USGS Home
Contact USGS
Search USGS

The Reds Wolman Lecture

Presented by Robert M. Hirsch, USGS,

At the Consortium of Universities for the Advancement of Hydrologic Science Incorporated (CUAHSI) Biennial Symposium on Hydrologic Science and Engineering

July 15, 2008, Boulder Colorado

The audio for this podcast can be heard or downloaded at
http://water.usgs.gov/dispatch/2008/podcast/
Transcript has been slightly edited from the talk to improve the readability.

Thank you Rick and thank you to the leadership of CUAHSI for inviting me to speak to you. It's a thrill and opportunity to speak to this distinguished audience that we have here today, and also to be giving a named lecture in honor of Reds Wolman.

I hope in this talk to be provocative and I hope you disagree with some things that I say and it spurs controversy in the discussion. Sometimes I like going without powerpoint because I really want to talk about concepts and philosophy and ideas rather than specific scientific details. I will just say a couple of things about Reds at the start, and come back to specifics about him throughout the talk.

He was born in 1924 in Baltimore. His father was already on the faculty of The Johns Hopkins University. Reds attained a bachelor's degree in geology. He was also an all-American lacrosse player at Johns Hopkins. Some said that he sometimes selected his graduate students based on lacrosse-playing abilities. I was not one of those selected on that basis.

He earned a Ph.D. in Geology at Harvard and came to work at USGS and worked from 1951 to 58 at the USGS doing fluvial geomorphology and of course published a very important textbook, Fluvial Processes in Geomorphology, along with Luna Leopold and John Miller.

In 1958, he was the Chairman of the Department called TIBDOG, that's the Isaiah Bowman Department of Geography at Johns Hopkins University. There he recognized the need to bring geography together with water sciences and engineering and was instrumental in creating in 1970 a merger of sanitary engineering and the geography department and that became DOGEE, the Department of Geography and Environmental Engineering, which he chaired for 20 years.

He is a great educator and continues to teach to this day. He is a great scientist and citizen with many contributions to our national thinking and laws and governance about water and a great scientist of the subject of rivers.

I'm going to start off actually with a quote from the historian Stephen Ambrose, who was talking about Lewis and Clark and the history of North America and I want to apply that to the history of water resources and water resources science. In a talk by Ambrose based on his famous book, Undaunted Courage, I heard him say that the 19th century was the century of discovery and description and Lewis and Clark were a part of that. The 20th century was the century of command and degrade. We commanded our natural resources and we degraded our natural resources. Ambrose went on to say that the 21st century should be the century of restoration.

I think our science has followed that pattern and needs to follow that pattern. Let me just talk about a little about how I see that. The 19th century and into the early 20th century was the era of description, starting with the people like John Wesley Powell and Frederick Newell, recognizing the need to measure the flow of our streams which they started in 1888 in New Mexico. Also, people like Darton and Mendelhall whose job it was to describe the ground-water resources of the United State s and how it behaves.

That's into 20th century so this time frame shifts a little bit into the first half of the 20th century. A couple of other names probably less known to you, and that's Dole and Stabler, who were two of what we would now call geochemists with the USGS who said, "You know, we ought to know something about the flux of the major ions from the continent to the oceans out of major river basins." They set about to measure for a year or two years at a wide variety of locations and created extremely important data sets which we can now look back on, to understand how our Nation's rivers have changed chemically over a hundred year timeframe. So that was the period of discovery and description.

The 20th century is of course "command and degrade." Probably the single most important aspect of that for water is the dam building era starting the 1930s right into the 1970s. Science is behind it with people like Hazen and others looking at the question of safe yield. But I would bring in an example from the work of Reds Wolman here. He, along with Gar Williams and some others, got into the description of the downstream effect of dams and the fact that the river channel below dams is forever changed. They began to describe that process and how it worked and published some excellent work on the topic of the downstream effect of dams.

The 20th century is the time of ground water development, particularly accelerating at the time of the development of high-capacity pumps and center pivot irrigation systems which led to significant draw-downs of major aquifer systems. The science that went along with that development was the work of people like Theis, Ferris, Jacob and others who described the process of what is needed to pump and to dewater and to develop an entire aquifer system.

Of course, we now see many of the consequences associated with that dewatering process. The 20th century was the time of point-source ground-water pollution leading us to Love Canal, Woburn, Superfund and all those kinds of activities and the science to try to understand solute transport in the subsurface.

The 20th century was the time of massive degradation of our surface water quality, culminating, if you will, in the Cuyahoga River burning just prior to the passage of the Clean Water Act. The Clean Water Act is a very important landmark moment, in 1972, in the history of our nation's surface water quality and the science of people like Clarence Velz and others who described this thing called the dissolved oxygen sag and the recognition that we need to stop arguing who caused the pollution and how much they need to clean it up and instead set some standards and say, "We'll clean it up to a certain technological standard." The writers of the Clean Water Act recognized that that may not be sufficient to get the water to the quality that we need. We then need to go on to see what the next steps are that we need to take. We are just now getting to that point since the 1990s and into this decade.

Interestingly, Reds Wolman wrote intensively about this topic in a seminal paper in 1971 in the journal Science. The paper is titled "The Nation's Rivers." In this paper he pointed out just how little we know about the nature of the degradation or improvement of our nation's rivers and the need for a systematic long-term measurement as well as analysis of the degradation and the many forms of degradation that occurred.

