Milton Friedman’s thermostat and sensitivity analysis of control policies

If you’re reading this blog, you probably already know that correlation does not necessarily imply causation. However, does a lack of correlation necessarily imply a lack of causation? Despite widespread misconception, even by Nobel Laureates, it does not. This is especially true for controlled systems, as explained in Milton Friedman’s thermostat example. The act of controlling a system for stability (e.g., a thermostat that expends energy to stabilize indoor air temperature) will tend to remove correlations between variables that are, in a certain sense, causally related (e.g., between energy expenditure and indoor temperature). More generally, the control of a dynamical system can introduce complex, path-dependent behavior that complicates the analysis of observed data using standard statistical techniques. This blog post will introduce these issues in the context of reservoir operations. I will then connect this to recent developments in using sensitivity analysis to improve understanding of how complex control policies respond to changing information over time, and demonstrate the benefits of using simulated state trajectories rather than random sampling approaches when analyzing controlled systems.

A Jupyter Notebook containing the Python code used for this analysis can be found in this GitHub repository.

Reservoir storage stabilization policy

First, consider a reservoir with a capacity of 20 MG, and daily inflows that average 10 MGD, with moderate persistence. Let’s assume the reservoir operator follows a “storage stabilization” policy that attempts to keep the storage constant at 10 MG. This simply requires the operator to release (S – S’ + I*) each day, where S is the current storage, S’ is the storage target, and I* is the projected inflow over the course of the next time step. How will this system evolve over time? With a perfect forecast, we have a situation analogous to Milton Friedman’s thermostat: the control policy keeps storage constant over time, leading it to become uncorrelated from both inflows and releases. Meanwhile, because storage is unvarying, the release depends only on inflows in a perfect linear relationship.

Figure 1: Time series of state variables for storage stabilization policy with perfect forecast.
Figure 2: Pairwise relationships between state variables for storage stabilization policy with perfect forecast.

The fact that inflow is uncorrelated with storage may appear paradoxical at first, but it makes sense once we understand the storage variable contributes no useful information to the system, since it is unchanging over time.

With imperfect forecasts, the operator sometimes needs to increase the release to avoid overtopping the reservoir, or decrease the release to avoid negative storage. This adds noise to the relationships above, but the reservoir releases are still minimally correlated to storages.

Figure 3: Pairwise relationships between state variables for storage stabilization policy with imperfect forecasts.

Power production policy

The reservoir storage stabilization policy and Milton Friedman’s thermostat example demonstrate just one particular aspect of a much broader point: the act of controlling a dynamical system will fundamentally alter the trajectories of the system through state-space. This has important implications for how observed data from these systems should be understood.

But first, consider an alternative reservoir control policy that attempts to take advantage of power price information when making releases for the purpose of hydropower production. The power price has a mean of 10 and lower persistance than the inflow distribution. We assume the following simple set of rules: (1) when price is below 8, release S/4 MG; (2) when price is between 8 and 12, release S/2; (3) when price is above 12, release (S + I*/2). Again, we also adjust the release to avoid overtopping the reservoir or incurring negative storages. Despite the simplicity of this policy, it produces surprisingly complex dynamics.

Figure 4: Time series of state variables for power production policy with perfect forecast.
Figure 5: Pairwise relationships between state variables for power production policy with imperfect forecasts.

In the bottom row of Figure 5, we see interesting and highly non-linear relationships driving the release policy as a function of the projected inflow, storage, and power price. Additionally, the relationships between the state variables themselves (i.e., power price and storage) can display non-linear and thresholding-type behavior. Comparing the storage stabilization policy and the power production policy underscores the importance of the control process in dictating the dynamics of the system and the regions of state-space that are explored.

Figure 6: Comparison of state-space inhabited by storage stabilization policy and power production policy.

