Jonathan D. Herman

Many-objective reservoir policy identification and refinement to reduce policy inertia and myopia in water management

This study contributes a decision analytic framework to overcome policy inertia and myopia in complex river basin management contexts. The framework combines reservoir policy identification, many-objective optimization under uncertainty, and visual analytics to characterize current operations and discover key trade-offs between alternative policies for balancing competing demands and system uncertainties. The approach is demonstrated on the Conowingo Dam, located within the Lower Susquehanna River, USA. The Lower Susquehanna River is an interstate water body that has been subject to intensive water management efforts due to competing demands from urban water supply, atomic power plant cooling, hydropower production, and federally regulated environmental flows. We have identified a baseline operating policy for the Conowingo Dam that closely reproduces the dynamics of current releases and flows for the Lower Susquehanna and thus can be used to represent the preferences structure guiding current operations. Starting from this baseline policy, our proposed decision analytic framework then combines evolutionary many-objective optimization with visual analytics to discover new operating policies that better balance the trade-offs within the Lower Susquehanna. Our results confirm that the baseline operating policy, which only considers deterministic historical inflows, significantly overestimates the systems reliability in meeting the reservoirs competing demands. Our proposed framework removes this bias by successfully identifying alternative reservoir policies that are more robust to hydroclimatic uncertainties while also better addressing the trade-offs across the Conowingo Dams multisector services.

Navigating financial and supply reliability tradeoffs in regional drought management portfolios

Rising development costs and growing concerns over environmental impacts have led many communities to explore more diversified water management strategies. These “portfolio”-style approaches integrate existing supply infrastructure with other options such as conservation measures or water transfers. Diversified water supply portfolios have been shown to reduce the capacity and costs required to meet demand, while also providing greater adaptability to changing hydrologic conditions. However, this additional flexibility can also cause unexpected reductions in revenue (from conservation) or increased costs (from transfers). The resulting financial instability can act as a substantial disincentive to utilities seeking to implement more innovative water management techniques. This study seeks to design portfolios that employ financial tools (e.g., contingency funds and index insurance) to reduce fluctuations in revenues and costs, allowing these strategies to achieve improved performance without sacrificing financial stability. This analysis is applied to the development of coordinated regional supply portfolios in the “Research Triangle” region of North Carolina, an area comprising four rapidly growing municipalities. The actions of each independent utility become interconnected when shared infrastructure is utilized to enable interutility transfers, requiring the evaluation of regional tradeoffs in up to five performance and financial objectives. Diversified strategies introduce significant tradeoffs between achieving reliability goals and introducing burdensome variability in annual revenues and/or costs. Financial mitigation tools can mitigate the impacts of this variability, allowing for an alternative suite of improved solutions. This analysis provides a general template for utilities seeking to navigate the tradeoffs associated with more flexible, portfolio-style management approaches.

Cooperative drought adaptation: Integrating infrastructure development, conservation, and water transfers into adaptive policy pathways

A considerable fraction of urban water supply capacity serves primarily as a hedge against drought. Water utilities can reduce their dependence on firm capacity and forestall the development of new supplies using short-term drought management actions, such as conservation and transfers. Nevertheless, new supplies will often be needed, especially as demands rise due to population growth and economic development. Planning decisions regarding when and how to integrate new supply projects are fundamentally shaped by the way in which short-term adaptive drought management strategies are employed. To date, the challenges posed by long-term infrastructure sequencing and adaptive short-term drought management are treated independently, neglecting important feedbacks between planning and management actions. This work contributes a risk-based framework that uses continuously updating risk-of-failure (ROF) triggers to capture the feedbacks between short-term drought management actions (e.g., conservation and water transfers) and the selection and sequencing of a set of regional supply infrastructure options over the long term. Probabilistic regional water supply pathways are discovered for four water utilities in the “Research Triangle” region of North Carolina. Furthermore, this study distinguishes the status-quo planning path of independent action (encompassing utility-specific conservation and new supply infrastructure only) from two cooperative formulations: “weak” cooperation, which combines utility-specific conservation and infrastructure development with regional transfers, and “strong” cooperation, which also includes jointly developed regional infrastructure to support transfers. Results suggest that strong cooperation aids utilities in meeting their individual objectives at substantially lower costs and with less overall development. These benefits demonstrate how an adaptive, rule-based decision framework can coordinate integrated solutions that would not be identified using more traditional optimization methods.