Interestingly and characteristically of Reds, he came back to this topic a couple of decades later in 1987 with a follow up paper looking at trends. He based this on set of measurements that he really inspired being taken through the NASQAN program of the USGS and also a program called the Benchmark program that is really the idea of Luna Leopold. In his 1987 paper he began to go back and say, "What have we learned since 1972 as we move forward in the restoration of our nation's surface water?"

Finally, one of the other aspects of degradation is the effect of urbanization on our nation's water landscape. Again, Wolman made absolutely critical contributions to the subject with a paper of his, "A Cycle of Sedimentation and Erosion in Urban River Channels." He recognizing the massive changes in sediment budgets and river channel dynamics that occurs as the process of urbanization proceeds. And it was that work which led directly to a conversation that Reds had with law-makers in the State of Maryland that led to the first regulatory measures to try to control these consequences of urbanization. It led to regulation that went nationwide to protect against some of those consequences.

So now I'm into the 21st century and the subject of restoration and I'll just mention some aspects of that. Then I'll get into what I think the science challenges are associated with that restoration. Of course we have going on today some of the most exciting pieces of integrated scientific work supporting this restoration concept in places like the Everglades ecosystem, the San Francisco Bay Delta, the Grand Canyon, the Platte River, and the list goes on and on.

In these systems we know that through the processes of resource development and modification of channels and the extraction of water and introduction of pollution, they have been severely degraded. Our society has set about to try to do something to restore them, maybe not to their pristine condition but to clearly a better condition. These present huge scientific challenges and are in fact funding a lot of really outstanding science across the nation.

The focus on issues of ecological flows, issues of ground-water remediation and the recognition that simply pumping and treating are not going to be sufficient. I will not talk more about ground-water remediation but I think that, of course, is one of the great challenges in this restoration for the future. We know we just cannot pump it out; we know we just cannot pump and treat it. We need to find ways of working with the microbial community and natural processes to make that remediation happen.

On stream restoration, or the idea that we want our river channels to look something like the river channels that were there before we started modifying the flows and the landscape, the subject has its roots in the work that Reds Wolman did along with Luna Leopold and others in terms of describing what is the shape and the meander pattern of rivers in our natural landscape.

Obviously, another way of restoration is through agricultural practices. We are adding immense amounts of agricultural chemicals to our landscape, with significant consequences for the downstream environment, including on into the estuaries and the oceans. How can we restore these systems at the same time as we maintain a high level of agricultural production?

And finally urban drainage design. I mentioned Reds. involvement and the question of urbanization and sediment transport. I think we are seeing the beginning of an exciting revolution in the whole concept of urban design. People of my generation were trained in the idea that the landscape needs to be designed in such a way that you get the water off the landscape as rapidly as possible and into these stream channels. This leads to the depletion of ground water and has effects on river channels and habitats. The recognition is now that we need to keep the water on the landscape in a way that is not harmful to the built infrastructure in order to help restore the aquatic ecosystem in these urban environments. There is a tremendous amount of science that needs to be done to understand how we go about that process of re-engineering our urban and suburban areas, to keep the water on the landscape rather than getting it off the landscape as rapidly as possible.

I want to say something about three characteristics that I think need to be present in hydrology to support the restoration issues of the 21st century. These are: documentation of change, process-based research, and interaction with the public. Then I'll start walking through some of those specific restoration issues.

The first point is that our work must involve the clear documentation of the changes in the hydrologic system, and I include within that the ecological system that depends on that hydrologic system. This requires the collection of data on a long-term, ongoing basis and the analysis of that data. This is absolutely central to our understanding what's going on.

I want to give you a very brief quote from a wonderful paper, a 2-page Perspectives paper that was published in Science magazine some months ago by Ralph Keeling and it was in recognition of the outstanding work done by his father David Keeling in creating the Mauna Loa Observatory record of carbon dioxide, truly a global signal of carbon dioxide so that we could all understand what was going on. And here's what Ralph Keeling said about the work of his father and the monitoring a global atmospheric CO₂.

"A continuing challenge to long-term earth observations is the prejudice against science that is not directly aimed at hypothesis testing. At a time when the planet is being propelled by human action [. . . and here he goes on to talk specifically about climate change but I would actually broaden it to a whole set of changes, not just climate change], we cannot afford such a rigid view of the scientific enterprise."

He goes on to say, "The only way to figure out what is happening to our planet is to measure it, and this means tracking changes decade after decade and poring over the records." If you know the work of David Keeling you recognize that almost immediately from the start of his collection of the record, he started making very important scientific observations about the behavior of atmospheric CO₂.

He really had no idea what kinds of things he would uncover in the process but it was the vigilance towards extraordinarily accurate data collection and the continuing analysis of that data, to derive meaning from it, leading to experiments and other observations. Ralph Keeling's article about his father points out the tremendous difficulty his father had in keeping the funding going from the wide range of funding agencies who kept saying, "Where's the hypothesis? This is just data collection."

And yet now we look back and say, the Keeling Mauna Loa CO₂ curve is probably the single most important geophysical or biogeophysical record that we have today, in terms of understanding what is happening to the Earth. We have to all work towards making sure that these critical records are collected. Many of us, who are always struggling to find the way to keep funding for long-term observations of the Earth, are always looking for a record that is as significant, as much of a sentinel, as Keeling's record.

Keeling's record is incredibly clean and straightforward and presents a very, very beautiful pattern. Most of the records that we can think of, that I think are really important, are vastly noisier than Keeling's CO₂ record. And hence the challenge is that there are some very important things going on those hydrologic records but there is a tremendous signal to noise problem: whether we're talking about flows, whether we're talking about chemistry, or whether we're talking about sediment. Our job is not just to collect the data but to tease the meaning out of it and to help the public understand what's in those records and what the data are telling us, and also to looking backwards in time to try to understand where those variables stood at various times in the past.

