Organizing Environmental Flow Frameworks to Meet Hydropower Mitigation Needs

Abstract

The global recognition of the importance of natural flow regimes to sustain the ecological integrity of river systems has led to increased societal pressure on the hydropower industry to change plant operations to improve downstream aquatic ecosystems. However, a complete reinstatement of natural flow regimes is often unrealistic when balancing water needs for ecosystems, energy production, and other human uses. Thus, stakeholders must identify a prioritized subset of flow prescriptions that meet ecological objectives in light of realistic constraints. Yet, isolating aspects of flow regimes to restore downstream of hydropower facilities is among the greatest challenges of environmental flow science due, in part, to the sheer volume of available environmental flow tools in conjunction with complex negotiation-based regulatory procedures. Herein, we propose an organizational framework that structures information and existing flow paradigms into a staged process that assists stakeholders in implementing environmental flows for hydropower facilities. The framework identifies areas where regulations fall short of the needed scientific process, and provide suggestions for stakeholders to ameliorate those situations through advanced preparation. We highlight the strengths of existing flow paradigms in their application to hydropower settings and suggest when and where tools are most applicable. Our suggested framework increases the effectiveness and efficiency of the e-flow implementation process by rapidly establishing a knowledge base and decreasing uncertainty so more time can be devoted to filling knowledge gaps. Lastly, the framework provides the structure for a coordinated research agenda to further the science of environmental flows related to hydropower environments.

Appendix 1

E-flow tools and conceptual frameworks were categorized into four types following Annear et al. (2004): (1) Assessment, (2) Incremental, (3) Model Building, and (4) Information-Base strategies. Assessment methods make e-flow determinations on the basis of simple evaluations of ecohydrologic conditions. These include policy-driven evaluations to set limits or thresholds to determine appropriate flow regimes (Stalnaker 1995). Incremental methods are among the most time-intensive and analyze modeled river-stage ecological (biotic and abiotic) relations within a stream channel to compare alternative flow scenarios (Stalnaker et al. Stalnaker 1995; Annear et al. 2004). Model building includes complex mathematical routines that aid decision making, such as optimization modeling (Jager 2014) or Bayesian belief networks (Webb et al. 2015). Lastly, information-base strategies are among the most holistic frameworks that typically build knowledge bases and fill information gaps in a series of steps. These frameworks use the current state of knowledge, information compiled at regional scales, and quantitative relations between river flow and ecology to assess river conditions to reduce uncertainty in order to make e-flow recommendations. We describe commonly applied methodological approaches under each framework below.

Assessment

Two of the most common assessment techniques are the Range of Variability Approach (RVA) (Richhter et al. 1997) and the Tennant method. The RVA identifies the extent of hydrologic alteration from predisturbance conditions using 66 flow metrics (e.g., Indicators of Hydrologic Alteration). The RVA approach is dependent upon obtaining pre and post-disturbance discharge data to determine deviation from the system’s natural range of variation (Richhter et al. 1997). The convenience of RVA is that complex hydrologic behavior is dwindled into a summary of descriptive and informative statistics. When used in isolation, however, the RVA approach has no quantitative stage-channel relationships or support for in-stream ecological unless accompanied by stage-specific information (e.g., Nislow et al. 2002) or biological information (e.g., Taylor et al. 2014). Another assessment technique, the Tennant or Montana method, estimates habitat quality at various flows using limited field measurements, hydrologic records, and photographs of the stream channel (Tennant 1976). This method can be used as a reconnaissance-level tool for determining acceptable seasonally variable flow magnitudes in situations where there are little or no major competing uses (Annear et al. 2004).

