River Science for Non-Engineers

>> Kale Gullett grew up in western South Dakota and attended the University of Wyoming. He received his Bachelor of Science Degree in Zoology, more specifically Fisheries Biology in 1994 and he received his Master of Science in Hydrology in 1996. After a year with the Forest Service in a private consulting firm in Fort Collins Colorado, Kale headed west and spent four and a half years working with the Yakama Nation in Washington State and four and a half years as a senior habitat biologist with the National Oceanic and Atmospheric Administration based in Washington and working on regional issues across Washington, Oregon, and Idaho. Kale's interests and specialties lie in watershed hydrology, fluvial geomorphology, aquatic ecology, fish passage, instream flow science, irrigated agriculture and reservoir or operations in management. Kale has been with NRCS at our Tech Center here for going on 7 years and he is a very experienced webinar presenter. So, with that, Kale I'm going to turn this presentation over to you. >> All right. Well, thank you very much Holli for that introduction and thanks to everybody out there for taking time out of your day to sit and watch the computer screen. I really appreciate it. I'll get right going because I've got quite a bit of slides today. So, I've got to get into it so we have a chance of getting done. Working in and around river is often accomplished by using tools and concepts of hydrology and fluvial geomorphology which are two related but specialized disciplines. Today, I'll cover some of the fundamental aspects of these two sciences and review some of the influence at land use and management has on stream flow in river morphology, trying to use some examples from different parts of the US. Further, I'll suggest some resources that might help you with analysis as well as the key short primers on river science. As I mentioned earlier, today's webinar is loaded to the brim so I won't have a lot of time for details on some things that really actually have a lot of details. So, please get a hold on me anytime to talk about this stuff or to get additional information. [ Pause ] All right. If you're new to work in and around rivers, especially on projects with outside partners or TSPs, you might encounter a bunch of passwords, models and jargon that sound a bit foreign. And truth is, river work is fairly specialized and not all NRCS folks were trained as hydrologists or geomorphologists. Many of these river science concepts have found their way into the mainstream only the last 20 years or so. I didn't intend to alienate anybody by today's webinar title. In fact, most engineers receive much of what I'll talk about today during completion of a bachelor's degree and many NRCS engineers has these tools and methods of hydrology and fluvial geomorphology, and the design-- excuse me, design and construction of various conservation practices. Today's webinar is aimed at conservation planners to interact with land owners and partners in the field and it's intended to provide a broad overview of the elements of river science that define rivers and may suggest why they look or act at certain way. [ Pause ] So, different professions have different views and perspectives about river corridors. Land planners often consider many aspects of land use and mainly include river corridor as when their activities penetrate corridor boundaries. For example, wastewater treatment activities can take place in confined structures far removed from a river. But the effects of releasing treated water into rivers is a primary consideration of water treatment planning. Engineers are often concerned with developing and implementing safe and practical solutions to human problems. For example, cities and counties often spend hundreds of hours and millions of dollars designing, building, and maintaining Civil Works projects that protect infrastructure from flood damage. To complete these activities, engineers sometimes partition river corridors into interrelated sets of theoretical and quantitative equations to describe and predict the effects of different amounts of water in a stream channel. Biologists are usually charged with understanding how physical, chemical, and biological processes affect riverine species and habitat. They commonly deal with managing habitats and populations and they're sometimes only focused on these aspects of river corridor management. Geomorphologists on the other hand, view river corridors as a landscape system responsible for moving runoff and watershed materials from headwaters to the ocean. They're interested in how hillslopes, channel networks, and floodplains respond to factors that change runoff and sediment production. Geomorphologists are relatively new science, born of the marriage of physical geography and geology. So, although each of the different professions I just discussed usually brings different views and responsibilities to a river project, I think they all share similar goals. Things like reducing erosion and conserving soil, improving water quality and supply, enhancing habitat, preserving streambank function, and creating functional river corridors are common goals that all relate to river conservation. Another commonality of this profession is terminology, although sometimes definitions differ a bit. If you had create a tag cloud of abstract keywords from popular published works in hydrology and fluvial geomorphology over the last 20 years in the United States, you get something that looks like this. Now, some of these terms are familiar but have an entirely different meaning when used in the context of river science, especially when you combine two or three of them. Others are specific to hydrology and fluvial geomorphology. So, today, I'll use many of these words and hope to explain them in a way that clears and doesn't muddy the water. I'll need a few definitions today and here's a good resource for most of the words in the previous slide. It was developed by a well-respected hydrologist with the US Geological Survey and it's available online. The references and resources I'll show today will be listed in a hyperlinked file included in the download package for the webinar. If any of the links are broken, just get a hold to me and I'd be happy to send you a digital or printed copy. So, let's start with the definition for hydrology. Hydrology is an earth science that deals with the flow of water across and through near surface environments. Key elements related to today's talk are how we measure and characterize streamflow and how changes or fluxes in streamflow affect river channels and floodplains. So, many of you remember something similar to this graphic from a junior high science class. The water cycle is composed of many elements driven by precipitation and other factors and includes both the surface and subsurface pathways. However, for today's purposes, we're going to focus on the part of the hydrologic cycle that concerns surface runoff, streamflow, river channels, and floodplains. Further, we'll focus mainly on drainage basins as a smaller unit of the planet where one or more rivers affects landscape appearance and function. Basically, a drainage basin is a landform where watershed materials, mainly water, sediment, and wood, move downhill under the influence of gravity. Within a drainage basin, rivers and streams act as systems or machines that manage available resources to accomplish a task. In this context, the