Wolman also wrote about the importance of observation in his 1971 paper, "The Nation's Rivers." I'll paraphrase a few statements from the end to that paper: He said that few observational programs combine the necessary hydrology with the measures of water quality and river characteristics and biology that are really needed in order to understand what's going on with our nation's rivers.

Notice that he recognized the terribly great importance of the physical processes in hydrology. In this case, the flow, in order to understand what's going on in water quality, biology, etc. And that the linkage between the physical, chemical, and biological is critical.

He goes on to say that interpretation of the data is vital and that estimates of inputs based on budgets of materials from individual outputs and land use are critical. So it's not just what's measured (not just "bean counting") but it is trying to put together what's going on in the atmosphere, at the land surface, in the subsurface, what are humans doing on a watershed, what's moving down the river, and trying to understand the dynamics and the conservative and non-conservative amounts of these pollutants.

And he recognized the tremendous importance of the effect of flow and the effect of seasonal variation in trying to understand these records, and even talked about the concept that we talk about a lot today, which is that there are thresholds and that there are tipping points. He didn't use the word "tipping points" but he did use the word "thresholds."

He also pointed out that flow regulation might be even more important than pollution in determining the consequences for aquatic ecosystems. So I thought that was rather prophetic on Reds' part, so many years ago, that he identified the ecological need for water quantity and I think those exist today.

So one part is clearly having the information and being engaged in using it. I think it's CUAHSI's responsibility in the realm of education to make sure that every student who comes along understands how to collect data and understands the value of data and how to extract meaning from the data. I think that's absolutely a critical piece.

The second point is that the documentation of changes needs to be coupled with process-based research and that's where a lot of you come in. Process-based research needs to be based on observations. Obviously on theory as well, but we need to use real data to test our hypotheses, to estimate our coefficients, using either long time series or experiments that we design in order to do that.

I would also argue that the restoration efforts that are going on on the landscape today are, in fact, experiments. And the opportunity to use those restoration efforts as our experimental designs I think is a critical part of where hydrologic research is going. I would also argue that policy decisions, good ones and bad ones, that are out there and occurring as we speak, also represent experiments.

For example, what I would call the corn-ethanol experiment going on in the United States right now, the very ill-advised experiment in my view, nevertheless presents a tremendous science opportunity to understand how ground water, surface water, nutrients, etc. behave because we have a rapid change in what's going on in our nation's watersheds. We need to look for those opportunities for experiments that are going on even if we didn't design them.

And then the third characteristic, as we move through documentation, move through process research. The third part is interacting with the public and the policy process in order to help those things have an impact on society. I'll tell you one of Reds. favorite humorous stories. It was in the aftermath of Hurricane Agnes and the monster flood that came through the Baltimore area in 1972. And there was a lot of concern by the politicians about flooding and what to do to prevent future situations of this kind. Reds was asked to speak to local legislators, and he got pretty wrapped up in the kind of research that he was doing. He talked about the effect of 100-year floods, and the effect of 50-year floods, and the effect of 10-year floods and the mean annual flood and all these different things and the consequences and what they mean to river channels. He was really riffing on all this great stuff that he was so well known for. And one of the legislators raised his hand and he said, "Hey Doc, I've got a question." And Reds always says "when they say, 'Hey, Doc!' you know you're in trouble."

"I've got a question. You know you're talking about 50-year floods and 10-year floods and all this kind of stuff. That's all very well and good, but I'm elected in every two years. Why don't you talk to us about the 2-year flood?" It was a wake up call to Reds, which is to say: find a way to talk to people and the policy makers in the manner that they're going to understand and is relevant to their career, their position, so that you can really have an impact. And in fact, he did have an impact. As I mentioned, his ideas lead to the changing of policies with respect to urban development. He really did have that kind of an impact.

I'm going to talk about five challenging topics for the future of hydrology. These topics are not finely divided, there's a lot of overlap between these topics, but I found it convenient to organize my thoughts in this particular manner.

The first one I'm going to talk about has to do with the scientific basis for water planning and allocation. And the subtitle is "beyond stationary." The second one is the subject of ecological flows. And the third is understanding and controlling non-point source pollution with an emphasis on nutrients. The fourth is the causes of aquatic ecosystem decline. And the fifth is the role of hydrologic systems in the sequestration of carbon. And then I'll follow it with some extensions on these thoughts, drawing from the writings of a couple of different people.

So the first of these topics is really a scientific basis for water planning but I might also say not just planning but what I would call the reallocation process for water in the United States. At the simplest level, and it doesn't matter whether you're talking about the West or the East, it is basically the same. There's a built infrastructure and a social contract of some kind. It says that the water that's out there in the environment, that there is a piece of it for the cities and there's a piece of it for the farmers and there's a piece of it for industry. And that traditionally that was all that mattered and it was allocated in that way.

Some changes have occurred, but the amount of water for farming, and the amount for cities haven't changed that much in many cases. But the thing that has happened that is new is that there's another player at the table. And that player is the fish or the aquatic ecosystem in general. So, as I like to say, we have a re-negotiation of the social contract over water going on in the United States. And now in the re-negotiation the fish, and really the aquatic ecosystem, has "a seat at the table" whether it's happening in the courtroom or happening through some kind of negotiated process.

