If you do not know much about America’s network of streamgages or, like many people, are unaware of the network’s existence, Greg Stewart is glad to fill you in. Stewart is data section chief at the U.S. Geological Survey’s Water Science Center in Augusta, Maine, and getting him to talk about streamgages is about as hard as getting Al Gore to talk climate change. A little prompt and the motor is running.
“It doesn’t take much to get me talking about this program,” Stewart, 39, said, during an hour-and-a-half drive from Augusta to a streamgage on the Piscataquis River near the town of Dover-Foxworth in central Maine. “Because I really love what I do.”
“The other day, I was looking at these typed, carbon-copy letters sent by our office in 1912 to a landowner, telling him how they’d built a gage house in his backyard,” Stewart said. “It’s incredible to be part of that history.”
The streamgage network does indeed have deep roots. The USGS, which operates and maintains the gages, set up its first station for monitoring a river’s flow in 1889 on the Rio Grande River in New Mexico. (The USGS tradition of spelling gauge as “gage” apparently began three years later, when the agency’s chief hydrographer is said to have left out the “u” in deference to the original Saxon spelling.) In 1901, USGS established its first New England streamgaging station on the Kennebec River at The Forks, Maine. From these beginnings, a vast system has emerged: currently, there are more than 7,600 automated streamgages across the country, constantly collecting information about river conditions for multiple uses, including forecasting floods and droughts, allocating water resources, designing bridges, operating dams, even measuring the effects of climate change. For so much of the nation’s work where water is involved, the data generated by the gages provide an indispensable foundation.
“The only way to try to get to a sustainable future is with good planning,” said Peter Evans, executive director of the Interstate Council on Water Policy. “And good planning starts with good measurements.”
The making of those measurements comes with a price of course, as Evans knows all too well. For years, he and ICWP have been pushing for greater federal funding of streamgages, with NEIWPCC’s support. But progress has been slow. For now, state, tribal, and local agencies are continuing to pick up more than their expected share of the tab, with predictable results. The USGS reports that from 1995 to 2008, 948 critical streamgages with 30 or more years of data records were discontinued. The main reason: a lack of funding. And unless that funding gap is addressed, more key gages will be shut down in the coming years.
During the drive to the Piscataquis River, Stewart explained that in Maine, the installation of new gages has largely offset the losses, so the total number has remained fairly steady. That has been the case nationally as well. But a new gage does not have the same history, the same years of valuable data, the same scientific memory, as an old one.
“It’s sad to see streamgages go,” Stewart said.
So far, the gage at the Piscataquis River has been spared, and its 108-year-long streak of providing data remains intact. Visual evidence of that long history was evident as Stewart pulled up to the site. A classically picturesque New England covered bridge spanned the river, and its presence is no coincidence. In the early days of streamgaging, USGS hydrologists measured the river’s water level, or stage, by standing on the bridge and performing what is known as a tape-down—lowering a steel tape into the flow and noting the depth.
The gage house also spoke to years of use. A weathered cement structure, it stood stolidly like a sentry on the riverbank, looking immutable. But its days are numbered. The USGS is knocking down these relics and replacing them with protective housings just large enough to contain the equipment, because in many cases, the gage houses no longer serve their original purpose; the huts provided shelter for hydrologists working on stilling well systems, in which underground pipes connected to the river allow water in a well directly beneath the hut to be at the same level as the river’s water surface. Use a float to measure the level in the well, and you have your stage reading. The system is ingenious, but also problematic. Stilling wells are expensive to install, the pipes rust, the wells can fill with sediment. In Maine, as in much of the rest of the country, the USGS is moving to modern alternatives.
At the Piscataquis River site, Stewart squeezed his strapping 6’7” frame into the hut to explain the technology inside. Data on the river’s stage are generated by a device based on a simple concept—the deeper the water, the greater the pressure. The device sends air down a small-diameter tube into the river, and by measuring the amount of pressure it takes to push the air bubbles out, the level of the water above the tube can be determined. While not maintenance-free, at least corroding pipes are not a worry. What happens next with the stage data is even more impressive.