primary task of a river is landscape change, collecting materials and moving them downstream. Gravity is the basic energy source. Water does most of the work, and rocks, soil, and streamside vegetation-- both living and dead-- provides resistance to that landscape change. So we need a basic stream anatomy overview before going much further. The channel is the part of the stream where water appears and it's usually rocketed by banks. The deepest part of the stream channel is known as the thalweg and it's a German word meaning valley way. The flat surfaces above and along the stream banks are called the floodplain and that's where water goes when the channel can no longer contain it. Sediment in the broadest sense is the sand, gravels, and boulders that appear in the channel and along the river corridor. And streamside vegetation often relies on the water table associated with the river and once it ends up in the channel, it becomes large *** debris or LWD. LWD can play a huge role as the structural element of the river channel in floodplain. [ Pause ] One of the fundamental pieces of information related to stream hydrology is streamflow, also known as discharge. Measuring the amount of water flowing down a stream channel can be done with different types of equipment, some simple and some relatively spendy like the velocity meter this gentleman is shown holding. When streamflow is measured in a channel, it gives you a snapshot in time of how much water was there the day you visited a site. Streamflow is determined by a measuring the dimensions of the channel, its width and depth, and then by using meters per second or other means to determine the velocity of the water flowing in the channel. Once these variables are known, the streamflow can be determined by multiplying the area of the channel by the velocity of the water flowing down it. Since velocity isn't equally distributed across a stream channel, accurate measurements of streamflow are derived by subdividing the channel into equal increments, taking measurements within those increments and then adding up the individual measurements within each of the smaller chunks. Streamflow is a volume rate function, it's a volume of water moving past a point per unit time. The basic units streamflow used out there are cubic feet or cubic meters per second. To put a CFS in the perspective, its equal to about 450 gallons per minute. One cubic foot per second flowing on to an acre-sized parcel for 24 hours would cover it to a depth of 2 feet. Since a streamflow measurement only gives you a snapshot in time, we set up gauging stations to record flow over time at multiple locations. I'm sure a few of you have seen things like this adjacent to a bridge over a river. Most streamflow gauges don't record the flow velocity-- it varies quite a bit across the channel with water depths and other factors. However, the stage or level of the water in the channel can be measured with some accuracy with minimal instrumentation. Stream gauges measure the stage of the water in the channel and a basic set up just shown here in the upper left. The stage or elevation of the water surface in the channel can be converted to streamflow by using a stage discharge relationship shown at the lower left. Flow data collected by folks using waders and meters as shown a couple of slides ago can be matched with river stage to develop a rating curve. As you can see on the graph, a stage of 3.3 feet equals a discharge of 40 cubic-- excuse me, 40 cubic feet per second. Now, this method has been used for over 100 years at many gauge stations across the country. However, gauges have to be visited regularly to maintain the accuracy of the rating curve. Stream gauges can collect data at a number of different time steps. Some records stage data every 15 minutes for example and over the long term, this data allows us to describe how much water was in the stream channel everyday of the year for many years. When plotted in graphical form, the data are known as a hydrograph, which shows streamflow over time. This hydrograph shows daily average flow in the Alsea River near Tidewater, Oregon from October 1989 through November 1999. Time is shown along the bottom or X axis and discharge is shown on the vertical or Y axis. Every dot represents the average flow for one day, and as you can see, flow in this river range from very small to almost 30,000 CFS over the 10 years represented. The span of time represented by gauge data is known as the period of record. So, here, the same data from the previous hydrograph although this time presented as a flow duration curve. Graphs like this allow us to estimate the frequency of flows at a station, and are often the result of the cumulative frequency analysis of all available gauge data for a site. On this curve, the highest flows are near the upper left and the lowest flows are at the lower right. Where a hydrograph shows the chronology of occurrence, flow duration curves are developed from all data, the day a certain flow occurred is not relevant. Reading a flow duration curve is easy. For example, the flow at the station that was equaled or exceeded 50 percent of the time was about 650 CFS. You get this by finding 50 on the X axis following the grid line up to the blue curve and then moving left over to the vertical axis to read the discharge amount. The Y axis is logarithmic, so each major grouping represents an order of magnitude increase in flow. Whether you've heard of a flow duration curve or not, I would bet that you've heard of one of the ways flow frequency analysis are commonly used. Flood frequency analysis is used to determine things we hear about in the news like a 100-year flood, a new term here is recurrence interval or the number of years it takes for a flow of some value to return. The recurrence interval of a 100-year flood is 100 years and it has a 1 percent chance of occurring in any given year. Recurrence interval is found by dividing 1 by the exceedance probability. While duration analysis uses average daily flows, flood frequency analysis use annual peak flows from a gauge. So if you run through the statistics, you can generate a curve that looks like this. You would read the value of the 100-year flood much the same way as we determine the 50 percent exceedance flow in the previous slide. So the 100-year flood based on this record is about 60,000 CFS shown by the blue lines connecting the X and Y axis that intersect at the flood curve. [ Pause ] It's important to pay attention to the way data are represented to get a better idea about the accuracy of probability and prediction. The two charts on the left were used-- or excuse me, were developed using the exact same 20 years of data. The panel on the upper left uses log axes which helps the data fit a straight line that appears the shoots straight as an arrow to some high value at the upper right. The panel on the lower left uses linear axes which makes the data look like different, curvilinear in fact. As you can see, bending the predictive line to fit the same high value at the right takes some imagination and could very well end up in a different place and value depending on how you fit the line. So there are errors and inherent ghosts when using data like this, especially when trying to predict a 100-year event based on only 20 years' worth of data. Also, bear in mind that this is probability theory and statistics. Recall that 100-year flood has a 1 percent chance of occurring in any