New stationarities: avoiding problems in the solutions
06/03/12 13:27
Most of the things that are wrong with water are easy to identify: the massive quantities of largely untreated sewage, industrial pollution that has been the legacy of the industrial revolution worldwide, chemical fertilizer and pest management runoff that is the legacy of the agricultural revolution in the most productive countries, building “bad dams” that are designed and/or operated in ways that significantly an negatively alter the ecosystems and livelihoods of rivers, invasive species, and the overconsumption and diversion of water resources, killing rivers for great lengths or draining lakes and marshes into cities, fields, and factories.
You could call these “first-order problems” with managing water.
You could call these “first-order problems” with managing water.
They largely consist of issues with managing water quantity and quality. And most of them are pretty straightforward to correct as well. The main obstacles in correcting them come from (a) prevention, and (b) political and regulatory will. In some cases, such as very bad dams or very severe pollution, the issues may develop a life of their own so that you end up with a Hudson riverbed full of PCBs or a Caspian sea, dead and hypersaline from the combination of massive upstream diversions and concentrated agricultural runoff. Rarely do first-order problems surprise you. They take time and energy and attention to build up a head of steam.
This is not to make light of first-order problems. They are certainly the dominant issues in most of the developing world. They were even dominant in much of the developed world through the 1960s and 1970s, when rigorous and functional regulatory frameworks became widespread.
Two advances followed this period from sustainability and ecological perspectives. Freshwater ecology has always lagged terrestrial and marine ecology in its scope and sophistication, but the 1980s and 1990s saw an acceleration of work that showed how flow regimes — the natural, seasonal periods of high and low water, as well as the frequency of extreme events such as very hot and cold water and droughts and floods — were critical to maintaining healthy ecosystems. Together, this work showed that inter- and intra-annual variability were the “master variables” for lakes and rivers and wetlands. The keystone concept from the 1990s was “the natural flow regime,” a signature often-cited paper by LeRoy Poff. Lakes and wetlands had a comparable concept: the natural hydroperiod.
Together, these concepts meant that water quality and quantity were less important to focus on in isolation. Indeed, quantity and quality were really determined by the flow regime. Sustainable water advocates pointed out that mainstream infrastructure management tended to level out, dampen, or even eliminate this seasonal and inter-annual variation. Big dams, reservoirs, or diversions that effectively turned freshwaters into invariant pools of water, profoundly disrupting ecosystems and a wide range of geophysical processes. Their suggestion was to develop new virtual “software” for the infrastructure “hardware” that altered how engineers operated and managed dams and networks of dams. This approach would try to mimic the natural flow regime in such a way that the normal seasonal variation could re-created through water releases. This management system has been called environmental flows or e-flows and has achieved significant (yet still uneven) adoption in the developed world but much less acceptance elsewhere.
These advances represent a major new era in water management, but the growing awareness that climate plays in driving the water cycle — all of the variables of the natural flow regime and hydroperiod — meant that climate change could undermine even these new approaches. This realization has come slowly and with great reluctance in some parts of the water community. And this has led to second-order problems with water management.
The basic issue around climate and water management was most notably identified in a paper by a blue-ribbon set of authors by Milly et al., called Stationarity Is Dead: Whither Water Management? Published in 2008, the paper introduced the term stationarity to a wide range of the water community. Most attention on the paper is focused on the first few paragraphs, which can be summarized as pointing out that water managers have long assumed that while the past is a good guide to the future, climate change (especially periods of relatively quick climate change, such as now) means that even good, long records of the past are not necessarily useful guides to what the future will look like. (There is a technical statistical definition of stationarity too that is the original source of the new climate-context meaning.)
This is a serious problem for many water managers, who need to develop systems and infrastructure that can perform consistently for decades. A dam needs to fit its climate like a key fits a lock. Such projects often need good evidence that the operational parameters for the kinds of hydrological conditions will be experienced. In effect, they need good numbers for such qualities as precipitation, flows, groundwater recharge rates, expected flood/drought intensity and frequency, and so on. All of these numbers ultimately reflect the ambient climate, so a shifting climate (with low certainty) makes traditional approaches to water management very difficult to implement. Stationarity also means that even advocates for what was viewed as sustainable water management are confronted with learning that most places where e-flows has been implemented have created a “new stationarity” — managing the natural flow regime as if it is not a reflection of climate or responsive to climate change.
This limitation is extremely problematic. Water infrastructure tends to be both critical and invisible to economies, since it provides water for energy, cities, agriculture, industries, and so on. Lose it, or lose its reliability, or lose the ability to build infrastructure that should last effectively for decades, and you have also knocked out the “stationarity” of economies as well. A dam ill-suited for its climate will likely stress and trash out the ecosystems all around it, even if it has been designed to be otherwise sustainable. Worse, the infrastructure will increasingly diverge with its climate over time.
In most cases, these problems may have a long fuse before being expressed on the same scale as with first-order problems, but they are still present and active. And in regions where the pace of climate change is very accelerated, such as in high-latitude and high-altitude areas, we are quite likely to be building new dams that have already lost their sync with their ambient climate.
These issues remain unsettled. Many who once saw themselves as the vanguard of thinking around sustainability feel undermined by critiques that they have created new problems. At the same time, active thinking and research has been emerging about how to develop get around these new stationarities.