And the question that needs to be answered is: "'well, how much water does the fish need?" How can we readjust all the other aspects of the delivery of water, which might involve building new infrastructure but might involve changing the way we manage it. This water planning also involves things like flood mitigation, building levies, as well as water supply issues. And it really is an issue of restoration because the big restoration projects like the Everglades and the San Francisco Bay-Delta are fundamentally about water allocation and water availability. In fact, both the Everglades and San Francisco Bay-Delta are about the fact that the cities need their water, the farmers need their water but the ecosystem is standing in the way of delivering the water they need. And they need to find the way to accommodate the ecosystem and still deliver the water that people are coming to expect and that is the basis for their regional economies.

For people of my generation, and most of the people in the room, the basis of what we learned when we took our classes in water planning and management was a legacy that came from something called The Harvard Water Program from 1955 to 1960. It was a group of very, very smart individuals who spanned hydrology, economics, operations research, and statistics. They laid out a truly path-breaking set of ideas about the way to think about the planning of river basin development.

But underlying everything that they said about water planning was the concept of stationarity. The flow of the river system has a mean, it has a standard deviation, it has a serial correlation structure. In fact Mike Fiering started synthetic hydrology as a part of that whole effort. And while we may not know that mean and that standard deviation perfectly, we can describe our uncertainty about it. And we can proceed forward in a rational planning manner.

We could describe things like trade-offs between capacity of the dam and flood risks or capacity of the dam and water supply and probability of a shortage and all these kinds of things. We could do risk-cost trade-offs or benefit-cost trade-offs. And it was all predicated on the idea that the hydrologic system was a stationary process.

Well, a number of things have come in the way of our thinking. I first and foremost want to mention land use change is one the reasons that we could no longer accept the stationary paradigm. And I want to point to some wonderful work that has been done recently in an example from the State of Wisconsin. Paul Juckem, Randy Hunt, Mary Anderson, and Dale Robertson, a terrific paper on the changes in the streamflow in the driftless area in Wisconsin over a 60 or 70 year period.

Some of this we don't quite understand but probably has to do with agricultural practices changing the nature of baseflow and stormflow. Just to be specific, baseflow increases as much as 37% in these basins, and stormflow decreases 23% somewhere around 1970.

There are changes that are occurring in our nation's streams, some of which we understand, some of which we don't. And it gets back to analyzing our records and understanding the role that agricultural practices can have and the role that ground-water development can have. You look at places like the Upper San Pedro River in Arizona where tremendous ecological consequences are coming about because of ground water development and its effect on flow.

Another recent, interesting piece of work I saw from my USGS colleagues in Kansas looks at the most recent drought there from 2002 to 2006. They looked at runoff as a percentage of precipitation in some fairly arid basins in Kansas in the recent drought and some of the previous droughts. Runoff was something like 3½ percent of precipitation during the severe drought situations.

The most recent drought which was by no means meteorologically a very severe drought, it was a modest drought, we were getting 1½ percent of precipitation coming off as runoff. Something has changed in the rainfall-runoff relationship. Again, this is probably due to agricultural practices and due to ground-water development. We as hydrologists, I think, have a big challenge in trying to describe this kind of change.

We can also move to the urban environment I mentioned earlier in terms of high flows. We also see tremendous changes in low flows. The Ipswich River in Massachusetts is one of the "poster children" for this. In a very wet part of the United States in Massachusetts just outside of Boston, the stream almost annually literally dries up. And that's because of ground-water development going on in an urban area. So one cause of non-stationary is what humans are doing on the landscape, whether in terms of ground-water development or the changes in the landscape characteristics due to urbanization or agricultural practices.

The other, of course, big challenge to stationarity is that of climate change. I'm going to borrow heavily here from the paper that Chris Milly, Julio Betancourt, myself and several others have published. This Perspectives piece was published in Science this Winter, and is called "Stationary is Dead, Whither Water Management." But I'd like to give you my own personal perspective on this. I'm just so delighted to see Peter Eagleson in the room because it reminds me of when I met Pete.

It was a conference about El Niņo and other long term ocean-atmosphere-landscape interactions. This meeting was about 20 years ago. And when I think about my education in hydrology, I never heard the words El Niņo. It was completely unknown idea to me. We were very rooted in the idea of stationarity and very rooted on the continents rather than thinking about the oceans.

The history of the last 20 years and our thinking has been that we started talking about El Niņo and its effect on landscape hydrology. Then people started talking about Pacific Decadal Oscillation. Then people started talking about Atlantic Multi-Decadal Oscillation. These are all, in my view, a set of quasi-periodic phenomena. We don't know why they start. We don't know why they end. We don't know when they're going to start. We don't know when they're going to end. We know something about their general characteristics. But the modeling community is yet unable to really describe when it's going to start, when it's going to end. So it's not particularly useful from a forecasting standpoint until we know it is actually upon us.

And of course then you can have the biggest quasi-periodic phenomena of them all and that is something called the Ice Ages. There again, we don't know why they started. We don't know why they ended. We don't know what the forces are that really created it. We have hypothese but I would argue that they're not well-demonstrated at this point.

All these factors, which have nothing to do with mankind mucking around with the climate system, should give us a tremendous amount of humility and recognition that the stationary paradigm doesn't fundamentally make a whole lot of sense. There are things that are happening at time scales of a few years up to tens of thousands or hundreds of thousands of years that we don't understand. And where the mean is going to be tomorrow is something that we really can't say much about. All of this leads me to think that we need to have more robust approaches to our water management and recognize that the future may not look so much like the past.

