ater supply and flood prevention are constant concerns for the managers of dams and reservoirs,
particularly in the American West. Reservoirs must maintain enough empty space to handle sudden storm flows while at the same time holding on to as much precious water as they can. In the past, problems like this were addressed with relatively simple, inflexible capacity requirements based on an analysis of historical behavior. Today, however, advances in hydrometeorological forecasting and big data analysis allow reservoir managers to operate on a far more flexible scale.
HDR Engineering is helping its clients across the United States integrate forecast-informed decision support techniques into their operations in order to make them more efficient and flexible. In this interview, David Ford, a water management and hydrologic engineering expert at HDR, and Mike McMahon, HDR’s senior hydrometeorologist, speak with Municipal Water Leader Managing Editor Joshua Dill about recent developments in the field of forecast-informed management.
Joshua Dill: Please tell us about your backgrounds and how you came to be in your current positions.
David Ford: I am a registered civil engineer in California with a PhD from the University of Texas at Austin; my field of specialty is hydrologic engineering. I apply hydrologic science to answer the water management questions HDR clients have. I started out my career working for the U.S. Army Corps of Engineers; I worked at an Army Corps research lab called the Hydrologic Engineering Center for 10 years. I then left and formed a consultancy, David Ford Consulting Engineers. We were in business for 28 years and were acquired in August 2018 by HDR. I have been based in Sacramento, California, for more than 40 years.
Mike McMahon: I am the senior hydrometeorologist at HDR Engineering. I have been a hydrometeorologist for the last 34 years. I worked in operations for the first 20 years with Lockheed and Boeing as an atmospheric scientist and researcher. Next, 12 years ago, I came to HDR and started working in hydrometeorology, which means that I focus on the part of the water cycle that is precipitation. You could say I do the front end of what David does once that water hits the ground. That involves the use of advanced analytics as part of operational/planning analysis and forecasting. I am the climate change and resiliency lead for the company as well. Currently, my work is split about 50/50 between hydrometeorology on the one hand and climate change and resiliency work on the other. I lived in California for 45 years prior to moving to Denver, where I have lived for the last 13 years. I absolutely love what I do.
Joshua Dill: Dr. Ford, does your work in hydrology focus on the behavior of lakes and rivers?
David Ford: I would call my field flood hydrology, because I focus mostly on runoff from big storms. For example, when Mike predicts the kind of major rain event that we recently had in California, we hydrological engineers take the rainfall predictions and use those to predict how much water is going to run off into the lakes and reservoirs so that those infrastructure systems can be better managed. We’re concerned with things like the ponding of water on the land surface and how much of it will infiltrate, or go into the soil. In California, we’re concerned with how much of that precipitation is going to fall as snow and accumulate in the mountains, and then the rate at which that snow will melt over time. We’re also interested in temperature measurements and temperature forecasts.
Joshua Dill: Please tell us about the benefits of forecast- informed operations, especially for reservoirs and dams.
David Ford: Implementing forecast-informed operations is a way for HDR to help our clients better use the infrastructure they have. In California and in the western United States, we have a seasonal rainfall pattern that allows us to be quite sure that we are not going to have rain in June and July. Reservoirs in the West are managed to accommodate that seasonal difference. In the rainy season, we keep the reservoirs low so that we can store flood runoff; in the summer season, we keep the reservoirs high, or filled, so that there is water for municipal clients and for commercial and industrial use. The challenge is deciding when to fill and when to empty those reservoirs. This technology allows managers to make those decisions more efficiently and in a timelier manner than they used to do.
Here is a great example. One of our clients is the Sonoma County Water Agency (SCWA), which manages the storage of water in Lake Mendocino in Sonoma County, just northwest of San Francisco. The rules for the allocation of water and the use of storage in Lake Mendocino were developed back in the 1980s, when there wasn’t a good method of forecasting rainfall, stream flow, and runoff into that reservoir. The rules were inflexible. They said that on a certain date, the reservoir needed to be drawn down to have lots of empty space for managing floods. It didn’t matter what the meteorologists were saying.
Now, however, we can use forecasting to inform those decisions. We can say, “We don’t see a big storm coming. Let’s hold that water in there to use later in the season.”That is important in California, because we want all the water we can hold to use throughout our long, hot, dry summers. If we follow these arbitrary rules based only on historical behavior and empty that reservoir, and no big storm comes in, we might be short on water during the summer. Forecast-informed operations let us manage that space a lot more efficiently.
Joshua Dill: Have existing guidelines and rules also become somewhat obsolete because of climatic changes?
Mike McMahon: Yes. Today, we need to be able to manage increased natural climate variability. One key factor in managing that variability is that a forecast-informed reservoir operations program is iterative. In other words, your decisions are based on what is happening with the hydrometeorology of a given basin right now as it pertains to the coming water year. This program enables you to iteratively adapt to what the climate is doing at any given time, whether that means droughts; floods; or a nice, normal water year. You can make adjustments as you go, rather than relying on a future year climate model.
Joshua Dill: It sounds like some reservoir managers are being exempted from older water allocation and storage rules and are now permitted to implement forecast-informed operations. How did this come about?
