management and Environmental Water
period are shown in Fig. 9. With the base-case management scheme
Distributions
(BCS), mean annual IFR allocations are more sensitive than hydro-
The allocation of water to different users is sensitive to changing power releases to climate change over time. Thus, with a drier cli-
climate-driven hydrology and adaptation of water management mate (MIROC-ESM), relative reductions in IFR allocations in the
frameworks. For different combinations of climate scenarios and future are greater than for hydropower allocations. Conversely, with
management schemes, Fig. 7 shows the differences in percentage a wetter climate (CNRM-CM5), relative IFR allocation increases

© ASCE 04016049-6 J. Water Resour. Plann. Manage.

J. Water Resour. Plann. Manage., 04016049


BCS AHD AHDF
30
20

1950−
2009
10
Change in percent of years (magnitude %) 0
−10
−20
30
20

2010−
2039
10
0
WYT
−10 ECD
−20 CD
30 D
20 BN

2040−
AN

2069
10
0 W
−10
Downloaded from ascelibrary.org by David Rheinheimer on 07/12/16. Copyright ASCE. For personal use only; all rights reserved.

−20
30
20

2070−
2099
10
0
−10
−20
4

5
SM

SM

SM

5
SM

SM

SM

M
C

C
O

O
−C

−C

−C
−E

−E

−E
IR

IR

IR

C
M

M
C

C
C

C
M

M
O

O
R

R
N
IR

IR

IR

N
C

C
M

M
GCM
Fig. 7. Change in distribution of water year types (ECD = extreme critical dry; CD = critical dry; D = dry; BN = below normal; AN = above normal;
and W = wet) for different periods, climate scenarios, and adaptation schemes compared to historical, base-case management (BCS); results are
insufficient for the 2070–2099 period under the AHDF scheme

1950−2009 2010−2039 2040−2069 2070−2099


1.00
0.75
0.50 BCS
Non−exceedance probability

0.25
0.00
GCM
1.00
MIROC-ESM
0.75
MIROC5
AHD

0.50
0.25 CCSM4
0.00
CNRM−CM5
1.00
1950−2009 mean
0.75
AHDF

0.50
0.25
0.00
20 40 60 80 20 40 60 80 20 40 60 80 20 40 60 80

Fig. 8. Annual nonexceedance probabilities (empirical cumulative distributions) of actual instream flow in the South Fork Yuba River for different
periods, adaptive schemes, and GCMs; results are insufficient for the 2070–2099 period under the AHDF scheme; dashed line represents the mean of
the historical period (1950–2009) for base-case management (BCS), assumed more ecologically desirable

are mostly greater than for hydropower. In between, IFR allocations counterintuitive. These differences can be explained by the chang-
tend to increase and change is greater than for hydropower, with ing flow regime with each of these climate scenarios. Although run-
powerhouse releases generally decreasing compared to the histori- off volumes change over time, so does the timing of runoff, to
cal period. Actual instream flows, which include IFRs and spill, are which both IFRs and powerhouse flows are also sensitive. Climate
more sensitive to climate than IFRs alone (data not shown). warming reduces snowpack and causes earlier runoff, which, in
The increase in IFR allocations for the slightly drier GCM turn, results in higher spill (data not shown) and suboptimal hydro-
(MIROC5) and decrease in hydropower allocations in the slightly power generation. This phenomenon is opposite from a previous
wetter GCM (CCSM4) with base-case operations are somewhat study of the same system (Rheinheimer et al. 2013), which showed

© ASCE 04016049-7 J. Water Resour. Plann. Manage.

J. Water Resour. Plann. Manage., 04016049


MIROC−ESM MIROC5 CCSM4 CNRM−CM5

Difference from historical (%)


Period of record
BCS
10
AHD
AHDF
0
Release type
PH
−10
IFR
1950− 2010− 2040− 2070− 1950− 2010− 2040− 2070− 1950− 2010− 2040− 2070− 1950− 2010− 2040− 2070−
2009 2039 2069 2099 2009 2039 2069 2099 2009 2039 2069 2099 2009 2039 2069 2099
Period

Fig. 9. Change in powerhouse release (PH) and instream flow requirement (IFR) under different management scenarios compared to the historical
period (1950–2009)
Downloaded from ascelibrary.org by David Rheinheimer on 07/12/16. Copyright ASCE. For personal use only; all rights reserved.