Then we add to that, this concept of the greenhouse effect. We clearly know that we are adding radiatively active gases to the atmosphere and that we can demonstrate that they are having an effect. We certainly see the temperatures rising and we see potential evapotranspiration increasing. The story of precipitation is a very very mixed bag to the present. The projections for the future are very regionally specific and the results from all the best modelers in the world continues to be mostly question marks as opposed to really definitive answers.

One of the few things that we can say right now about what global climate change, the greenhouse effect, means for surface water hydrology is the effect on hydrology in areas where snow and ice are a critical part of the hydrologic system. Mike Dettinger's presentation and Danny Marks. presentation yesterday were great examples of that, demonstrating that warmer temperatures mean more rain, less snow, earlier melt, and change in the pattern of runoff.

One of these things I find fascinating in all the studies to date on the effects of warming on these kinds of systems is that we have yet to identify any overall change in the amount of runoff coming out of these systems in addition to the change in timing. That timing change maybe very significant in terms of aquatic ecosystems and in terms of delivering water to downstream users because snowpack is storage, it's mother nature storage. We need storage in order to have a reliable water supply. We have less storage.

I think there are tremendous questions lying out there about what does this warming potentially means for ground-water systems in a whole variety of environments? And I think we haven't even begun to touch that issue. I think the need for monitoring particular places like mountain front environments is going to be a very crucial part of that.

We really have to say that we cannot accept the concept of stationarity any more. But, given that all of our thinking about the design of water resources systems was based on that, where are we? Are we completely at sea? Some might say we should just abandon the use of our historical records and not bother to look at them. But I think we would all agree that that's not the right answer and we've got to be informed by the past and informed by the kind of fluctuations that occurred in the past. But we need to be advancing the science. I'm going to quote here directly from a few paragraphs from this paper that we published in Science.

"In view of the magnitude and ubiquity of the hydroclimatic change apparently now under way, however, we assert that stationarity is dead and should no longer serve as a central, default assumption in water-resource risk assessment and planning. Finding a suitable successor is crucial for human adaptation to changing climate. . . Stationarity is dead because substantial anthropogenic change of Earth's climate is altering the means and extremes of precipitation, evapotranspiration and rates of discharge of rivers. . ."

"We need to find ways to identify nonstationary probabilistic models of relevant environmental variables and to use those models to optimize water systems. The challenge is daunting. Patterns of change are complex; uncertainties are large; and the knowledge base changes rapidly."

We go on and talk about the value of coupled atmospheric-land surface models. But we must be vigilant to relate those back to the actual data. We need to collect the data and we need to analyze them and compare them to the results we're getting from the models.

"Modeling should be used to synthesize observations; it can never replace them. Assuming climatic stationary, hydrologists have periodically relocated streamgages so that they could acquire more perspectives on what was thought to be a fairly constant picture. In a nonstationary world, continuity of observations is critical."

"The world today faces the enormous dual challenges of renewing its decaying water infrastructure and building new water infrastructure. Now is an opportune moment to update the analytical strategies used for planning such grand investments under an uncertain and changing climate."

I think the tremendous challenge for the academic community is that we really need something of the magnitude and intellectual vigor of the Harvard Water Program, which means not just hydrologists but again statisticians, economists, operations researchers and ecologists to help us come up with really a new paradigm.

What do I think are the rules for water management and how people should proceed today. In simple terms I'd like to say: number one, we can't have blinders, meaning we can't just say the future's going to look like the past. But we also can't have the arrogance to say that, "Oh, we have a GCM output and it tells us it's going to be 20% dryer over here or 5% wetter over there." I believe that the models are nowhere near to saying those things at this time in a meaningful manner. I mean we need to be humble, always humbled by these quasi-periodic phenomena going on around us, including the Ice Ages. Our decisions should be informed by history and by models, but not a slave to history or models. We need to admit to a wide range of options for what the future hydrology is going to look like and prepare for surprises and keep at the research.

It doesn't exactly sound like I'm talking about restoration but I come back to saying that most of the water planning that is going on today is in the context of restoring an aquatic ecosystem while maintaining an ability to deliver water to people and farms. So, this non-stationarity discussion is highly relevant to restoration which is at the center of planning for future supplies.

I'm going to move to my next topic (which is tightly related to the previous one) the subject of instream flow or ecological flows. As I said, the "fish has a seat at the table." Again, when people like me were educated, the question was how much water can we reliably withdraw from this river? That was the whole business of safe yield analysis.

Today's question is not how much water can we reliably withdraw from this river? Today's question is how much water do we need to leave in the river? I think science was needed then and science is needed now to deal with that. I would also argue that this issue of ecological flows (how much water needs to stay in the river) is the gridlock issue in water resource management decisions in the Nation. If markets for water are going to work and move the water around to its highest uses, there needs to be a clear set of property rights. As long as there is an unclear property right for the ecosystem then all other property rights to that water are going to be unclear. And the ability or willingness of people to make investments in that infrastructure with that unclear set of rights will be very limited. Our ability to make progress to satisfy all of the needs, ecological as well as human, I think it's going to be extremely limited.

I think there's a wonderful transformation occurring in the science and I like to simply characterize it as a shift from an old paradigm and a new paradigm with respect to just how much water we need to leave in the river. Part of the old paradigm was the "minimum instream flow." "Just tell me how many cfs I have to have going down this river in a critical time of the year. And that's going to be enough to wet the bed and have the right conditions for spawning and everything will be fine. Just give me that one number."