“The data collection platform is the brain,” Stewart said, pointing to a WaterLOG H-522, which combines a data logger with a built-in system for transmitting readings via satellite. Once an hour, the H-522 sends the Piscataquis River stage data on the same path followed by data from all the nation’s streamgages. First, the information goes to a Geostationary Operational Environmental Satellite (GOES), which, because it orbits the earth at the same rate as the earth’s rotation, stays above the same point on the earth’s surface at all times, meaning expensive tracking antennas are not required. The GOES transmits the information to a giant receiving dish at the National Oceanic and Atmospheric Administration facility in Wallops Island, Virginia, whereupon the information is immediately sent back into the sky at a much higher power to a domestic communications satellite that bounces the data down to one of 21 smaller Local Readout Ground Stations that USGS maintains. (Maine’s data goes to an LRGS in Pennsylvania.) From there, the data moves on to USGS’s National Water Information System or NWIS, where after processing (more on that in a moment), the information is posted on the USGS website delivery system (http://waterdata.usgs.gov/nwis).
“I can see the data in our database in about 90 seconds [from the time it leaves the gage house],” Stewart said. “The website doesn’t update quite that fast. It can be 15 or 20 minutes.”
So, anyone with an Internet connection, sitting anywhere in the world, can access the latest data on the Piscataquis at Dover-Foxworth, or any other streamgage for that matter, within roughly a quarter-hour after measurements are taken. This rapid access is a technological feat that would stun Stewart’s predecessors, but it is not being done to show off. Users of the data—especially state emergency managers trying to predict floods and issue timely flood warnings—benefit immensely from monitoring river conditions in real-time without leaving their desk. What most users need to monitor, however, is not necessarily a stream’s stage; what matters most is the total amount of water moving down a river, known as streamflow or discharge. That is where the processing at NWIS comes in.
At NWIS, so-called stage-discharge rating curves are used to convert streamgage stage data to the streamflow numbers seen by the world on the Web. These rating curves show what the discharge is in cubic feet per second (CFS) for every conceivable level of stage reported by a gage. Developing these curves is a big part of a USGS hydrologist’s life. Bear in mind that to determine discharge, you need to know the area of the water at the point in the channel where data is collected—and to get area, both height (stage) and width must be known—plus you must know the water’s velocity. Since streamgages measure only one part of that equation (stage), hydrologists routinely hit the water to make manual discharge measurements at gage sites so they can ultimately correlate the streamflow associated with each and every stage level—and in the process, build the all-important rating curves used by NWIS.
The tricky part is that, since river channels are prone to change, a curve that is accurate one year may not be so accurate the next. Ice moves rocks in a stream around; sand or gravel can be scoured off a bank and into a riverbed; dunes and bars shift routinely. All that can throw off a rating curve, in that stage levels no longer equate to the same amount of discharge.
Stewart explained that natural variation in a channel’s configuration is a big reason each of Maine’s 70 active streamgages is visited by a member of his six-person streamgaging team at least six times a year and sometimes more than 20 times.
“We’re going to make sure the stage is reading correctly,” Stewart said, “but we’re also making sure that the stage versus discharge relationship is still valid. Things change on a regular basis. That’s why we’re out there making measurements.”
In measuring discharge manually, the USGS relies on techniques and tools, such as mechanical current meters, that fundamentally are not much different from those employed more than 100 years ago. But modern technology is moving in. Hydrologists increasingly use sophisticated devices such as hydroacoustic meters, which determine water velocity by sending pulses of sound through a stream to measure the speed of solid particles moving with the flow. Use of these innovations has improved discharge measurements, and reflects USGS’s overall embrace of technology in pushing the science of streamgaging forward.
This embrace was hard to miss at the Piscataquis River. You could see it in the complex equipment inside the gage house, in the solar panel affixed to the side. You could hear it in Stewart’s voice as he spoke of his office’s new practice of using handheld computers or PDAs to connect to a data collection platform during a streamgage visit, so the data inside can be downloaded and then dumped directly into a database back at the office.