given year. However, the chance that a homeowner will experience a 100 year flood at least once over the term of a 30-year mortgage is 26 percent. All right, so as if you've ever need to find streamflow data or interested in diving into some of the stuff I've just been talking about, I'll go over a few internet resources available out there. The first is known as the National Water Information System or NWIS, and it's maintained by the US Geological Survey. Information from thousands of instruments including streamflow, groundwater, spring flow, atmospheric data, and glaciers are all available at this portal. This graphic shows the front end for active surface water sites across the United States. Each gauging station is a gray triangle. Zooming in on a site of interest will show you the gauges in that area. Clicking on the access data link will take you to available data for that site. So here's what the site data interface looks like when you click on that access data link. A whole bunch of information can be accessed here including streamflow, weather, water quality, and river stage. Some stations like this one for the North Platte River near Northgate, Colorado also include real-time streamflow data. Real-time data are useful for a number of purposes. For example, these data, accessed last week for the same North Platte River gauge, tells me whether or not I could get my boat down the river without dragging which the answer to that is probably not since daily flows are way below average based on the 98-year period of record for the gauge. The average daily flows are these reddish triangles. And here's the actual flow appear for example on July 12. Also, if I drove in to the nearest town which is Walden and walked in to the Elkhorn Cafe where I'd be related to some of the patrons, I'd know who had irrigation water and who didn't because of the low water year in the basin. And this is important because it helps you know where to sit down because some folks are going to be grumpy. So, the National Water Information System makes streamflow and other hydrologic data available to anyone with an internet connection. However, it generally doesn't present data analyzed using statistical hydrology methods in the form of flow durations or flood probabilities. Another application is known as StreamStats and it uses a powerful set of tools to deliver statistical streamflow products. This graphic shows the front end of the tool as well as the states which are shown in green where the tool is fully functional. Most states are functional or undergoing implementation, and the remaining 12 are the states in white, slated for completion in the next few years depending on funding. Within the operational states-- and I went to Colorado again-- the interface works much the same as the previous NWIS tool. You use a map to zoom in on the area of interest until you find the gauge. Clicking on the gauge gives you access to links that take you to data, which are usually shown in blue and underlined. The data, however, look a bit different than the NWIS interface, and as you can see, the screenshot shows or I hope it shows that statistical predictors like the estimates or estimates for the 2, 5, 10, 25, and 50 year peak flows are given for this location. Another cool feature of StreamStats is that it-- it enables the user to select a point in an ungauged stream or on an ungauged stream to generate drainage basin and streamflow information. You zoom in to an area of interest to the layer where blue line coverages of stream networks from the National Hydrography Database kicks in. And select the point anywhere along the stream and StreamStats will delineate the basin and then use a number of different types of data and tools to generate streamflow estimates. And here is the shot of the type of data generated from the small watershed I delineated in the previous slide. Now, the software produces a lot of information including drainage basin metrics, peak and low flow estimates and flow duration data. However, these data specially the streamflow are largely estimates derived using equations, and although they can give you an idea of streamflow and other facets of the hydrology of a basin, they can only get you so far. For example, they shouldn't be used for any sort of design work without calibration at a project site. So, how does StreamStats generate all that data? Well, a part of the answer is related to the same factors that affect streamflow from a drainage basin. Many things affect the way a basin cycles precipitation and how that appears as streamflow. In addition to the type and amount of precipitation basin geology, soils, size, slope aspect and shape or-- and land cover affect the timing, duration, magnitude, and intensity of streamflow at the basin outlet. Precipitation type has a strong influence on the shape of the hydrograph from rivers in different areas. For example, snowmelt dominated rivers shown by the red line on this graph usually have their highest flows in the spring when snow melt-- snow melts and runs off with low flows over the rest of the year when ground water takes over. Rivers in the rocky mountain states have hydrographs like the red line. Basins dominated by rain shown here as the blue line exhibit high flows for 3 to 5 months throughout the winter and early spring. The Southeastern United States has a lot of rivers with hydrographs that look like the blue line. Basins with both rain and snow precipitation shown by the black line usually have a bimodal or two *** hydrograph which reflects two distinct periods of rainfall and snowmelt runoff. Examples of these types of rivers can be found along the West and Pacific Northwest coast. Basin scale, slope and shape influence stream low and hydrograph shape as well. For example, as shown here on the left, smaller basins produce less streamflow with quicker rise and fall than a larger basin. And as shown on the right, basin slope and shape affect discharge as well. Narrow steep basins tend to produced sharply peaked hydrographs with abrupt floods, while more symmetrical basins have broader peak flows. In addition, land cover influences the shape of a hydrograph. Now, this figure shows the effect of the same amount of rain in two similar watersheds, one forested and the other urban. Generally, streamflow for a river draining an urban watershed is flashier, meaning, it quickly spikes and recedes to very low flows especially when compared to a forested landscape. The forested setting processes rainfall differently primarily due to greater vegetative cover, lack of impervious or hard surfaces like roads and parking lots, and intact shallow aquifers that act like hydrologic capacitors releasing flows slowly back to the channel after a rain event. So here is an example of how land cover change affects streamflow using the lower Skagit River in Washington State. This area lies along Puget Sound, north of the city of Seattle. The panel on the left shows a pre-development state of this part of the lower Skagit. And although you probably can't read the legend along the bottom, you may notice that the area around the river is composed of mostly blues, greens, yellow or orange. Now, this means that the river was mostly surrounded by floodplains and wetlands prior to development. The panel on the right shows the present day land cover in the same area of the basin. You'll notice mostly white and magenta, which means that land cover has shifted to mostly agriculture and