David Ford: It came about because of the recognition by both the dam owners and operators and the regulators that the technology could now support forecast-informed operations. The operation of Lake Mendocino during flood periods, for example, is subject to rules developed by the Army Corps. The water agency follows those rules, which require it to keep empty space in that reservoir. But the Army Corps recognizes that improved forecasting allows those decisions to be made in a more adaptive manner. With the cooperation and collaboration of a group that included the U.S. National Weather Service, which is part of the U.S. Department of Commerce, the Army Corps, the California Department of Water Resources, various university researchers, and the National Marine Fisheries, SCWA was able to incorporate improved forecasting into its operational decisionmaking, increasing water availability without increasing flood risk.
A similar thing happened near where I live in Sacramento. There’s a big reservoir just upstream from us called Folsom Reservoir. The release from Folsom flows down the American River, which goes right through the urban area of Sacramento.The American River’s capacity is limited, so it is important to operate Folsom Reservoir in such a way that flood releases don’t exceed that capacity, but also in such a way to ensure that there’s water in that reservoir for all the water supply clients. It is almost the opposite of the Lake Mendocino situation. At Folsom Reservoir, the focus is on flood management, and forecast-informed operations are used to inform decisions to draw the reservoir down in anticipation of a flood in a flexible and proactive fashion. That’s all possible because the regulators and the operating agencies have recognized that the technology has improved enough to make this possible.
Mike McMahon: Both Folsom and Lake Mendocino have minimum in-stream flow requirements that are designed to support fish populations below the dam. In years of drought, they have to release as much as they can for the fisheries downstream, but in years of flood, particularly in the case of Lake Mendocino, they can’t release too much, because it is a hazard to those downstream and because it could adversely affect the fisheries within the Russian River. Particularly in the Russian River, there is an issue with salmon spawning grounds. In the case of Folsom, water temperature is important to the fisheries in the American River below the dam. Forecast-informed reservoir operations can help control these temperature variables as well. At Kerr Dam on Flathead Lake, they are also using forecast-informed operations to address hydropower management issues and minimum stream flows for two environmentally sensitive trout species. Operating a reservoir more efficiently has ancillary benefits beyond flood control and water supply.
Joshua Dill: Aside from dam and reservoir management, what other applications are there for forecast- informed reservoir operations?
David Ford: We are providing similar services for storm water management. Every urban community in the United States has storm drains, small open ditches, canals, channels, and detention storage facilities that have to be effectively managed. The scale is different, but the technology can also help with that. Weather forecasting, albeit on much smaller scale, can help emergency managers make decisions about when roads should be closed. HDR is helping provide services like that to our clients, too. We have ongoing projects in North Carolina and in San Antonio, Texas.
Mike McMahon: We’re doing the same kind of thing in wastewater as well, especially in combined sewer systems that manage storm water and wastewater simultaneously. There is a cost benefit to doing this. I’ll give you an example. Orange County Sanitation District is looking at infrastructure options to enable it to better manage its system capacity, but it is already regularly using forecast information to manage its system. With the increased efficiencies the district could get from using forecast information, it might not have to fund costly infrastructure to improve its system conveyance. There is a distinct cost benefit to increasing your water management efficiencies through improved decision-support information rather than just relying on bigger, better, and faster infrastructure.
Joshua Dill: Are these applications of forecast-informed operations paving the way for a broader adoption of the technology?
David Ford: The applications we discuss here represent the front edge. I would predict a great shift to forecast- informed operations for virtually every reservoir in the United States that has significant flood-management or water supply capabilities. It’s just a better way to use the infrastructure that is already there.The great thing is that the regulatory agencies understand that. The leadership of the Army Corps in Washington, DC, is well aware of the opportunities that have arisen, and it is embracing them. There are still details to be worked out, but the technology is not experimental any more. HDR is excited to be on the front edge of this technology and to help our clients implement it.
Mike McMahon: I will extend my career by 5 years just so that I can work on it. It is that important. Rolling out this paradigm change is complex, but it is exciting to be part of making it a reality. Right now, our task is not to achieve the highest conceivable level of efficiency, but to start doing better than we are right now. Over time, the technology will improve, and it will be able to be rolled out across the United States.
Joshua Dill: How can HDR help agencies deal with this new technology and navigate the difficulties of implementing it?
David Ford: One of the most important things that HDR can do to help our clients move forward with this technology is what Mike is doing: forecasting weather and interpreting those forecasts. The second thing, which HDR is also on the forefront of, is the use of information technology. This all works because we have computers to take in massive amounts of data about the weather, the watershed’s response, the snow melt, and the runoff, and to use that to analyze alternative operations schemes and help our clients make decisions.
Mike McMahon: There is a third component that I think is worth mentioning: the issue of dam and levee safety. When we start talking about operating reservoirs in a different way, there is a natural concern about dam safety. Forecast-informed reservoir operations will not reduce dam safety—it will be able to enhance reservoir management to the point that increases it. The decision-support information that David just mentioned will provide water managers with the tools they need to improve operational efficiencies, and that includes dam safety.