But this progress does not come cheaply. Nor is it inexpensive to send hydrologists out on the road into rural Maine to fine-tune rating curves or to race out to fix broken gages within 24 hours of noticing a problem during the daily check of all gage data. Stewart estimates that each streamgage in Maine costs $12,000 annually to operate and maintain. And that is one of the lowest estimates in the country. In Rhode Island, the cost estimate is $12,900. In Massachusetts: between $15,000 and $17,000. Cost-of-living adjustments contribute to the difference regionally, but sometimes the gap is a matter of geography. In Alaska, some streamgaging stations cost more than $50,000 a year. To service them, hydrologists have to fly in via seaplanes.
But as Stewart knows, even Maine’s comparably paltry gage cost still sounds high.
“We do understand that it’s very expensive,” Stewart said. “We’re not living in an ivory tower saying ‘Oh, well, this is what you have to pay.’ We’re very conscious of the cost. But in order to meet our criteria, to do it to our standards, it really does cost that much.”
Struggle for Support
The funding to cover the cost of running the nation’s streamgaging network comes from multiple sources, one of which has been around a very long time. More than 110 years ago, the USGS started its Cooperative Water Program (CWP), which supports streamgaging and other water studies, and over the years, the underlying rationale for the program has remained constant: since government agencies at all levels need waterresource information, then the smart move is to pool financial resources, and let the experts—USGS—do the work and keep the results in one database accessible to all. The CWP was designed to be a 50:50 costshare partnership, with the federal government (through the Interior Department, which oversees USGS) kicking in half the monies needed and local and state agencies, or “cooperators,” appropriating the rest. In recent years, it has not been so even. According to CWP budget figures on USGS’s website (http://water.usgs.gov/coop/description.html), cooperators are picking up roughly 64 percent of the CWP’s total funding. That percentage may be moving higher. While the fiscal 2011 federal budget was still in a state of flux as this IWR went to press, President Obama’s proposed FY11 budget calls for $63.6 million in federal monies for the CWP, $2 million less than fiscal 2010. (Typically, about 40 percent of the CWP’s budget is directed at streamgaging.)
The picture is no brighter for another key source of federal funds for streamgages—the much younger USGS National Streamflow Information Program. USGS initiated NSIP in 2001 after Congress raised concerns about a big decline in the number of streamgages in the mid-1980s just as demand for streamflow information was increasing. To meet NSIP’s primary goal of establishing a stable “backbone” network of streamgages, USGS identified 4,744 core gages around the country that the agency would fully fund to ensure they were not compromised by the vagaries of fluctuating funding from multiple partners. Nice idea—but the execution has fallen short. In its 2009 NSIP Implementation Status Report, USGS said it was operating 2,940 or 62 percent of the backbone gages. And the report did not shy away from the cause of the problem, noting that in the 2009 federal budget, NSIP was funded at $22.4 million—a far cry from the $117 million per year USGS says it needs to fully implement the program. NSIP did get a $5 million increase in 2010, but the president’s FY11 budget is proposing $27.2 million, a drop of roughly $500,000.
The failure to fully fund NSIP means USGS continues to rely extensively on other sources, such as the roughly $30 million a year provided for streamgages by other federal agencies, primarily the Army Corps of Engineers and Bureau of Reclamation. It also keeps the burden on the CWP, and in particular the cooperating state and local agencies that, for now, continue to help sustain so many of the nation’s streamgages. CWP funds, for example, support the Piscataquis River gage, with USGS and the cooperator, Maine’s Department of Transportation, splitting the cost. But as the economic malaise continues, streamgage proponents fear cooperators—with budget problems of their own—will be unable to continue to shoulder more than their fair share of the CWP’s financial requirements.
“What’s going to happen is that [the cooperators’ contribution] is going to shrink,” said Peter Evans of the Interstate Council on Water Policy, “and if USGS is not in a position to step up its investment, we’re going to see an acceleration in how fast we lose streamgages.”