a general benefit for hydropower generation to changing runoff management schemes. In particular, if maintenance of historical
timing. However, that study assumed perfect hydrologic foresight, distributions of annual allocations are desired, water year types
which was not the case here; other studies with imperfect foresight used to achieve this can, at the least, be regularly updated with retro-
have shown decreases in hydropower generation with warming spective information (AHD management scheme) to account for
and little change in precipitation in the Yuba River region generally, changing distributions of the inflow hydrology that define the water
including this system (Mehta et al. 2011; Rheinheimer et al. year types. This is already being considered in some instances. For
2014). example, in the western United States, outside of California, the
Updating WYT definitions with a rolling period of record U.S. Bureau of Reclamation has recently started using a 30-year
diminished the sensitivity of IFR allocation changes to climate moving period of record, updated every 10 years, to compare
change and increased the sensitivity of powerhouse releases, also water year types, using percentiles, and there are even concerns that
shown in Fig. 9. Both adaptive management schemes (AHD and variability is increasing to the point of rendering 30 years too
AHDF) result in increasing future IFR releases with the dry GCM long a period (B. McInerney, National Weather Service/Colorado
and decreasing releases with the wet GCM compared to base-case River Forecasting Center, personal communication, 2013). Further
management. For example, mean IFR allocations with a wetter IFR management benefits arise from better characterizing future
climate (CNRM-CM5) are nearly the same in 2040–2069 as in meteorological conditions and integrating that knowledge into
1950–2009 with the most anticipatory management scheme decision-making (AHDF). This might include not only the 20-year
(AHDF). This is by design, and, along with Fig. 8, demonstrates outlook considered in this study and suggested by Wilby et al.
that the adaptive approach has the desired effect of reducing— (2009), but also probabilistic seasonal weather predictions (Towler
though not eliminating—IFR variability from climate change. et al. 2013). However, adaptive approaches to flow release sched-
ules are at regulatory odds with the long-term licensing procedures
practiced by FERC (Viers 2011), which issues licenses based
Discussion on static WYT and IFR definitions. Thus, any adoption will likely
require policy changes as well.
With the historical approach to environmental water management, Although the adaptation mechanism used here—a rolling
it would appear that instream flows disproportionately benefit N-year period of record—is useful, other, complementary ap-
from warming compared to hydropower. However, this is not nec- proaches are needed to better understand the risks to environmental
essarily the case, due to the changed flow regime and higher spill and human water management associated with climate change and
with warming, as noted earlier, which has been a recurring theme in possible adaptation options. In particular, quantitative, risk-based
climate change and environmental flows research in the Sierra decision-making frameworks are needed for improving long-term
Nevada (Rheinheimer and Viers 2015). With most climate scenar- aquatic ecosystem management (Borgomeo et al. 2014; Johnson
ios, this is a lose–lose outcome from climate change, as hydro- and Weaver 2008). Although climate impact assessments have been
power releases are reduced considerably, but aquatic ecosystems mostly top-down, whereby climate projections are downscaled and
are increasingly impacted by changing flows, despite increased used in impact studies, proposed risk-based decision frameworks
allocated IFRs and actual streamflow, because they are less likely tend to be bottom-up, so that local vulnerability thresholds are
to mimic the historical hydrology for which local biota are adapted mapped to climate projection ranges (Brown et al. 2012; Singh et al.
(Fig. 8). Other studies have shown that these volumetric impacts to 2014; Wilby and Dessai 2010). This enables uncertainty to be built
aquatic ecosystems are exacerbated by increasing stream temper- into decision-making as robust decision frameworks (Weaver et al.
atures in both regulated and unregulated systems (Null et al. 2013). Bottom-up approaches recognize that climate models are
2013a, b). collectively incapable of predicting far-term future climates, but in-
This paper shows that adapting WYT frameworks for future hy- stead can provide ranges of future climates (Stainforth et al. 2007).
droclimates maintains instream flow targets better than nonadaptive Such risk-based analyses can help identify if/when hydroclimatic
frameworks. The approach in this study, conducted with GCMs that changes might overwhelm capacity—infrastructural or otherwise—
represent the anticipated range of future conditions, was developed to absorb change, enabling planners to develop appropriate correc-
to highlight uncertainty—making for more complex results, but tive measures. In the context of the present study, a risk-based
also mirroring real-world uncertainty. The above analyses highlight approach could help identify how aggressively a WYT-based
the need to explicitly account for potential and uncertain changes IFR scheme needs to be improved to reduce long-term risk to
in regional hydroclimates when designing long-term water high-priority water users.