Synthetic Drought Scenario Generation to Support Bottom-Up Water Supply Vulnerability Assessments

AbstractExploratory simulation allows analysts to discover scenarios in which existing or planned water supplies may fail to meet stakeholder objectives. These robustness assessments rely heavily on the choice of plausible future scenarios, which, in the case of drought management, requires sampling or generating a streamflow ensemble that extends beyond the historical record. This study develops a method to modify synthetic streamflow generators by increasing the frequency and severity of droughts for the purpose of exploratory modeling. To support management decisions, these synthetic droughts can be related to recent observed droughts of consequence for regional stakeholders. The method approximately preserves the spatial and temporal correlation of historical streamflow in drought-adjusted scenarios. The approach is demonstrated in a bottom-up planning context using an urban water portfolio design problem in North Carolina, a region whose water supply faces both climate and population pressures. Synth...

Internationally coordinated multi-mission planning is now critical to sustain the space-based rainfall observations needed for managing floods globally

At present 4 of 10 dedicated rainfall observing satellite systems have exceeded their design life, some by more than a decade. Here, we show operational implications for flood management of a ?collapse? of space-based rainfall observing infrastructure as well as the high-value opportunities for a globally coordinated portfolio of satellite missions and data services. Results show that the current portfolio of rainfall missions fails to meet operational data needs for flood management, even when assuming a perfectly coordinated data product from all current rainfall-focused missions (i.e., the full portfolio). In the full portfolio, satellite-based rainfall data deficits vary across the globe and may preclude climate adaptation in locations vulnerable to increasing flood risks. Moreover, removing satellites that are currently beyond their design life (i.e., the reduced portfolio) dramatically increases data deficits globally and could cause entire high intensity flood events to be unobserved. Recovery from the reduced portfolio is possible with internationally coordinated replenishment of as few as 2 of the 4 satellite systems beyond their design life, yielding rainfall data coverages that outperform the current full portfolio (i.e., an optimized portfolio of eight satellites can outperform ten satellites). This work demonstrates the potential for internationally coordinated satellite replenishment and data services to substantially enhance the cost-effectiveness, sustainability and operational value of space-based rainfall observations in managing evolving flood risks.

Scalable Multiobjective Control for Large-Scale Water Resources Systems Under Uncertainty

Advances in modeling and control have always played an important role in supporting water resources systems planning and management. Changes in climate and society are now introducing additional challenges for controlling these systems, motivating the emergence of complex, integrated simulation models to explore key causal relationships and dependences related to uncontrolled sources of variability. In this brief, we contribute a massively parallel implementation of the evolutionary multiobjective direct policy search method for controlling large-scale water resources systems under uncertainty. The method combines direct policy search with nonlinear approximating networks and a hierarchical parallelization of the Borg multiobjective evolutionary algorithm. This computational framework successfully identifies control policies that address both the presence of multidimensional tradeoffs and severe uncertainties in the system dynamics and policy performance. We demonstrate the approach on a challenging real-world application, represented by the optimal control of a network of four multipurpose water reservoirs in the Red River basin in Northern Vietnam, under observed and synthetically generated hydrologic conditions. Results show that the reliability of the computational framework in finding near-optimal solutions increases with the number of islands in the adopted hierarchical parallelization scheme. This setting reduces the vulnerabilities of the designed solutions to the system’s uncertainty and improves the discovery of robust control policies addressing key system performance tradeoffs.