I think we all recognize today that that's not the answer. The answer is really about the whole range of flows and variability as being an important part of it. The difficulty is that the practitioners of ecological flows, and I have a lot of respect for those practitioners, is that this is still an art form and not a science. Which is: we want the hydrograph to look something like this. And it's not exactly what the original hydrograph was because we can never get back to the original hydrograph. But we are designing it as an art form and I would argue that we need to be pushing the science to get a better handle on what kind of variability we actually need.

The second aspect of the old paradigm was that the channel was a static entity. It has a width, it has a depth, and a substrate. We simply need to know how much water we have to send down that river. The new paradigm says that it's a dynamic channel. Its width can be changing, its depth can be changing, and it substrate can be changing, all as a function of the delivery of water and sediment to that system. I'll go back to one more Reds Wolman story. When he was a young scientist at the USGS, he and his colleagues, as Reds describes it, he said, "We sat around with our feet up on the desk talking about the process of meander migration. And we knew it existed and we knew that meanders moved downstream. But, we concluded that the process must be so slow that we will never observe it in our lifetime." So, Luna Leopold walked into the room and said, "Get your feet off the desk and go out and figure out a way to measure meander migration. It might be that slow but you've got to find out first."

They went out into a place called Watts Branch and they came up with some innovative techniques for measuring meander migration. And then fortuitously not long after they did their instrumentation there was a modest size flood that actually did an enormous amount of erosional work. Interestingly, the river channel looked exactly the same after this flood had happened as it had before. The channel cross-sectional areas were almost exactly identical to what they were before but everything had moved downstream.

And my point here, and Red's point when he tells this story, is "get out and measure it." Of course this goes back to Reds' work on the downstream effect of dams. We've got to be out there describing how our river channels are changing and building the modeling capability and the understanding to describe how they evolve over time as a result of the way we manage the river.

The third part of the old paradigm about ecological flows was that this was strictly a surface-water engineering problem and the way that you managed it was at the valve at the dam. How much water you released, maybe not just the valve at the dam but also the withdrawal by a city or an industry, which determines how much water we have going down the river. That was the question that needed to be answered.

The new paradigm says, yes, it's about surface-water engineering but also very, very, much about ground water, because of the effect the ground water has on maintaining base flow in rivers and the temperature and chemistry of instream flow. It becomes much more an integrated ground water and surface water problem, not just strictly a surface-water engineering problem.

Finally the old paradigm was that this was a single species problem. It got managed for a single species. We now recognize that, we've got to manage for multiple species sometimes with conflicting life-cycle needs. How to balance that out is going to be one of the great challenges in this field.

Lack of answers to these questions, I think, leads to gridlock. We need to build on the success of good work that has gone on in the past related to instream flow. We need to draw on biology, geomorphology, hydraulics, sediment transport, ground water hydrology and all these things. And integrate advanced measurement technologies such as GPS, acoustics, remote sensing and models in order to really help with this particular challenge.

The third topic I want to just mention quickly is non-point source pollution. Or another way to put that is some of the biogeochemical cycles in which man has played an enormous role in modifying. We're talking about nitrogen, we're talking about phosphorus, we're also talking about carbon. And also some things that exist out there in a smaller quantity but are still quite important: like mercury and sulfur and selenium and few others as well. But I'm going to confine my comments here to nitrogen. We could make similar statements about some of these others.

The Mississippi River Basin: humans apply 7 million metric tons of nitrogen to the land surface each year in chemical fertilizers. And there are several other sources of nitrogen not included here: atmospheric deposition, sewage, manure, and natural inputs from forests. The flux to the Gulf of Mexico is about 1 to 1½ million metric tons per year of nitrate to the Gulf measured as N.

Interesting numbers, 7 million applied to the land, 1 to 1½ going to the Gulf. Clearly not a conservative substance, there is a lot of processing, a lot of action whether in storage or modification (primarily denitrification) occurring along the way. So the need to understand what are the controlling factors on these processes is actually critical because these numbers suggest that this is something that we can do something about because clearly nature is doing a lot in the way of processing that along the way.

What have been the changes over a hundred year timeframe? And I come back to some of the names I mentioned in the beginning, Dole and Stabler. I thank my lucky stars on a regular basis, in fact I have a picture of Stabler on my computer, as I always want to thank him. These guys went out and said, "We're going to measure this stuff," back in 1905, 1906, and now we have good data points back there. Not just one observation but a set of observations over the annual cycle and good interpretation of those results.

We now know that the nitrogen flux of the Mississippi River is roughly speaking 2.6 times the level it was at the beginning of the 20th century. And there are some tributaries of the Mississippi River, like the Cedar River in Iowa, and the Minnesota River in Minnesota where the increase is as much as six-fold or even eight-fold over what they were at the beginning of the 20th century. These are enormous changes in our environment with a very important, biologically extremely important, substance, nitrogen, which also has a significant public health effect in the form of nitrate.

We also know that there are tremendous year-to-year variations in the flux of nitrogen down our major river systems. 1988, the year of a great drought, the Mississippi River Basin carried something like 0.5 million metric tons downstream. 1993 the year of the great flood, 1.6 million metric tons. So we see three-fold differences year-to-year purely driven by the vagaries of natural meteorological variation.

The ability to say, what's the effect of man and what's the effect this natural variation? That is one of the great challenges for hydrologic science. In systems like the Mississippi River or Chesapeake Bay or many others around the country, the Nation is determined to try to do something about this problem of nutrient enrichment of our Nation's rivers, and ultimately our lakes and estuaries, and we need to understand the processes and the management of these processes so that can do something about it. I think it's a critical challenge for hydrologic sciences. Understanding the signal-to-noise problem and explaining that signal-to-noise situation to the public. I think is one of our great challenges.