Evans estimates that between 100 and 150 gages are being lost every year, and his belief that the nation needs more gages, not less, has motivated him to address the issue at hearings in Washington and to lead ICWP’s extensive efforts to increase federal support. NEIWPCC Deputy Director Susan Sullivan, in her role as ICWP’s chair, has joined Evans in the effort, signing an ICWP letter in April 2010 that urged the chairman and ranking member of the House Subcommittee on Interior, Environment and Related Agencies to back substantial increases for the NSIP and CWP.
At NEIWPCC, we have also acted on our own, with Executive Director Ron Poltak sending a similar letter in May 2009 to the same subcommittee. But the efforts have had little impact. Evans said politics play a part.
“It goes back to the 24-hour news cycle, and the importance of sound-bite salesmanship,” Evans said. “If the Secretary [of the Interior] thinks, ‘Well, if I’m going to Congress to show them that we’re doing things that their constituents really value, am I going to ask for $100 million for streamgages or am I going to ask for $100 million to involve kids in natural resource programs? Am I going to put us on a schedule for completing the implementation of the data collection system that we’ve been talking about for a decade, or am I going to ask them to put funding into a plan to fully implement President Kennedy’s vision for a land and water conservation grant program?’ The data collection just doesn’t sell as well.”
In a NEIWPCC interview with Anne Castle, the Interior Department’s Assistant Secretary for Water and Science (seen on page 8 of the Fall 2010 IWR Newsletter), she disputed the notion that something as ostensibly mundane as streamgaging does not sell in Congress. Castle also said that while there could always be more gages, the existing number is “adequate.” In Maine, Greg Stewart assessed his state’s situation differently.
“I would say we’re definitely short streamgages,” Stewart said. “There are very, very obvious gaps in our dataset. I really think our network should be about 120 to 150 streamgages.”
That is about double what Maine has now—but far short of the titanic total Stewart would need to have a gage in every flowing body of water in Maine’s 33,264 square miles. Such complete coverage is unrealistic even in far smaller states, so to fill in the blanks, the USGS uses a statistical tool known as regression analysis to develop mathematical equations to estimate streamflow in ungaged streams. These equations are built using data from gaged waters, and the more streamgages a state has, the more accurate the estimates. If there are not enough gages to represent all basin and climate conditions in a state, hydrologists are hamstrung in their ability to do a comprehensive job.
“We really can’t generate equations for small watersheds, we really can’t,” Stewart said. “We just don’t have enough streamgages.”
In Massachusetts, where 17 streamgages were shut down last year alone, you might expect to hear similar concerns. But Gardner Bent, a hydrologist at the USGS Water Science Center in Northborough, Mass., said the state still has “fairly good coverage.” Though he added the impact of the closings will be felt—eventually.
“The effect is more long-term,” Bent said. “Years down the road, [the lost gages] would help us in developing equations to predict low flows or flood frequency.”
With serious consequences like these, and the unlikelihood of a substantial federal funding increase anytime soon, USGS has been looking into ways to make its hydrologic data collection programs more efficient—and that includes seeking help from the private sector. In the fall of 2009, a team made up of USGS staff and representatives from two environmental equipment-makers, YSI and In-Situ, visited the streamgage at the Piscataquis River and a gage station in Colorado to observe and analyze the hydrologists’ activities. The team based its evaluation on a set of industry principles for identifying and eliminating waste. In a May 2010 report by three study participants, the authors concluded that the USGS process of evaluating, manipulating, and publishing data is “an extraordinarily complicated process requiring too many steps and too many software and hardware tools.” The report’s recommendations for improving the process are now being considered by USGS, but do not expect to see wholesale changes. When asked about the study, Stewart said it did not find anything needing improvement that the USGS Augusta office had not already identified, and that the study “just validated that the staff and processes we use in Maine are the most cost-effective you will find anywhere.”
Stewart’s response reflects the pride that USGS staff take in the way the agency conducts science—thoroughly, deliberately, and with an emphasis on quality not quantity. Getting USGS to do anything more quickly is like urging a revered surgeon to pick up the pace. Sometimes it pays to remember that results are what matter.