Incremental Methods

Instream-Flow-Incremental-Methodologies (IFIM) are the most commonly applied techniques used to estimate e-flows (Tharme 2003) despite several shortcomings. IFIM approaches use field measurements of ecohydrologic conditions of the river channel at incremental discharges to model flow-ecology relationships. IFIM approaches can consider hydrology, biology, habitat, sediment transport, and water quality over a range of given discharges or under various flow regime alternatives (Bovee et al. 1998). The IFIM approach can range in complexity from describing simple relations between hydrologic indices and aquatic habitats to more complex hydrodynamic models linked to multiple river components (Tharme 2003). Although even the most complex IFIM application can be scientifically sound and provide assessments of management alternatives, the IFIM approach is: (1) only applicable to the specific study reach (Moir et al. 2005), (2) typically assumes that higher habitat suitability translates into a biological response (Anderson et al. 2006), and (3) often focuses only on individual species (Anderson et al. 2006).

Model Building

Mathematical models are often applied to balance water allocation among competing users, and also predict the ecological effects of modified flow regimes, particularly when there is substantial uncertainty in our estimates. Two examples of mathematical models are reservoir optimization algorithms and Bayesian belief networks (BBN). Reservoir optimization algorithms find optimal balances among multiple water demands (e.g., ecological and societal needs, Yeh 1985; Wurbs 1993). Algorithms can be relatively simple, or address more complex variables by incorporating stochasticity in forecasted inflows (Stedinger et al. 1985) or uncertainty in reservoir operations (Shresha et al. 1996). Reservoir optimization algorithms are highly useful for assessing the feasibility of flow alternatives, but still require a priori knowledge of the water needs for the ecosystem. BBN have increasingly been used to support water-management decisions (e.g., Hart and Pollino 2009; Stewart-Koster et al. 2010; Chan et al. 2012). BBN provide a simple graphical depiction of complex probabilistic reasoning about the relationship among important key variables. The key feature of BBN is that they allow us to model uncertainty in the relationships among variables. Other advantages of BBN include: (1) accommodating missing data and small sample size, (2) incorporating unorthodox data (e.g., expert opinion), and (3) being updated as new data are available (Korb and Nicholson 2004). More recent software packages (e.g., Netica, Hugin) have provided some solution to the limitations of these models such as the inability to include feedback loops between input and output variables, and the lack of time-dependent variables (Hart and Pollino 2009; Landuyt et al. 2013). These models have been used to aid the evaluation of many environmental flow issues (e.g., the effects of water withdrawals on fishes, Chan et al. 2012; the effectiveness of restoration strategies, Shenton et al. 2013).

Information-Base Approaches

Information-Base approaches including the Building Block Methodology (BBM, King and Louw 1998) and the Downstream Response to Imposed Flow Transformations (DRIFT) have been developed as conceptual alternatives to traditional e-flow techniques. These methods build on expert knowledge to support complex, stakeholder-driven management decisions and they emphasize monitoring post implementation. The BBM addresses all riverine ecosystem component needs (including societal) using existing knowledge and expert opinion in a structured workshop process. The DRIFT method builds upon the BBM and quantifies biophysical and sociological linkages to flow regimes and then evaluates biophysical, social, and economic responses under various flow scenarios (King et al. 2003). The strength of the DRIFT procedure is that stage-specific ecohydrologic assessments are conducted as opposed to relying only on predicted responses to flow.

The need for quantitative predictions to support e-flows led to the development of a process known as the Ecological Limits of Hydrologic Alteration (ELOHA) (Poff et al. 2010). ELOHA has been considered the most holistic e-flow framework to date (Richter et al. 2012) and has formally been applied in nine states of the US (Kendy et al. 2012). Within the ELOHA framework, streams are classified based on similar hydrology as a foundation for later assessing hydrologic alterations and flow-ecology relationships (see Arthington et al. 2006; Poff et al. 2010) within those stream groups (i.e., with similar hydrology). The flow-ecology relationships are then used in social processes to identify acceptable “ecological limits” to inform flow alteration thresholds and water-policy standards. However, ELOHA was not constructed to address applications requiring river-specific socio-economic and ecological issues (Kendy et al. 2012), a possible reason for its limited application in hydropower contexts (but see McManamay et al. 2013a, Rolls and Arthington 2014). Even so, ELOHA is flexible in that it provides context for the ecologic and hydrologic conditions of rivers, and reduces information gaps by isolating the most relevant hydrologic and ecological indicators, which may be relevant to the initial stages of e-flow implementation in hydropower.