Implications for sensitivity analysis of control policies

So what does any of this have to do with sensitivity analysis? Recent advances in adaptive control (e.g. Direct Policy Search, Reinforcement Learning) have allowed for the development of complex operating policies for managing water resource systems under uncertainty by adapting to evolving conditions over time (see recent reviews by Giuliani et al., 2021, and Herman et al., 2020). These operating policies can be built from Artificial Neural Networks, Radial Basis Functions, Decision Trees, or other functional forms, and can exhibit highly non-linear behavior. This complicates efforts to understand how they work “under the hood,” leading to skepticism from some researchers and decision-makers who prefer simpler and more transparent operating policies. Recent work has attempted to “open the black box” by combining sensitivity analysis and visualization in order to illuminate how different operating policies monitor and respond to different sources of information (e.g., Quinn et al., 2019; Hamilton et al., 2022).

In most applications of global sensitivity analysis, researchers want to understand the impacts of uncertain model parameters (e.g., soil permeability) or inputs (e.g., precipitation) on model outputs (e.g., runoff) (see Pianosi et al., 2016, for a review of sensitivity analysis in environmental models). Importantly, in general the inputs to the sensitivity analysis are considered to be exogenous factors and forcings that do not respond to the dynamics of the rest of the system model. For this reason, (quasi)random sampling approaches are often used to generate samples from across the feasible range of each parameter.

However, when sensitivity analysis is applied to the problem of better understanding adaptive control policies (e.g., reservoir releases), then the policy inputs may be either exogenous (e.g., inflows, power prices) or endogenous (e.g., storage). The latter, as we have seen, can cause highly non-linear and path-dependent dynamics within the system, resulting in simulated trajectories that are not uniformly distributed across state space. Any non-uniformity in the distribution of states is potentially valuable “information” about the dynamics of the control policy and the broader system, which we would like to capture with our sensitivity analysis. For this reason, it is preferable to use simulated system trajectories as inputs to the sensitivity analysis, rather than randomly generated inputs, in order to ensure that our sensitivity analysis reflects the actual data generating process of the system.

Figure 7: Comparison of state-space inhabited by storage stabilization policy, using randomly generated input data vs. simulated state trajectories.
Figure 8: Comparison of state-space inhabited by power production policy, using randomly generated input data vs. simulated state trajectories.

Due to the non-linear, non-independent, non-normally distributed nature of these simulated data, many common variance-based global sensitivity analysis methods may not be appropriate. However, moment-independent methods such as information theoretic sensitivity analysis and the delta-moment independent method can help overcome some of these challenges. See Hamilton et al., 2022, Hadjimichael et al., 2020, and references therein, as well as this blog post by Keyvan Malek, for more discussion of these issues.

For the sake of simplicity, I use R-squared as a simple indicator of variance-based sensitivity rather than more sophisticated measures such as Sobol sensitivity. I also apply a discretized version of information theoretic sensitivity analysis (ITSA). Both indices range between 0 and 1, but they cannot be directly compared to each other (e.g., R2 is often higher than ITSA given the same data) . Tables 1 and 2 show the sensitivity indices for the two operating policies.

ITSA, simR2, simITSI, unifR2, unif
Projected inflow0.520.950.260.69
Power price0.
Table 1: Comparison of sensitivity indicies for policy inputs governing reservoir releases, under storage stabilization policy
Projected inflow0.290.480.140.30
Power price0.
Table 2: Comparison of sensitivity indicies for policy inputs governing reservoir releases, under power production policy

The most obvious contrast is between the uniformly sampled data and the simulated data; the former tends to estimate significantly lower sensitivity. This is due to the missing information on the path-dependent relationships between states within the system, as a result of using randomly sampled input data rather than actual simulated trajectories. Another interesting comparison is ITSA_sim vs R2_sim for the power production policy. While R2 classifies projected inflow and storage as having roughly equivalent influence on release decisions, ITSA finds substantially more influence from storage. This finding makes sense when comparing the two relationships in Figure 5; both have similar shapes and variances if you squint, but the storage-release relationship has more well-defined fine structure that cannot be discerned by variance-based approaches.