urban uses. Only about 10 percent of the predevelopment floodplain and wetland landscapes exist. And here is the effect that changing land cover has had on streamflow in the Skagit River and three other Puget Sound systems, two affected by urbanization and one not so much. The red markers indicate flood probabilities for 1883 while the black markers show flood probabilities for 2002. Peak flows in the Skagit and Mercer Rivers have increased by more than 100 percent. By comparison, the Skokomish and Cedar rivers have experienced much smaller changes, owing to less urbanization and greater retention of forested cover and their upper watersheds. So since land cover influences the frequency and magnitude of streamflow, it follows that there must be some corresponding effect on the way a stream channel looks and functions in response to hydrologic alteration. Now, this is a decent segue into a discussion of fluvial geomorphology because over time, river systems tend to reach a balance between water and the material that it carries-- mainly sediment and wood. All right, time for another definition. Fluvial geomorphology is the study of landform changes driven by flowing water. [ Pause ] And many talks covering fluvial geomorphology include this graphic in some form, so we'll use it right out of the chute and refer back to it from time to time. Known as Lane's balance, this figure suggests that channel response and stability are dependent on a proportionality or balance between the load and size of sediment and the slope and streamflow of a system. Originally used by the Bureau of Reclamation in the design of stable channels, this simple tool illustrates how altering any one of the four components forces channel adjustment. For example, if you move the bucket of water here on the right side of the balance to the right, the left side would rise up and the needle in the middle would move towards degradation. This response is commonly observed, increasing slope or adding water to a system usually results in degradation or incision, with the channel eroding downward, becoming narrower and deeper. You could also include a vegetative component into this relationship because streamside vegetation has a strong influence on geomorphology. A few large scale physical side boards dictate the way a river basins look and function. Climate, geology and the physical geography of a watershed govern the amount of water and sediment a river will process on any given day of the year. River channel and floodplain geometry adjust themselves to the prevailing streamflow and sediment regime and vegetation plays a big part in resisting adjustments. Again, this is the general relationship that Lane's balance captures. At the drainage basin scale, we tend to lump river systems into three parts. Gravity powers the work done and a number of forces resist erosion along the way as rivers try to get rid of extra energy and find that balance. So moving from left to right, we have zone 1, the headwaters or uppermost reaches of the stream which serves as supply zone. Steep slopes, bare, sometimes young land forms and extreme climate and precipitation ranges supply sediment to river channels. Zone 2, also known as the transfer zone is where rivers begin to respond to the materials delivered from upstream. Floodplains become more extensive, channel morphology can be diverse and streamside vegetation plays a bigger role in floodplain and channel structure and function. Zone 3, the deposition zone is where we see the largest rivers in a given landscape, extensive floodplains and wide deep channels. So when you assume that river systems reorganize from headwaters on down, we can make some generalizations about the basic architecture of any river from headwaters to the ocean. As you move downstream, slope and the average size of bed material decreases. Streamflow, channel width, depth and average velocities increase. In addition, the average amount of sediment stored in the channel and floodplain increases as does the relative influence of vegetation. As I mentioned in the previous slide, river channels and floodplains adjust to fluxes in water and sediment over time. These are known as alluvial systems, and are often referred to as self-formed because of their plasticity to changes in streamflow or sediment regimes-- think of Lane's balance. Non-alluvial channels on the other hand are those not formed in alluvium and tend to remain functionally stable over long time periods-- any river in a bedrock canyon is a good example. I've talked about floodplains and you may have heard of river terraces. They're related to one another but are significantly different. For alluvial rivers, the floodplain is the surface adjacent to the river channel built and maintained under the present streamflow and sediment transport regime. It's a geomorphic surface and I'd like to point out that the 100-year floodplain you hear about relative to FEMA and flood insurance may not always be related to the geomorphic floodplain. Sometimes it's an administrative boundary that changes after a flood. Terraces are abandoned floodplains that are usually associated with a different river than you see today. So, here's the shadow of the Snake River at the foot of the Tetons, showing the floodplain which brackets the channel on either side. Above there you'll find terraces which in this setting were formed by successive periods of glacial advance down the valley and retreat back up. Again, these terraces are abandoned floodplains of a much difference Snake River. So, let's get into some of the specifics of channel form, function and process. Specifically, we'll cover the following categories of metrics used to describe and analyze stream corridors. Profile describes the form and shape of a river from headwaters down. Pattern or planform is how river looks from above. And dimension includes elements of a river's size and shape that vary with streamflow, sediment, and the influence of vegetation. Channel profile describes the overall longitudinal shape-- from headwaters to mouth-- of a stream system. It's fairly intuitive concept, since water flows downhill. Base level is a concept that describes the factor that limits how far-- in the vertical or elevation direction-- a channel can erode or aggrade. For all rivers, the ultimate base level is global sea level. However, local controls on base level are afforded by dams, large landslides and geologic factors such as waterfalls. Now, rivers are generally lumped in one of four channel pattern categories that are commonly recognized according to appearance as well as physical variable such as slope, number of channels, vegetative characteristics and the size of material comprising the bed and banks of the river. These four categories are straight, braided, anabranched and meandering. Bear in mind that each of these can exist within the same river along its entire course. Channel pattern can be referenced at a variety of spatial scales. Straight rivers or even straight reaches of rivers are relatively uncommon in nature. Straight rivers are relatively unstable and while they may have a slight bend or two like a meandering river, the location and occurrence of these bends appears to be largely random. Naturally, straight rivers are associated with bedrock, faulting or other tectonic features. The most common examples of straight rivers reaches in a present day are commonly associated with human modification. Braided rivers are common at high