“Some people feel like it takes forever for [USGS] to make decisions,” said Evans of ICWP, which helped coordinate the efficiency study. “At the same time, the independent credibility and competence the USGS results bring to any decision process are renowned. So, on one hand, people say ‘Well, we wish we could see preliminary results from their studies. We wish that they would speed up the process.’ At the same time, people like to know that boy, when USGS says ‘Here’s the data,’ or ‘This is what the model shows,’ you can take that to the bank.”
Gauges of Change
This trust in USGS’s work is nothing new; the quality of its data has long been unquestioned, and when the data have been gathered in much the same way for many years—as with streamgaging—the result is a historical record with tremendous value. The decades of data provide insights into everything from the frequency, intensity, and duration of extreme events to the effects on streamflow of man-made changes to the landscape. The records are also providing evidence that contributes to our understanding of one of the most pressing (and hotly debated) environmental issues of our times–climate change.
“Because we’ve always used similar methods in streamgaging with high levels of quality assurance, the legacy is this incredibly long-term, powerful dataset that is virtually unsurpassed in terms of looking at the environmental changes over the past 100 years,” said Bob Lent, director of the USGS Water Science Center in Maine and Greg Stewart’s boss. “In northern New England, we have some of the only long-term unimpacted streamgages east of the Mississippi. So that when we want to understand how climate has affected hydrologic processes, we really can do that in a way that nobody else can.”
Sitting at a conference table at the Water Science Center, Lent spoke briskly with barely a pause between sentences, his train of thought fast and fluid. Lent said he and his staff began looking into climate change’s effects more than ten years ago, and the process proved productive due to streamgage records—and the area’s seasonal water cycle. Climate change research has shown that a key hydrologic effect is earlier snowmelt runoff, and in northern New England, such runoff plays a dominant role; snow is stored in the mountains for many months in winter, then runs off in a rush in the spring, replenishing surface drinking water supplies and providing a vital flow in the life cycles of fish and plants. What Lent and his staff discovered in analyzing streamgage records is that while the annual amount of runoff in the region’s rivers is actually increasing, the traditional timing is off.
“Drinking water supplies probably have more water than they did in the past, but the majority of that water is coming one to two to three weeks earlier than it did in the past,” Lent said. “So they have a longer period of low flow that they have to look at. As the winters warm, we’re seeing fundamental changes in northern New England hydrology.”
Lent rattled off some of the other climate change findings made by the Augusta team: summer baseflows—that is, the groundwater that seeps into streams—have decreased in Maine, while in the rest of New England, baseflows in summer have risen. The annual thawing of winter ice on lakes, known as “ice-out,” has been happening earlier and earlier; the same goes for river-ice breakup. Rare surges in flows that statistically occur as seldom as once every 50 years are now greater than in the past and, if predictions hold true, the pace of change is going to accelerate, with serious implications for designers of bridges, culverts, and other floodexposed infrastructure.
At the root of so many of these findings are the streamgage data, as well as notes taken over the years during visits to gage sites.
“What I think is remarkable is that we’ve been able to use the records that Greg and his crew, and their predecessors, and their predecessors’ predecessors, have made,” Lent said. “We’ve been able to go back and recreate field conditions from 100 years ago—for example, the thickness of the ice they had to drill through. They didn’t need to know how much ice there was, but they needed to remove that amount from their calculations to determine the area of the river they were measuring. If properly used, these records, not only the published records but also the unpublished notes and remembrances, provide an incredibly powerful tool for unlocking some of the connections between climate change and natural resources.”
It is easy to imagine Bob Lent’s and Greg Stewart’s successors, and their successors’ successors, paying a similar compliment in the future to the Augusta team. Strict adherence to USGS’s quality assurance process, in which each hydrologist’s measurements and analysis are scrutinized intensively by colleagues, ensures the value of the team’s records will endure. During the visit to the Piscataquis River gage, Stewart said three of four USGS surface-water experts conducting a recent internal review of Augusta’s streamgaging program called it the best program in the country.