© ASCE 04016049-8 J. Water Resour. Plann. Manage.

J. Water Resour. Plann. Manage., 04016049


The WYT-based approach might also be modified further as definitions, which is a relatively easy/straightforward adaptation
needed to ensure adequate environmental protections and minimal strategy. However, this approach needs to be complemented with
economic disruptions with climate change. Advances in climate more research to better quantify the real risks of climate change on
forecasting and real-time management decision frameworks might river ecosystems to identify how aggressive adaptation should be.
enable moving from monthly to daily instream flow decision mak- Water management frameworks should be both adaptive and
ing, as is already done for hydropower and other human uses. This anticipatory, as well as regionally concordant. Given current ana-
could enable, for example, semiopportunistic releases of short-term lytic capabilities, water planning and management can be discussed
flood pulses coordinated with electricity demand forecasts to with awareness of at least some of the long-term implications of a
minimize hydropower revenue losses. Water managers might also given management scheme, even with uncertainty about future hy-
consider using a statistics-based method for discretizing water droclimatic conditions and ecosystem responses. This necessitates
availability for index-based allocations to reduce the ambiguity that quantifying known uncertainties, yet also allowing for flexibility to
may arise from regularly changing human-centric classifications. accommodate unknown uncertainty.
Terms such as wet and dry become less meaningful if water year
type thresholds are regularly changed as hydrologic conditions
Downloaded from ascelibrary.org by David Rheinheimer on 07/12/16. Copyright ASCE. For personal use only; all rights reserved.

change. Instead, water years might be classified as flow percentiles Acknowledgments