We also have a need to understand the time lags in these systems. We have some idea that, for example in the Chesapeake Bay watershed; that if you could cease the input of nitrate at the land surface instantly today; it would take at least a decade before you see really significant changes in the amount of nitrogen reaching the Chesapeake Bay. So much of it is traveling through the shallow subsurface; this is one of the things that we are discovering now. I think of these things like the nitrogen flux down a major river system as being one of our analogs to Keeling's CO₂ curve but it is much tougher, because there is tremendous amount of natural variability, not the very clean simple picture that the CO₂ curve has.

We need to track it. We need to understand. We need to describe the sources of variation and we may need to help the public to understand it. And this is not just the problem in United States, but it's a global problem of nutrient over-enrichment with dead zones all over the world associated with this enrichment.

Our fourth topic is what I call causes of aquatic ecosystem decline. I'll come back to the Clean Water Act and just remind you that the Act said that every city and every industry needs to clean up to a certain technological standard and then we're going take a look at the water quality and say, "Does it meet our desires for what water quality should look like in this river?" If it doesn't, then we are going to try to do something about that.

We cleaned up to the basic level of technology. It's interesting to note that this first phase of Clean Water Act basically said, "We don't need any science at this point in time." It's all about the technology of cleaning up the point source pollution. In fact I would argue that the science went rather dead for almost 20 years because it played no role in the nation's management of its surface water quality. But at the end of that era of cleaning up the point sources, we recognized "You know what? There's still a problem in the rivers and we've got to do something about it."

The provision of the Clean Water Act to deal with this is called the TMDL provision. Some people think that stands for "Too Many Damn Lawyers." It actually stands for "Total Maximum Daily Load." It says you've got to go back to look at the watershed and figure out where this came from; which tributary, which source, was it the atmosphere, was it this point source, etc. And then you've got to come up with a plan to do something about it. Well, it's a huge political problem once you've figured that out and actually have a plan. Making it happen can be very difficult. But there's a tremendous amount of science that needs to be done even in getting up to that.

We look at our rivers and we say, "We as a society are not satisfied with the biotic condition of the river in a particular location. The fish that we want to have there are not there, we can't swim in it, the mussel population is gone or whatever it might be. We're not satisfied." The question is, why. What's impairing it? What are the causes of the impairment? The potential list is enormous. It could be the flow, the thing that I've just talked about. It could be photosynthesis and algal populations and hypoxia. Those are possible explanations. It could be toxic chemicals that are killing them. It could be construction of physical barriers. It could be invasive species whether plants or animals. It could be disease. It could be harvesting.

I think one of the big questions out there is this: Is it chemical substances which are not, in and of themselves, lethal to the organisms? There are no dead bodies lying around or fish that died because of these chemicals but these chemicals are in fact affecting the reproductive capability of the species. In particular, I point to the hormones which come through the medicines that we take, through the natural processes of our body, through the pharmaceuticals we flush down the toilet, and through the pharmaceuticals that are used in animal feeding operations which we know are entering our waters.

This is a tremendously difficult scientific question because, yes, we know these substances are present, but present at minute levels which are clearly not toxic. They are not killing things by and large but we have very good reason to believe that they have some reproductive effects. How do you measure that and how do you set the regulatory and management process in place that will control these things?

So, the concern is that in spite of the desire to improve the ecosystem of a particular river, we can select what we think is our reason for why it is impaired and we can set about to spend hundreds of millions of dollars to clean up that impairment. And we might very well be going after the wrong cause and after all that expenditure we discover that the system is still impaired, and we need to go back and look at it from another perspective.

I think this is an issue where scientific stove-piping is a really crucial thing. We all come at it from the biases of the field that we were trained in. If your field is photosynthesis and algae then you are going to work with the algae. If you're trained like me, you'll be thinking about river channels, you can think it has to do with flow and substrate. If you're trained in fisheries biology, you may think about harvest etc. We all come at it with prejudices of our own subfield and we need to be aware of interdisciplinary science to look open-mindedly at all possible causes of this ecosystem degradation. I think it's critical so that you can identify the right cause and the right solutions to these problems and really does demand a very interdisciplinary look.

My final topic, is carbon sequestration and I want to bring this up from two perspectives where hydrology fits into this topic of sequestration. The first one is geologic sequestration of carbon dioxide. The Department of Energy right now has enormous plans just at the experimental pilot plant level. The idea is to scrub the CO₂ from our power plants and inject it into the ground and put it into permanent storage in the subsurface.

Some of the proponents will just say, "Well you know, we know how to do this. We've been injecting CO₂ into the ground in oil fields for years. It's straightforward. We know how to do it, so let's get on with it." And this is a significant way in which we can contribute to at least slowing the growth of the greenhouse gases in our atmosphere. The projections of how much CO₂ we would have to inject to have a significant impact from just the stationary sources of CO₂ in the Nation, it is on the order of 3.6 billion tons per year of CO₂ we would have to be injecting into the ground. At the present time or in recent years, we've been injecting something like 30 million tons per year of CO₂ into oil fields. So, we're talking about an injection of about 120 times more CO₂ than what we've been doing in the past. This is a supercritical fluid. It would create tremendous pressure waves, have effects on the aquifer matrix, effects on the integrity of wells, and geochemical changes associated with the introduction of the supercritical CO₂ at the point at which it mixes with water creating carbonic acid.