latitudes and altitudes, and likely reflect with all rivers on the planet look like before the advent of vascular vegetation about 400 million years ago. Today, many braided systems are associated with glaciers or in low gradient streams dominated by sands and small gravels. Braided rivers have many channels separated by high spots or areas of very shallow flow. Braided river systems exhibit complex flow patterns that can change over short periods of time. Consequently, channel migration rates can be extreme and the amount of material moved can be high but irregularly distributed across the channel. Braided rivers are commonly devoid of vegetation. Braided channels are often seen after debris flow events for some period of time until the river is able to rework itself into the pattern that existed before an event. Another channel pattern I'll discuss today is known as anabranched. At first glance it resembles a braided river-- many channels with islands in between. The main functional difference between braided and anabranched rivers is the presence of vegetated islands. Although, some geomorphologists don't recognize this as a major group, I include it here for two reasons. First, there's strong evidence that anabranching is an intermediate step between a braided and meandering pattern. And second, anabranched rivers highlight the important role of vegetation in keeping rivers together. In fact, there's an interesting bit of work out there that details the change in river morphology from meandering to braided after the Permian-Triassic extinction event that killed the dinosaurs. Terrestrial plants died as well and rivers changed from the meandering to braided in a relatively short period of time. The final pattern we'll discuss today is meandering. Meandering rivers wander from side to side along a floodplain in a pattern resembling a sine wave. Meandering doesn't have to be contained within the banks of a river-- here, it's a purely hydrodynamic phenomenon. This slide shows meandering patterns observed in the Gulf Stream off the east coast of the US in the fall and winner of 1996. Hopefully, you can see the meander patterns in the main current as well as rings that have been spun off as the meanders migrated across the ocean. And here's the large meander loop formed by melt water running off the surface of a glacier. The meander pattern is nature's way of minimum uniformly distributed work. [ Pause ] So, before talking more about meandering channel types, I need to talk a little about the relationship between streamflow and channel form. Major floods have the capacity to do a tremendous amount of work, but they occur infrequently and thus aren't able to totally account for the morphology of rivers and floodplains overtime. On the flip side of the coin, low flows are the most frequent but since they lack the competence to move large amounts of sediment, they don't have much influence on channel and floodplain morphology. So, that leaves something in the middle, medium flows that occur on a relatively frequent basis lasting a few days or weeks, every year or two. So, the medium flow that's received the most attention and research in years past is known as bankfull flow-- here defined as the discharge where water just begins to leave a stream channel and spread out onto the floodplain. Commonly, this discharge occurs every one to three years on average in stable alluvial, temperate streams-- however, bear in mind that there's a good bit of variability with return interval of bankfull flow depending on where you're at. Graphically, bankfull stage, or the level of water in the channel that coincides with the bankfull discharge, appears as the dotted line above the light blue band in the graphic shown. There's a bunch of research out there done over the last 50 years that details correlations between drainage basin area and bankfull flow. In addition, there are numerous instances where channel dimensions like width and depth and thus, channel area are a function of bankfull flow. These correlations allow folks to develop predictive relationships of channel dimension as a function of drainage area known as Regional Curves. So, bankfull flow is one important driving factor behind the shape and appearance of a stream channel. Over the long run, bankfull flow does the most geomorphic work by transporting bedload, maintaining channel form, and driving channel migration. Bedload transport begins at flows equal to about 60 percent of bankfull in a gravel channel and most particles would be in motion at about 90 percent of bankfull flow. So, if you see a channel flowing at bankfull, you can assume that many of the particles inside that channel are moving somewhere or another. And here's a shot of Youngs Branch near Groveton, Virginia within a few inches of bankfull stage. So, identifying bankfull in the field can be difficult and is often the source of a fairly significant range of error. Now, some systems lend themselves to easy identification, others have been modified to the point where finding bankfull could be difficult. These are a few of the more commonly used indicators of bankfull flow. Here's a cross section from a small stream in Surry County, North Carolina. The red lines indicate the channel cross-section and an estimate of where bankfull stage would be found. It's coincident with the top of the bar surface along the right hand side of the picture. But again, there are quite a few streams out there where identifying bankfull stage can be difficult or entirely inappropriate given watershed condition and channel type. Okay, so back to meandering. Meandering rivers are the most common pattern in the world, and as I showed a few slides ago, the pattern exists in nature outside of river settings. Now, meander patterns can be predictable-- one complete sequence occurs about every 10 to 14 bankfull channel width. That's kind of why I made that little side trip on discussion of bankfull. Within a meander sequence, the primary structural channel units are pools and ripples and they're spaced about every 5 to 7 bankfull channel widths apart. And finally, meanders can persist over long time spans. Well-developed meander patterns carved in the bedrock can be found all the over the world. And here's an example of a meander pattern that's remained stable over a long span of time. The meanders of the Colorado River had kept pace with the uplift of the Colorado plateau-an ongoing process that started millions of years ago. As I mentioned, the length of a full meander sequence or the wavelength is related to bankfull channel width. This red line shows one meander wavelength and the distance from end to end is equal to about 12 bankfull channel widths in this river in Wyoming. As you can see, the winding channel is longer than the straight line distance down the valley. This concept, which can be measured, describes the sinuosity or bendiness of the river. Higher sinuosity means the river is more bendy. Within in meander wavelength, the primary structural elements are pools which are low spots in the stream and riffles which are high spots. As I mentioned earlier, these channel units are often spaced from 5 to 7 bankfull channel widths apart. Some interesting hydraulics operate in meanders that both maintain channel morphology and enable the channel to move across and down the floodplain over time. Flow along the outside margin of a meander bend which is shown here in red excavates the outer bank and moves material