“The pride my staff takes in the data, the quality of the data, to make sure the data are the best they possibly can be, is tremendous,” Stewart said. “I think that’s one of the reasons that our program has been so successful.”
Overcoming the Hurdles
The relevance to climate change research has helped raise streamgaging’s profile—a welcome development for proponents of the network, who are not letting up in their efforts. Since 2007, USGS and ICWP have been holding regional roundtable meetings around the country to discuss the Cooperative Water Program and the work it supports, which in addition to streamgaging includes the newer and smaller USGS groundwater and water-quality networks and a litany of other research projects. The roundtables bring together CWP stakeholders—local, state, and federal officials; USGS representatives; leaders of water organizations—to discuss the program’s strengths and weaknesses, as well as potential changes, such as taking a more regional, multi-state approach to monitoring. NEIWPCC joined with ICWP and USGS to cosponsor the New England roundtable on November 9-10 in Chelmsford, Mass.
“Addressing the challenges related to streamgaging in our region is a whole lot easier when all the interested parties are together in one room,” said NEIWPCC’s Susan Sullivan, a speaker at the roundtable. Also on the list of speakers: Greg Stewart—a good choice considering his passion for the topic. At the Piscataquis River streamgage, Stewart kept finding more things to show and explain, including a brilliantly simple device known as a crest-stage gage, in which cork inside a vertical steel pipe rises as a stream’s stage increases; as the water recedes, the cork adheres to a long piece of wood inside the pipe, leaving physical evidence of a flood crest that can be used to confirm peak flows recorded by the more modern gage house equipment. Just upstream of the site, Stewart walked down to the bank to point out a cableway strung across the river. The USGS uses the cable to send an acoustic Doppler current profiler or ADCP out over the water to measure velocity and depth during periods of high flow.
Talking with Stewart revealed that data collection has its human element too; despite extensive safety training, accidents still happen. Stewart said he and all of his streamgaging staff have fallen through ice at some point while manually collecting data, but nobody has been seriously hurt in the process.
Another occupational hazard—traffic. Gage sites are sometimes located near busy roads with few parking options; even though crews follow detailed USGS traffic control plans, putting out signs and cones warning drivers of work being done ahead, that is not always enough. In Massachusetts, a USGS staffer conducting water research suffered multiple injuries after a car struck him and sent him flying through the air. Even gage houses themselves face risks.
“We’ve had people break into a gage house, rip the battery off, and throw it into the river,” Stewart said. “We’ve had people shoot gage houses. We’ve had mice chew through wires. We had somebody knock one into the river. Anything you can imagine has probably happened.”
But even with all the challenges, including the financial ones, the work goes on and the job gets done, just as it always has. Certainly, there is room for improvement, particularly with the funding. But the network and its long history remain something unique and worthy of celebration.
Actually, a little celebrating was exactly what Maine’s USGS staff had in mind in 2001 when they invited the Secretary of the Interior, the USGS Director, and other officials to the site of the state’s oldest streamgage, the one on the Kennebec River at The Forks. The plan was to commemorate the gage’s centennial, its 100 years of continuous operation. The invitations were accepted, preparations were complete, and everything was set. Then, three days before the celebration was to take place, planes guided by terrorists flew into the World Trade Center and the Pentagon. In the immediate aftermath of 9/11, nobody from Washington was heading into the woods of Maine to praise a streamgage.
The staff in Augusta called off its plans, and believing the moment had passed, never did hold the celebration—an understandable development but a shame all the same. Lost in the process was a rare opportunity for the streamgage network to get a bit more of the attention it so richly deserves.
Photos by S. Hochbrunn, NEIWPCC unless otherwise noted. Reprinted with permission from NEIWPCC's Fall 2010 IWR Interstate Water Report. To learn more about NEIWPCC or view the original PDF newletter containing this article, please visit http://www.neiwpcc.org/.
Gauging the Gages
Feb 8th, 2011 - Stephen Hochbrunn, NEIWPCC
Nationwide Network of Monitoring Stations Keeps Eye on Rivers and Streams, Collects Critical Data—But Funding is Ongoing Challenge