What are the effects on the aquifer and what are the far-field effects on the saline aquifers that we may be trying to inject it into in terms of the displacement of the saline water that is already in those systems? I think this is scientifically really exciting challenge for the ground-water modeling community to think through and then develop schemes for monitoring these processes because I think the experiments will go forward, as I think they should, to try and find out how to do this and how to do it safely.

The other aspect of research on change in sequestration is the role of terrestrial ecosystems and carbon. We know that there's a lot of carbon moving in our ground water and in our surface-water systems and in agricultural activities, and the coupling with the nitrogen and phosphorus is really crucial.

A recent paper in Global Biogeochemical Cycles identifies small farm ponds, aggregated over the entire earth, as a tremendous site for the burial of carbon by these processes because we have very organically rich material and it's going into storage in the long term. So, that the role that hydrology can play in understanding the potential for the burial of carbon. I think is absolutely very critical.

This is where I'm going to be a bit more provocative and say that sometimes I have the feeling that hydrology is kind of a bit player in the greenhouse warming inquiry; that we're just a tail on this big atmospheric and oceanic dog and we are just looking at subtle effects which are coming from greenhouse warming. An alternative is really a hydrologic focus on what I call the great global fertilization experiment that's going on. The fertilization is with carbon, nitrogen and phosphorus primarily where mankind over the last century has dramatically changed the way these elements are distributed on the planet and all of them relate back to the carbon cycle and how the carbon is sequestered.

Many, many consequences are associated with this global fertilization experiment and only one of the consequences is this climate change phenomenon. There are many others, particularly the ecosystem-oriented ones that I think are critical. I want to go on by quoting from someone who I've admired for a long time, not a hydrologist but actually a physicist, Freeman Dyson, and some of his words about the subject of carbon and global change.

He started writing about this in 1978 in a commentary to the Department of Energy. He followed up and I'll read to you in some words from a paper he wrote in 1990 and he continues with more recent papers even in the last year or two. In 1990 he said that 12 years had gone by since his words about how the real issue is what's happening in carbon and his argument that this climate thing is the second order issue.

The first order issue is what happens to the carbon. It has been "12 years . [since] these words were written and nothing much has changed. The 10 percent enrichment of atmospheric carbon dioxide has grown to 15 percent." Of course today, it's about 25 percent. Continuing to quote Dyson, "The financial resources of government research programs investigating the carbon dioxide problem have grown even more rapidly. But the increased funds have mostly been poured into computer simulations of the global climate, rather than into observations of the real world of roots and shoots, trees and termites."

Dyson continues: "I do not blame only the government bureaucrats for the excessive emphasis on computer simulations. We scientists must share the blame. It is much more comfortable for a scientist to run a computer a computer model in an air-conditioned supercomputer center rather than to put on winter clothes and try to keep instruments correctly calibrated outside in the mud and rain. Up to a point, the computer models are useful and necessary. They are only harmful when they become a substitute for real-world observations. In the twelve years since 1978, the results of computer models have tended to dominate the political discussion of the carbon dioxide problem. The computer results are simpler and easier for politicians to understand than the vagaries of the real world."

I'd like to go on to quote Dyson just a little bit more. In a recent talk on the subject he said "The problem of CO₂ in the atmosphere is a problem of land management and not a problem of meteorology." I think that's an overstatement, but I love the kind of bold way in which he states it. "Global warming causes problems indeed but global warming as a scientific focus, takes away money and attention from other problems that are more urgent and important."

These are now my words: Within the field that we deal with, there are issues of resource depletion and its impacts on water for human use for future generations and for ecosystems, issues of nutrient enrichment, issues of habitat and reproductive health of organisms and many others. I'll quote Tracy Mehan, former Assistant Administrator for Water at EPA, who said that sometimes the subject of global climate change is "sucking all the oxygen of the room," when it comes to discussions of environmental and natural resource issues.

But, don't say, "Oh Bob Hirsch is a climate change nay-sayer." What I'm saying is: think about balance. Think about of the relative importance of the issues, and the issues which we can make a contribution to, which are of global, national, and regional importance and I will argue there are many, and that climate is just one of these issues.

I think we have a responsibility, and I'm going back to the words of Ralph Keeling, "to record the earth's vital signs." To measure the movement of water and its chemical constituents and the biota it supports. All of these things are really part of the vital signs of the Earth. I'll quote another one of my heroes, Bob Meade, retired from the USGS. I spent a few days with him on the Mississippi River collecting data several years ago.

And he commented "You know the Mississippi River is the urinary tract of the Nation and doctors check our urine to find out what's going on in our body. We need to be looking at our rivers, large and small, and seeing what's in them in terms of sediments and chemistry to understand what's really happening in our landscape." We need to learn from it, learn about its health, in measuring and understanding what's going on in our hydrologic system. Bob Meade always points out that this idea is not original with him, but comes from tropical ecologist Harald Sioli, who said "River water is the urine of the landscape."

Keeling said "The only way to figure out what's happening to our planet is to measure it. This means tracking changes decade after decade and poring over the records." Society needs to understand what things need restoration and we, the hydrologic community, needed to point the way to this restoration.

We need to be engaged in planning it, observing it, understanding it, and using the knowledge that we gain from these things that are being done for restoration to enable us to do it even better in the future. I think this is a very exciting time for hydrology and I congratulate CUAHSI for the contribution it makes and for the fact that CUAHSI has spoken very loudly of the importance of measurement and the field component of hydrologic sciences. I salute you and I'm glad to be present for this occasion. I thank you.