(e.g., quintiles) for greater semantic objectivity and transparency.
Finally, while a river’s flow regime is a master variable control- The authors gratefully acknowledge Guido Franco and the
ling other river characteristics like water quality, biogeochemical California Energy Commission for funding portions of this re-
processes, and sediment transport (Poff et al. 1997), the flow re- search (CEC# 500-10-030), as well as the University of California
gime itself accounts for only a portion of freshwater ecosystem Center for Information Technology Research in the Interest of
variability, with other key variables that include temperature, geo- Society (CITRIS).
morphology, etc. (Worrall et al. 2014). Bulk water allocation tools
such as water year types potentially miss other key water variables,
necessitating the development a broader suite of tools to adapt References
river management to nonstationary environmental conditions, such
as the screening tool approach advocated by Grantham et al. (2014) Borgomeo, E., Hall, J. W., Fung, F., Watts, G., Colquhoun, K., and
to evaluate dams for improved environmental flow releases. Lambert, C. (2014). “Risk-based water resources planning: Incorporat-
ing probabilistic nonstationary climate uncertainties.” Water Resour.
Res., 50(8), 6850–6873.
Brekke, L., Thrasher, B. L., Maurer, E. P., and Pruitt, T. (2013). “Down-
Conclusions scaled CMIP3 and CMIP5 climate and hydrology projections: Release
of downscaled CMIP5 climate projections, comparison with preceding
This research acknowledges two broad challenges confronting information, and summary of user needs.” U.S. Dept. of the Interior,
river ecosystem managers as transition is made to nonstationary Bureau of Reclamation, Technical Services Center, Denver.
hydrologic regimes with long-term uncertainty: whether or not Brown, C., Ghile, Y., Laverty, M., and Li, K. (2012). “Decision scaling:
to adapt management practices as hydroclimatic changes are slowly Linking bottom-up vulnerability analysis with climate projections in
realized, and whether current management practices such as WYT- the water sector.” Water Resour. Res., 48(9), W09537.
based instream flow requirements are sufficient given management CBC v2.8 [Computer software]. COIN-OR Foundation, Towson, MD.
objectives, or whether new approaches are needed. Specifically, Doherty, J. (2010). “PEST: Model-independent parameter estimation user
this study examined the implications of assuming a stationary hy- manual.” Watermark Numerical Computing, Brisbane, Australia.
droclimate in a WYT-based IFR scheme and assessed the utility of DTA Sacramento. (2007). “Modeling and evaluation report: Synthesis
using a rolling period of hydrologic record to regularly redefine of unimpaired hydrology [preliminary draft].” Devine Tarbell &
water year types. Associates, Sacramento, CA.
Fung, F., Watts, G., Lopez, A., Orr, H. G., New, M., and Extence, C. (2013).
In the case of the upper Yuba River hydropower system in the
“Using large climate ensembles to plan for the hydrological impact
Sierra Nevada, where climate change will greatly affect hydrology,
of climate change in the freshwater environment.” Water Resour.
the utility of water year types for key water management decisions Manage., 27(4), 1063–1084.
is diminished by long-term hydroclimatic change. If long-term GAMS v24.0 [Computer software]. GAMS Development Corporation,
management objectives are to maintain historical and predictable Washington, DC.
operations for both environmental and human water uses, it is Georgakakos, A. P. (2012). “Value of adaptive water resources management
apparent from this study and others that these objectives are not in Northern California under climatic variability and change: Reservoir
maintained. The utility of historically-defined WYTs is further management.” J. Hydrol., 412–413, 34–46.
weakened by the uncoordinated definition and use of WYTs re- Grantham, T. E., Viers, J. H., and Moyle, P. B. (2014). “Systematic screen-
gionally, which confounds water management at local and regional ing of dams for environmental flow assessment and implementation.”
scales. Spatially variable hydrologic and management changes BioScience, 64(11), 1006–1018.
would possibly exacerbate those incongruities between and across Heino, J., Virkkala, R., and Toivonen, H. (2009). “Climate change and
basins. freshwater biodiversity: Detected patterns, future trends and adaptations
With historical, nonadaptive rules, changes in mean runoff were in northern regions.” Biol. Rev., 84(1), 39–54.
Jiménez Cisneros, B. E., et al. (2014). “Freshwater resources.” Climate
mostly equitably shared between hydropower and environmental
Change 2014: Impacts, Adaptation, and Vulnerability. Part A: Global
releases on a mean annual basis, though hydropower benefited
and Sectoral Aspects. Contribution of Working Group II to the 5th As-
more from increased runoff. However, this outcome is not neces- sessment Rep. of the Intergovernmental Panel of Climate Change, C. B.
sarily desirable, as ecological conditions are maintained by the Field, et al., eds., Cambridge University Press, Cambridge, U.K.,
entire flow regime, not annual runoff magnitude. 229–269.
To maintain historical instream flow requirements using the his- Johnson, T. E., and Weaver, C. P. (2008). “A framework for assessing
torical WYT-based IFR scheme, this study and Null and Viers climate change impacts on water and watershed systems.” Environ.
(2013) clearly demonstrate the efficacy of regularly updating WYT Manage., 43(1), 118–134.

© ASCE 04016049-9 J. Water Resour. Plann. Manage.

J. Water Resour. Plann. Manage., 04016049


Knutti, R., and Sedláček, J. (2013). “Robustness and uncertainties in change combinations for critical hydrologic indicator thresholds:
the new CMIP5 climate model projections.” Nat. Clim. Change, 3(4), Application to a watershed in Pennsylvania, U.S.” Water Resour.
369–373. Res., 50(4), 3409–3427.
Livneh, B., et al. (2015). “A spatially comprehensive, hydrometeorological Stainforth, D. A., Downing, T. E., Washington, R., Lopez, A., and New, M.
data set for Mexico, the U.S., and southern Canada 1950–2013.” Sci. (2007). “Issues in the interpretation of climate model ensembles to in-
Data, 2, in press. form decisions.” Philos. Trans. R. Soc. London A: Math. Phys. Eng.
Madani, K., and Lund, J. R. (2009). “Modeling California’s high-elevation Sci., 365(1857), 2163–2177.
hydropower systems in energy units.” Water Resour. Res., 45(9), Stakhiv, E. Z. (1998). “Policy implications of climate change impacts on
W09413. water resources management.” Water Policy, 1(2), 159–175.
Maurer, E. P., Hidalgo, H. G., Das, T., Dettinger, M. D., and Cayan, D. R. Strayer, D. L., and Dudgeon, D. (2010). “Freshwater biodiversity conser-
(2010). “The utility of daily large-scale climate data in the assessment of vation: Recent progress and future challenges.” J. North Am. Benthol.
climate change impacts on daily streamflow in California.” Hydrol. Soc., 29(1), 344–358.
Earth Syst. Sci., 14(6), 1125–1138. SWRCB (California State Water Resources Control Board). (1999).
Mehta, V. K., et al. (2011). “Potential impacts on hydrology and hydro- “Water right decision 1641 (revised).” 〈http://www.waterboards.ca.gov/
power production under climate warming of the Sierra Nevada.” J. waterrights/board_decisions/adopted_orders/decisions/d1600_d1649/
Water Clim. Change, 2(1), 29–43.
Downloaded from ascelibrary.org by David Rheinheimer on 07/12/16. Copyright ASCE. For personal use only; all rights reserved.