downstream. When water flows into the bend, it generates a cork-screw like motion known as helical flow that rotates from the channel bed upward toward the stream bank. This gives the river a lot of scour or erosion ability to move the bank overtime. Material eroded from the outside of a meander bend is dumped in a process known as deposition along the inside edge of downstream meanders. These depositional features are known as point bars. The scour and fill action of meanders creates pools and riffles as bends are scoured and bars are deposited. The migration of meanders by scour and fill drives floodplain evolution and exposed bars are colonization sites for riparian vegetation. Here's an example of channel migration and floodplain evolution, moving from right to left, along the Bow River in Alberta, Canada. The tip of the point bar along the left margin of the photo is relatively barren, but as you move up onto the floodplain, you begin to see at least four different age classes of cottonwood trees. Aging these trees would enable you to determine the relative age of the point bar as well as channel migration rates. So overtime, meanders move across the floodplain leaving scars where former main channel positions have been closed off. Floodplain features like meander scars and oxbow lakes help you identify the lateral extent of the active floodplain. Okay, now, a few slides on the basics of sediment transport or how sediment moves in river channels. As I mentioned earlier, sediment generally includes a range of materials from very small to very large, like a boulder. Sediment transport mechanisms are commonly broken into three categories related to the size of the material and transport: Wash, suspended and bed load. Wash load, also known as dissolved load isn't a big player in the morphology of channels and floodplains. However, suspended load and bedload are basic building blocks in stream channels and floodplains. Suspended load is mostly composed of silts and clays transported aloft in the water column by turbulence, hence the main. It's the largest component of the total material transported by a river or stream and it's said to be supply limited-- the amount transported is limited only by the supply on hand. It plays a big part in channel and floodplain maintenance and can be a significant water quality problem if a stream is out of balance or land use introduces a bunch of fine sediment to a system. Bedload, which encompasses everything from sand up to large boulders, has a strong influence on morphology of a river. Bedload moves in a layer immediately above the bed of a river by sliding, rolling, or bouncing along. Measuring and predicting bedload is notoriously hard and error prone. One of the problems with bedload is that it moves in irregular waves or bursts that don't lend themselves easily to prediction or measurement. Bedload is transport limited, meaning that it can only be moved when flows get high enough. So, combining all of these concepts allows us to evaluate or describe what happens when changes are made to the water, sediment and large wood regime of a river. Most rivers across working lands have been modified in one way or another, and responses to these modifications is a whole 'nother talk. However, in the last bit of our time today, I'd like to cover two of these cases, channelization and the effects of dams. Rivers are often realigned and placed into engineered channels in an attempt to protect adjacent land uses from flooding. Here, the Walla Walla River was placed between levees in 1963, to protect orchards and other high value crops. However, a big flood in 1964 breached the levees and the river assumed a fairly obvious meander pattern as it migrated beyond its constructed margins. Other changes can be a less dramatic but end up looking the same over time. Many stream channels were realigned and straightened in an effort to control drainage and maximize farmable acreage. Here's an example of a reach of stream channelized sometime in the '30s. The red X serves as a registration point to anchor your eye through the next few slides. Some 40 years later, in the absence of continued manipulation, the channel has changed quite a bit. You can see that it's assumed a meandering pattern, the light unvegetated bands adjacent to the meander bends or point bars, and we're starting to see some streamside vegetation. Jumping ahead another 20 years, the channel has expanded its meander pattern, point bars are more extensive and streamside vegetation continues to increase. And by 2002, we see what appears to be a vegetated buffer. The menders are more strongly pronounced and streamside trees are more prevalent. There even appears to be a channel shift near the bottom of the photo on the opposite side of floodplain from our red X. And here's a shot from 2007, this time in color, not a lot of difference between this and the 2002 picture except that the meander loops have extended and moved down valley, and the riparian corridor appears a bit more continuous. So, over the last 80 years or almost 80 years represented by this sequence of aerial photos, the stream has changed quite a bit from its channelized condition in the '30s to the meandering system bracketed by a decent riparian corridor in 2007. The same sequence of events can be found in hundreds, maybe thousands of stream miles across the Eastern and Midwestern United States. I like to use the preceding slides because they illustrate an important tool in fluvial geomorphology. The Channel Evolution Model originally developed by Schumm, Harvey and Watson, first published in 1984. In short, the model describes a sequence of five relatively predictable stages in the recovery of a channel from major disturbances like forest clearing, urbanization, dam construction, or channelization. The end stage-- or point where the channel reaches equilibrium-- results in a mini-me or small version of the original channel inside of a smaller floodplain that's inset into terraces composed of the pre-disturbance floodplain. The model accounts for many of the processes in fluvial geomorphology that I have covered today. Channel pattern and slope changes, the development of bars and structural units like pools and riffles, revegetation and changes in sediment size and composition along the channel as it tends to balance itself are all accounted for by the model. Further, when applied to a channel that hasn't reached equilibrium, the Channel Evolution Model provides a sort of roadmap for future changes that will likely occur overtime once you've figured out what stage of recovery or evolution the channel you're looking at is in. So, you might imagine dams have a strong effect on rivers, both upstream and downstream of the dam itself, and here's a well-studied example from the West Cost. Trinity and Lewiston Dams were built on the Trinity River in California and were finished between 1962 and 1963. Each of these two facilities provides storage for irrigated agriculture and produces hydropower. Downstream of the two dams, the Trinity River at a place known as "gold bar" looked like this in 1961 prior to the closure of the dams. Notice the relative width of the channel and the presence of an island in the middle of the river. [ Pause ] Here's the same location seen in 1970 about seven years after the dams were closed. Now since the dams both decreased streamflow to this reach and captured all the sediment the one supplied this part of the river, you can see that