Giuliani, M., Lamontagne, J. R., Reed, P. M., & A. Castelletti. (2021). A State-of-the-Art Review of Optimal Reservoir Control for Managing Conflicting Demands in a Changing World. Water Resources Research, 57, e2021WR029927.

Hadjimichael, A., Quinn, J., & P. Reed. (2020). Advancing diagnostic model evaluation to better understand water shortage mechanisms in institutionally complex river basins. Water Resources Research, 56, e2020WR028079.

Hamilton, A. L., Characklis, G. W., & P. M. Reed. (2022). From stream flows to cash flows: Leveraging Evolutionary Multi-Objective Direct Policy Search to manage hydrologic financial risks. Water Resources Research, 58, e2021WR029747.

Herman, J. D., Quinn, J. D., Steinschneider, S., Giuliani, M., & S. Fletcher (2020). Climate adaptation as a control problem: Review and perspectives on dynamic water resources planning under uncertainty. Water Resources Research, 56, e24389.

Pianosi, F., Beven, K., Freer, J., Hall, J. W., Rougier, J., Stephenson, D. B., & T. Wagener. (2016). Sensitivity analysis of environmental models: A systematic review with practical workflow. Environmental Modelling & Software, 79, 214-232.

Quinn, J. D., Reed, P. M., Giuliani, M., & A. Castelletti (2019). What is controlling our control rules? Opening the black box of multireservoir operating policies using time-varying sensitivity analysis. Water Resources Research, 55, 5962–5984.

So What’s the Rationale Behind the Water Programming Blog?

By Patrick M. Reed

In general, I’ve made an effort to keep this blog focused on the technical topics that have helped my students tackle various issues big and small. It helps with collaborative learning and maintaining our exploration of new ideas.

This post, however, represents a bit of departure from our normal posts in response to some requests for a suggested reading guide for my Fall 2019 AGU Paul A. Witherspoon Lecture entitled “Conflict, Coordination, and Control in Water Resources Systems Confronting Change” (on Youtube should you have interest). 

The intent is to take a step back and zoom out a bit to get the bigger picture behind what we’re doing as research group and much of the original motivation in initiating the Water Programming Blog itself. So below, I’ll provide a summary of papers that related to the topics covered in the lecture sequenced by the talk’s focal points at various points in time. Hope this provides some interesting reading for folks. My intent here is to keep this informal. The highlighted reading resources were helpful to me and are not meant to be a formal review of any form.

So let’s first highlight Paul Witherspoon himself, a truly exceptional scientist and leader (7 minute marker, slides 2-3).

  1. His biographical profile
  2. A summary of his legacy
  3. The LBL Memo Creating the Earth Sciences Division (an example of institutional change)

Next stop, do we understand coordination, control, and conflicting objectives in our institutionally complex river basins (10 minute marker, slides 6-8)? Some examples and a complex systems perspective.

  1. The NY Times Bomb Cyclone Example
  2. Interactive ProPublica and The Texas Tribune Interactive Boomtown, Flood Town (note this was written before Hurricane Harvey hit Houston)
  3. A Perspective on Interactions, Multiple Stressors, and Complex Systems (NCA4)

Does your scientific workflow define the scope of your hypotheses? Or do your hypotheses define how you need to advance your workflow (13 minute marker, slide 9)? How should we collaborate and network in our science?