Given the low confidence in predicting future water cycle conditions, three general approaches seem to be taking the lead right now:
This is not to make light of first-order problems. They are certainly the dominant issues in most of the developing world. They were even dominant in much of the developed world through the 1960s and 1970s, when rigorous and functional regulatory frameworks became widespread.
Two advances followed this period from sustainability and ecological perspectives. Freshwater ecology has always lagged terrestrial and marine ecology in its scope and sophistication, but the 1980s and 1990s saw an acceleration of work that showed how flow regimes — the natural, seasonal periods of high and low water, as well as the frequency of extreme events such as very hot and cold water and droughts and floods — were critical to maintaining healthy ecosystems. Together, this work showed that inter- and intra-annual variability were the “master variables” for lakes and rivers and wetlands. The keystone concept from the 1990s was “the natural flow regime,” a signature often-cited paper by LeRoy Poff. Lakes and wetlands had a comparable concept: the natural hydroperiod.
Together, these concepts meant that water quality and quantity were less important to focus on in isolation. Indeed, quantity and quality were really determined by the flow regime. Sustainable water advocates pointed out that mainstream infrastructure management tended to level out, dampen, or even eliminate this seasonal and inter-annual variation. Big dams, reservoirs, or diversions that effectively turned freshwaters into invariant pools of water, profoundly disrupting ecosystems and a wide range of geophysical processes. Their suggestion was to develop new virtual “software” for the infrastructure “hardware” that altered how engineers operated and managed dams and networks of dams. This approach would try to mimic the natural flow regime in such a way that the normal seasonal variation could re-created through water releases. This management system has been called environmental flows or e-flows and has achieved significant (yet still uneven) adoption in the developed world but much less acceptance elsewhere.
These advances represent a major new era in water management, but the growing awareness that climate plays in driving the water cycle — all of the variables of the natural flow regime and hydroperiod — meant that climate change could undermine even these new approaches. This realization has come slowly and with great reluctance in some parts of the water community. And this has led to second-order problems with water management.
The basic issue around climate and water management was most notably identified in a paper by a blue-ribbon set of authors by Milly et al., called Stationarity Is Dead: Whither Water Management? Published in 2008, the paper introduced the term stationarity to a wide range of the water community. Most attention on the paper is focused on the first few paragraphs, which can be summarized as pointing out that water managers have long assumed that while the past is a good guide to the future, climate change (especially periods of relatively quick climate change, such as now) means that even good, long records of the past are not necessarily useful guides to what the future will look like. (There is a technical statistical definition of stationarity too that is the original source of the new climate-context meaning.)
This is a serious problem for many water managers, who need to develop systems and infrastructure that can perform consistently for decades. A dam needs to fit its climate like a key fits a lock. Such projects often need good evidence that the operational parameters for the kinds of hydrological conditions will be experienced. In effect, they need good numbers for such qualities as precipitation, flows, groundwater recharge rates, expected flood/drought intensity and frequency, and so on. All of these numbers ultimately reflect the ambient climate, so a shifting climate (with low certainty) makes traditional approaches to water management very difficult to implement. Stationarity also means that even advocates for what was viewed as sustainable water management are confronted with learning that most places where e-flows has been implemented have created a “new stationarity” — managing the natural flow regime as if it is not a reflection of climate or responsive to climate change.
This limitation is extremely problematic. Water infrastructure tends to be both critical and invisible to economies, since it provides water for energy, cities, agriculture, industries, and so on. Lose it, or lose its reliability, or lose the ability to build infrastructure that should last effectively for decades, and you have also knocked out the “stationarity” of economies as well. A dam ill-suited for its climate will likely stress and trash out the ecosystems all around it, even if it has been designed to be otherwise sustainable. Worse, the infrastructure will increasingly diverge with its climate over time.
In most cases, these problems may have a long fuse before being expressed on the same scale as with first-order problems, but they are still present and active. And in regions where the pace of climate change is very accelerated, such as in high-latitude and high-altitude areas, we are quite likely to be building new dams that have already lost their sync with their ambient climate.
These issues remain unsettled. Many who once saw themselves as the vanguard of thinking around sustainability feel undermined by critiques that they have created new problems. At the same time, active thinking and research has been emerging about how to develop get around these new stationarities.
Given the low confidence in predicting future water cycle conditions, three general approaches seem to be taking the lead right now:
- Expand the concepts of the natural flow regime and e-flows and make managed water resources more dynamic. Several basins, for instance, have been experimenting with short time re-allocation cycles for e-flows that “update” the software for managing the infrastructure — such as every four to six years. Tanzania is doing some excellent work here around the Pangani basin. This is a type of “manual re-load” for the software. In effect, this is focusing on the stakeholders and institutions that collectively manage a particular basin and its infrastructure.
- Develop a set of ecological and hydrological markers that allow water managers to determine what maintains and cradles the “resilience” and health of the system. This is a more technical approach. Ideally, this approach would mean facilitating how a particular freshwater ecosystem evolves through time rather than remaining “stationary” in a climate that has probably long passed. One way to pursue this route in practical terms is to begin to blend ecosystems into built infrastructure networks, so that wetlands rather than reservoirs begin storing water.
- The simplest and most engineered method may be to consider building in stages, with these stages lasting for decades. This approach assumes that big infrastructure is at greater risk for diverging from its climate. If the Three Gorges Dam or Hoover Dam were built as a series of dams over fifty years, with updates to the design and operations included to better track the shifting climate, would they be more effective?
blog comments powered by Disqus