Deriving global parameter estimates for the Noah land surface model using FLUXNET and machine learning

With their origins in numerical weather prediction and climate modeling, land surface models aim to accurately partition the surface energy balance. An overlooked challenge in these schemes is the role of model parameter uncertainty, particularly at unmonitored sites. This study provides global parameter estimates for the Noah land surface model using 85 eddy covariance sites in the global FLUXNET network. The at-site parameters are first calibrated using a Latin Hypercube based ensemble of the most sensitive parameters, determined by the Sobol method, to be the minimum stomatal resistance (rs,min), the Zilinkitevich empirical constant (Czil), and the bare soil evaporation exponent (fxexp). Calibration leads to an increase in the mean Kling-Gupta Efficiency (KGE) performance metric from 0.54 to 0.71. These calibrated parameter sets are then related to local environmental characteristics using the Extra-Trees machine learning algorithm. The fitted Extra-Trees model is used to map the optimal parameter sets over the globe at a 5 km spatial resolution. The leave-one-out cross-validation of the mapped parameters using the Noah land surface model suggests that there is the potential to skillfully relate calibrated model parameter sets to local environmental characteristics. The results demonstrate the potential to use FLUXNET to tune the parameterizations of surface fluxes in land surface models and to provide improved parameter estimates over the globe.

Policy tree optimization for threshold-based water resources management over multiple timescales

Abstract Water resources systems face irreducible uncertainty in supply and demand, requiring policies to respond to changing conditions on multiple timescales. For both short-term operation and long-term adaptation, thresholds or “decision triggers”, where a policy links observed indicators to actions, have featured prominently in recent studies. There remains a need for a general method to conceptualize threshold-based policies in an easily interpretable structure, and a corresponding search algorithm to design them. Here we propose a conceptual and computational framework where policies are formulated as binary trees, using a simulation-optimization approach. Folsom Reservoir, California serves as an illustrative case study, where policies define the thresholds triggering flood control and conservation actions. Candidate operating rules are generated across an ensemble of climate scenarios, incorporating indicator variables describing longer-term climate shifts to investigate opportunities for adaptation. Policy tree optimization and corresponding open-source software provide a generalizable, interpretable approach to policy design under uncertainty.

An open-source Python implementation of California's hydroeconomic optimization model

Abstract This short communication describes a new open-source implementation of the CALVIN model (CALifornia Value Integrated Network), a large-scale network flow optimization model of Californias water supply system. The model is cross-platform, uses common data formats, and connects to several freely available linear programming solvers. Given inputs including hydrology, urban/agricultural demand curves, and variable operating costs, the model minimizes the systemwide cost of water scarcity and operations including surface and groundwater reservoirs, wastewater reuse, desalination, environmental flow requirements, and hydropower. Key outputs include water shortage costs and marginal economic values of water and infrastructure capacity. We benchmark the scalability of different solvers up to roughly 5 million decision variables, using shared-memory parallelization on a high performance computing cluster. Runtimes are reduced by two orders of magnitude relative to the original model when no initial solution is provided, in addition to the benefits such as accessibility and transparency that come with an open-source platform. While this model is specific to California, the data and model structure are separated, so a similar framework could be used in any system where water allocation has been formulated as a network flow problem.

Environmental hedging: A theory and method for reconciling reservoir operations for downstream ecology and water supply: ENVIRONMENTAL HEDGING

Building reservoir release schedules to manage engineered river systems can involve costly trade-offs between storing and releasing water. As a result, the design of release schedules requires metrics that quantify the benefit and damages created by releases to the downstream ecosystem. Such metrics should support making operational decisions under uncertain hydrologic conditions, including drought and flood seasons. This study addresses this need and develops a reservoir operation rule structure and method to maximize downstream environmental benefit while meeting human water demands. The result is a general approach for hedging downstream environmental objectives. A multistage stochastic mixed-integer nonlinear program with Markov Chains, identifies optimal “environmental hedging,” releases to maximize environmental benefits subject to probabilistic seasonal hydrologic conditions, current, past, and future environmental demand, human water supply needs, infrastructure limitations, population dynamics, drought storage protection, and the rivers carrying capacity. Environmental hedging “hedges bets” for drought by reducing releases for fish, sometimes intentionally killing some fish early to reduce the likelihood of large fish kills and storage crises later. This approach is applied to Folsom reservoir in California to support survival of fall-run Chinook salmon in the lower American River for a range of carryover and initial storage cases. Benefit is measured in terms of fish survival; maintaining self-sustaining native fish populations is a significant indicator of ecosystem function. Environmental hedging meets human demand and outperforms other operating rules, including the current Folsom operating strategy, based on metrics of fish extirpation and water supply reliability.