  1. Dewey, J. (1958), Experience and Nature, Courier Corporation.
  2. Dewey, J. (1929), The quest for certainty: A study of the relation of knowledge and action.
  3. Hand, E. (2010), ‘Big science’ spurs collaborative trend: complicated projects mean that science is becoming ever more globalized–and Europe is leading the way, Nature, 463(7279), 282-283.
  4. Merali, Z. (2010), Error: Why Scientific Programming Does Not Compute, Nature, 467(October 14), 775-777.
  5. Cummings, J., and S. Kiesler (2007), Coordination costs and project outcomes in multi-university collaborations, Research Policy, 36, 1620-1634.
  6. National Research Council (2014), Convergence: facilitating transdisciplinary integration of life sciences, physical sciences, engineering, and beyond, National Academies Press.
  7. Wilkinson, M. D., M. Dumontier, I. J. Aalbersberg, G. Appleton, M. Axton, A. Baak, N. Blomberg, J.-W. Boiten, L. B. da Silva Santos, and P. E. Bourne (2016), The FAIR Guiding Principles for scientific data management and stewardship, Scientific Data, 3.
  8. Cash, D. W., W. C. Clark, F. Alcock, N. M. Dickson, N. Eckley, D. H. Guston, J. Jäger, and R. B. Mitchell (2003), Knowledge systems for sustainable development, Proceedings of the National Academy of Sciences, 100(14), 8086-8091.

Perspectives and background on Artificial Intelligence (15 minute marker, slides 10-16)

  1. Simon, H. A. (2019), The sciences of the artificial, MIT press.
  2. AI Knowledge Map: How To Classify AI Technologies
  3. Goldberg, D. E. (1989), Genetic Algorithms in Search, Optimization and Machine Learning, Addison-Wesley Publishing Company, Reading, MA
  4. Coello Coello, C., G. B. Lamont, and D. A. Van Veldhuizen (2007), Evolutionary Algorithms for Solving Multi-Objective Problems, 2 ed., Springer, New York, NY.
  5. Hadka, D. M. (2013), Robust, Adaptable Many-Objective Optimization: The Foundations, Parallelization and Application of the Borg MOEA.

The Wicked Problems Debate (~22 minute marker, slides 17-19) and the emergence of post normal science and decision making under deep uncertainty.

  1. Rittel, H., and M. Webber (1973), Dilemmas in a General Theory of Planning, Policy Sciences, 4, 155-169.
  2. Buchanan, R. (1992), Wicked problems in design thinking, Design Issues, 8(2), 5-21.
  3. Kwakkel, J. H., W. E. Walker, and M. Haasnoot (2016), Coping with the Wickedness of Public Policy Problems: Approaches for Decision Making under Deep Uncertainty, Journal of Water Resources Planning and Management, 01816001.
  4. Ravetz, J. R., and S. Funtowicz (1993), Science for the post-normal age, Futures, 25(7), 735-755.
  5. Turnpenny, J., M. Jones, and I. Lorenzoni (2011), Where now for post-normal science?: a critical review of its development, definitions, and uses, Science, Technology, & Human Values, 36(3), 287-306.
  6. Marchau, V. A., W. E. Walker, P. J. Bloemen, and S. W. Popper (2019), Decision making under deep uncertainty, Springer.
  7. Mitchell, M. (2009), Complexity: A guided tour, Oxford University Press.

Lastly, the Vietnam and North Carolina application examples.

  1. Quinn, J. D., P. M. Reed, M. Giuliani, and A. Castelletti (2019), What Is Controlling Our Control Rules? Opening the Black Box of Multireservoir Operating Policies Using Time-Varying Sensitivity Analysis, Water Resources Research, 55(7), 5962-5984.
  2. Quinn, J. D., P. M. Reed, M. Giuliani, A. Castelletti, J. W. Oyler, and R. E. Nicholas (2018), Exploring How Changing Monsoonal Dynamics and Human Pressures Challenge Multireservoir Management for Flood Protection, Hydropower Production, and Agricultural Water Supply, Water Resources Research, 54(7), 4638-4662.
  3. Trindade, B. C., P. M. Reed, and G. W. Characklis (2019), Deeply uncertain pathways: Integrated multi-city regional water supply infrastructure investment and portfolio management, Advances in Water Resources, 134, 103442.
  4. Gold, D., Reed, P. M., Trindade, B., and Characklis, G., “Identifying Actionable Compromises: Navigating Multi-City Robustness Conflicts to Discover Cooperative Safe Operating Spaces for Regional Water Supply Portfolios.“, Water Resources Research,  v55, no. 11,  DOI:10.1029/2019WR025462, 9024-9050, 2019.