the channel width has decreased and vegetation has started to encroach and narrow the channel. The margins of the island are covered by vegetation, and the channel of the river along the left side of the island-- which is towards the lower right of the photo-- is barely visible. Here's the same location shown in 1997. Streamside vegetation has increased in age and structure, and the channel flowing along the left side of the island has been closed off. The remaining main channel has become narrower and deeper than the pre-dam condition. Now, the river's response to damming shown at this specific location can also be seen for many miles downstream. So, Trinity and Lewiston Dams trapped most of the sediment that once supplied the channel, and their operations and resulting outflows generally scour smaller material from downstream channel reaches. The main channel became simplified and floored with larger sediments that no longer move frequently under the new flow regime of the river. Although operational changes to the dams in the late '90s helped restore some of the geomorphic forces that maintained the channel, there's still a profound lack of diversity and habitat. So, a combined effort of state, federal and tribal entities is working to inject gravel back into the channel in an effort to replenish channel structure and provide better spawning habitat for native fish. Ongoing studies are attempting to quantify the effects of gravel injection on the morphology and function of the Trinity River. All right, my final example for today looks at what can happen upstream of a dam and incorporates the concepts of base level change, urbanization, increased sediment production and the role of riparian vegetation in channel stabilization. Conestee Dam on the Reedy River near Greenville, South Carolina was finished to its present day form in 1892. It was built atop a bedrock control in the river but increased the height or base level of that geological control by about 28 feet. The dam formed Lake Conestee about 6 miles south of downtown Greenville. So, here's what Lake Conestee looked like in 1943. The dam is shown by the red arrow near the lower right of the photograph and a red star indicates the location of a feature known as Taylor's Island. The red and black dashed line indicates the perimeter of the lake at the time the dam was built. As you can see, the channel of the Reedy River indicated by the blue lines is already started to reclaim the former lake. And here's an aerial of the lake from 2009. The location of Taylor's Island, perimeter of the lake and location of the dam are noted using the red star, dashed line and arrow respectively. So, Lake Conestee is almost completely filled with sediment and areas of formerly open water more than 20-feet deep are occupied by a river channel and a vegetated floodplain. And finally a shot from 2011. The dam imparted an increase in the base level of the Reedy River in this reach of about 28 feet. As the Greenville area developed, changes in land cover contributed sediment to the river that couldn't be passed by the dam because of base level rise. The dam created a big flat spot in the river that extended for quite a way upstream, and this slope reduction equated to a reduction in sediment transport capacity that resulted in deposition. If you look at a time series of aerial photos of the area, the site looks much like the delta of a large river emptying into the ocean. As this delta created new geomorphic surfaces, vegetation became established, defended the floodplain from erosion, and a meandering channel developed. As you can see at the lower right of the photo, the new channel now ends only a few feet from the crust of the dam. This is really interesting site, the destruction of a lake and creation of a river channel in less than 100 years. Sometimes, things happen fast along rivers. So, much of this information could be found all over the internet, in numerous textbooks, and thousands of journal articles, but I think there's a few out there, a few short primers that are pretty good reads because they're short and written without too much technical gobbledygook. The first is called "Stream Dynamics: An Overview for Land Managers" and it was written by distinguished forest service hydrologist named Burchard Heede. Next, I'd recommend Chapter 7 in the Big Blue Book, NEH-653. I'd also recommend Appendix A of the "Stream Simulation Culvert Guide" I've spoken of in other webinars. Although the guide is focused on culverts, Appendix A is a great overview of stream geomorphology. And finally, I'd recommend "Going with the Flow: Understanding effects of Land Management on Rivers, Floods and Floodplains." This is a cooperative effort between Oregon State University and Oregon Sea Grant written for land managers. All right, that's it for today. You've set through a relatively epic amount of material and I really appreciate your time and attention. Before I turn this over to Holli, I'd like to mention another webinar slated for the end of January 2013. The working title is "Urban and Channelized Streams, Working in Highly Modified Environments to Enhance Stream Function and Habitat Quality." So with that, we'll turn this back over to Emily or Holli and maybe have a few questions. Be pleased to hear comments or questions. And again, thanks for your time. >> If you'd like to ask a question, you can dial star 1 on your phone to put yourself into the Q. You can also use the notes tab on the top right of your screen. >> Thanks Emily and thanks Kale for a very informative presentation. Your slides are always just really interesting to see and to watch the sequence of events that takes place every time. We've got a series of questions that are coming of true and false type things. Let me start out now. One person is asking about the inter-- oh, hold on, I forgot-- pass the one I need here, yup. Okay, the USGS StreamStats site looks to be a powerful tool. The person is asking for the website or is there another way that you plan to provide that information? >> Well, yeah. As I said, what I've got is a single word document that is going to be part of the download materials for the presentation. But the actual web address is streamstats.usgs.gov. >> Okay that's a simple one. All right, let me move to the next question. The question is the NRCS Hydro, is that a tool that can be used by non-engineers for watershed analysis? [ Pause ] Is that something that you have familiarity with? >> Well, I think I do but, you know, I'm not to the point at least presently where I'd feel comfortable enough to answer the question. I guess that's one of those where you'll have to say, would it be okay if I got back with you to answer that question before I shoot my mouth off and say something wrong? >> Okay, we'll make sure-- I'm sorry. >> I'm sorry, yeah. As long as we can get the question after his name, I'll be more than happy to try to collect my thoughts based on where NRCS hydro is at this time. I haven't been involved with that for a couple of years I suppose. >> Okay. All right, here's a true and false question and you can comment as needed. Meandering appears to be a way for a stream to manage its energy. Is this is a true statement? >> Yes. I think that, you know, if you think about any of the pictures that you've seen of maybe if you're familiar with a lot of the works of Luna Leopold, train wreck in South Carolina that had a bunch of really long rails on it. When it wrecked, all those