wrd1641_1999dec29.pdf〉.
Milly, P. C. D., et al. (2008). “Stationarity is dead: Whither water manage- Taylor, K. E., Stouffer, R. J., and Meehl, G. A. (2012). “An overview of
ment?” Science, 319(5863), 573–574. CMIP5 and the experiment design.” Bull. Am. Meteorol. Soc., 93(4),
Moriasi, D. N., Arnold, J. G., Van Liew, M. W., Bingner, R. L., Harmel, 485–498.
R. D., and Veith, T. L. (2007). “Model evaluation guidelines for Towler, E., Roberts, M., Rajagopalan, B., and Sojda, R. S. (2013). “Incor-
systematic quantification of accuracy in watershed simulations.” Trans. porating probabilistic seasonal climate forecasts into river management
ASABE, 50(3), 885–900. using a risk-based framework.” Water Resour. Res., 49(8), 4997–5008.
Nash, J. E., and Sutcliffe, J. V. (1970). “River flow forecasting through U.S. Fish and Wildlife Service. (2008). “Formal endangered species act
conceptual models: Part I—A discussion of principles.” J. Hydrol., consultation on the proposed coordinated operations of the central
10(3), 282–290. valley project (CVP) and state water project (SWP).” Rep. No.
Nevada Irrigation District. (2006). “Yuba-Bear hydroelectric project.” 81420-2008-F-1481-5, U.S. Fish and Wildlife Service, Sacramento,
FERC Project No. 2266: Annotated Existing License, Marysville, CA. CA.
Null, S. E., Ligare, S. T., and Viers, J. H. (2013a). “A method to consider U.S. Forest Service. (2012). “Revised FS preliminary conditions and rec-
whether dams mitigate climate change effects on stream temperatures.” ommendations provided under 18 CFR § 4.34 (b)(1) in connection with
J. Am. Water Resour. Assoc., 49(6), 1456–1472. the application for relicensing for the Drum-Spaulding hydroelectric
Null, S. E., and Viers, J. H. (2013). “In bad waters: Water year classification project (FERC No. 2310).” Washington, DC.
in nonstationary climates.” Water Resour. Res., 49(2), 1137–1148.
Viers, J. H. (2011). “Hydropower relicensing and climate change.” J. Am.
Null, S. E., Viers, J. H., Deas, M. L., Tanaka, S. K., and Mount, J. F.
Water Resour. Assoc., 47(4), 655–661.
(2013b). “Stream temperature sensitivity to climate warming in
Vörösmarty, C. J., et al. (2010). “Global threats to human water security
California’s Sierra Nevada: Impacts to coldwater habitat.” Clim.
and river biodiversity.” Nature, 467(7315), 555–561.
Change, 116(1), 149–170.
WEAP v3.4 [Computer software]. Stockholm Environment Institute U.S.
Olivares, M., and Lund, J. (2012). “Representing energy price variability in
Center, Somerville, MA.
long- and medium-term hydropower optimization.” J. Water Resour.
Weaver, C. P., Lempert, R. J., Brown, C., Hall, J. A., Revell, D., and
Plann. Manage., 10.1061/(ASCE)WR.1943-5452.0000214, 606–613.
Sarewitz, D. (2013). “Improving the contribution of climate model
Pacific Gas and Electric Company. (2007). “Drum-spaulding project.”
information to decision making: The value and demands of robust
FERC Project No. 2310: FERC License as of May 11, 2007,
decision frameworks.” Wiley Interdiscip. Rev.: Clim. Change, 4(1),
San Francisco.
39–60.
Palmer, M. A., et al. (2008). “Climate change and the world’s river basins:
Anticipating management options.” Front. Ecol. Environ., 6(2), 81–89. Wilby, R. L., et al. (2009). “A review of climate risk information for adap-