things were compressed into this really-- I guess, I probably should say about a train wreck, but beautiful sign way of looking forms. And I think, again, the most sort of common way of minimum variance in the context of some famous hydrologist is that, you know, meandering is nature's way of minimum uniformly distributed work. >> Okay, next question. Three-year storms are the channel forming flows, is this true? >> Well, again, you'll hear all over the place that the return interval or the recurrence interval for bank flow is roughly about two years. And I think that there's a pile of data out there that would support that especially in stable alluvial, temperate streams. But as you move into systems that have been modified, again, where sediment and stream flow have been affected, or if you move into settings where, for example the American South West, a channel forming flow may take 90 or a 100 years to happen. That's again based on precipitation and climate. So, you know, we-- I think sometimes we're settled with a lot of this rules of thumb. By and large, I think, you know, for a lot of the places NRCS works in the lower 48 specially in the Midwest, two-year storm could be considered a channel forming flow but, again, we're kind of getting into some details there specially where tankful flow and effective flow have been confused a little bit, so. But as a rule of thumb, I think it's a good place to start and then you can either change, you know, you can either increase or decrease that return interval flow depending on where you're at based some field data. >> Okay, Emily, are there any questions that has come in on the audio? >> No questions on the line. >> I think we are becoming a notes webinar question and answer period. We've got a couple of people, Kale that has passed along their compliments on your presentation. Thank you, Carl and Todd. And also, Carl indicates that please do and it's okay to get back to him on the NRCS Hydro. >> Great. >> Karen is making the-- asking the question will the PowerPoint presentation be available for future reference? And I would say that yes, we always make the webinar replay and a version of the presentation file available for you after the fact. Those are available at our text and our website in the Science and Technology Training Library for NRCS employees. Let's see, Dan is asking a question. Does SVAP 2 reflect what we have discussed today? >> Yeah. You know, actually, quite a few of the slides in the-- probably too many that I've presented are drawn directly from the set that I use for one day SVAP workshop that I've taught in a few places and, you know, it's a great question and thanks by the way. It does not everything it can. SVAP does, you know, they try to capture the various types of forces that effect, you know, stream condition both visually and functionally. So, for example, in 2007-- '06 I guess, when we rewrote SVAP into its current version SVAP 2, the channel condition element is largely composed or comprised of the channel evolution model. And in attempt, you know, really they capture the fact that an awful a lot of the streams we work on, you know, our meandering systems in working lands that have been modified to some extent or another. >> Okay, we've got one more question from Brett. Brett asks, could you please offer your professional view of the difference between form and function driven fluvial geomorphology series? That's a big one. >> Well, there's a big one in that. That might be Brett Fessel from Michigan and I'm not sure if it's the same person. But in any rate, you know, there's-- I guess if I understand the question correctly, I think that there's getting to be a couple of big camps or schools of thought out there in the term or I guess in the world of geo-fluvial whizz *** stuff, form versus function. And I think that when you talk about estimates or classifications according to form or how a river looks, obviously, they're based on data. And I think a good example of this would be the Rosgen's stream classification system. He based his development on that, on data that were collected, you know, starting in the early '80s. The first paper was published in 1994 that's been going [inaudible]. I think that the danger here is where you start to extend form-based theories in fluvial geomorphology into places that don't capture the same types of settings where the original data were collected. For example, if you look at streams and main, a lot of the things that you'll see in main although, you know, the recently glaciated, you know, you'll see an awful a lot of channels that looks like they fit Rosgen's classification system or form-based approach. They don't really-- once you start getting into some of the pieces and parts, there's places where it just doesn't make any sense. And so, then, I believe it's better to reach back into your toolbox and go with something that is, you know, more focused in function, you know, focused on the process in play that might be responsible, you know, at least over the last bit of time that we've, you know, that we're able to sort of evaluate the functions that have occurred out there that make a channel look the way that it does. So, I guess the real-- my real professional view is that they are first and foremost form-based as excellent for helping people immediately get an idea in their head of the stream when you want to talk to somebody on the phone. But sometimes, in terms of trying to figure out design or what factors might be leading to bank stability problems, then you have to be able to shift gears and go onto function or process driven stuff. >> Okay, well you must have hit the nail on the head because you've gotten applause on that answer, Kale. Okay, we got one more in the queue. The question is, the use of *** debris in a stream can be good for stream habitat, do you agree? >> I do agree. I think one of the problems we've run into in the past, excuse me, especially when it comes to things like stream restoration is the idea that, well, wood is good so let's throw it out there. And then of course, what will happen is since maybe the wood is too small for the stream, people will want to pile a bunch of rocks on top of it to hold it in place or go with the great lengths to cable it down. And so, along the lines of this-- so the previous question, it's one of those times when I think a function driven or a process-based approach to geomorphology is good. And it's good because you need to be able to place how wood react in a system, you know, especially in the context of where you're at. Obviously, *** debris in the eastern United States is nowhere near what it used to be. That's an incredible interesting topic if you start to kind of dive into historical literature as far as what the Mississippi looked like or in the New England states what, you know, what the size of an old growth tree was up there. It's just the way. And when you take that information like just the size of wood and what effect that likely had on streams, the watersheds we're working in today are nowhere near what they used to be like, so. But as a general question and answer, yeah, habitat wise, wood is good. >> Well, all right. Well, I'm not seeing any more questions coming in and you've been on the hot seat for about an hour and ten minutes now. So, I think we'll just conclude the webinar at this point. And again, thank you Kale for the excellent presentation. >> Thank you, Holli. Thanks to everybody else. [ Silence ]