PEST v12.3 [Computer software]. Watermark Numerical Computing, tation and development planning.” Int. J. Climatol., 29(9), 1193–1215.
Brisbane, Australia. Wilby, R. L., and Dessai, S. (2010). “Robust adaptation to climate change.”
Petts, G. E. (2009). “Instream flow science for sustainable river manage- Weather, 65(7), 180–185.
ment.” J. Am. Water Resour. Assoc., 45(5), 1071–1086. Wilby, R. L., Fenn, C. R., Wood, P. J., Timlett, R., and LeQuesne, T. (2011).
Poff, N. L., et al. (1997). “The natural flow regime.” BioScience, 47(11), “Smart licensing and environmental flows: Modeling framework and
769–784. sensitivity testing.” Water Resour. Res., 47(12), W12524.
Poff, N. L., Olden, J. D., Merritt, D. M., and Pepin, D. M. (2007). “Homog- Worrall, T. P., Dunbar, M. J., Extence, C. A., Laizé, C. L. R., Monk, W. A.,
enization of regional river dynamics by dams and global biodiversity and Wood, P. J. (2014). “The identification of hydrological indices for
implications.” Proc. Natl. Acad. Sci., 104(14), 5732–5737. the characterization of macroinvertebrate community response to flow
Rheinheimer, D., Viers, J., Sieber, J., Kiparsky, M., Mehta, V., and Ligare, regime variability.” Hydrol. Sci. J., 59(3–4), 645–658.
S. (2014). “Simulating high-elevation hydropower with regional climate Xenopoulos, M. A., Lodge, D. M., Alcamo, J., Märker, M., Schulze, K.,
warming in the west slope, Sierra Nevada.” J. Water Resour. Plann. and Van Vuuren, D. P. (2005). “Scenarios of freshwater fish extinctions
Manage., 10.1061/(ASCE)WR.1943-5452.0000373, 714–723. from climate change and water withdrawal.” Global Change Biol.,
Rheinheimer, D. E., and Viers, J. H. (2015). “Combined effects of reservoir 11(10), 1557–1564.
operations and climate warming on the flow regime of hydropower by- Yarnell, S. M., et al. (2015). “Functional flows in modified riverscapes:
pass reaches of California’s Sierra Nevada.” River Res. Appl., 31(3), hydrographs, habitats and opportunities.” BioScience, 65(10), 963–972.
269–279. Yarnell, S. M., Viers, J. H., and Mount, J. F. (2010). “Ecology and
Rheinheimer, D. E., Yarnell, S. M., and Viers, J. H. (2013). “Hydropower management of the spring snowmelt recession.” BioScience, 60(2),
costs of environmental flows and climate warming in California’s upper 114–127.
Yuba River watershed.” River Res. Appl., 29(10), 1291–1305. Yates, D., Sieber, J., Purkey, D., and Huber-Lee, A. (2005). “WEAP21 - A
Rupp, D. E., Abatzoglou, J. T., Hegewisch, K. C., and Mote, P. W. (2013). demand-, priority-, and preference-driven water planning model:
“Evaluation of CMIP5 20th century climate simulations for the Pacific Part 1—Model characteristics.” Water Int., 30(4), 487–500.
Northwest U.S.” J. Geophys. Res.: Atmos., 118(19), 10884–10906. Young, C. A., et al. (2009). “Modeling the hydrology of climate change in
Singh, R., Wagener, T., Crane, R., Mann, M. E., and Ning, L. (2014). “A California’s Sierra Nevada for subwatershed scale adaptation.” J. Am.
vulnerability driven approach to identify adverse climate and land use Water Resour. Assoc., 45(6), 1409–1423.

© ASCE 04016049-10 J. Water Resour. Plann. Manage.

J. Water Resour. Plann. Manage., 04016049

You might also like