Red River Basin Flood Damage Reduction Framework

Red River Basin Flood Damage Reduction Framework Red River Basin Flood Damage Reduction Work Group Technical and Scientific Advisory Committee Technical Paper No. 11 Principal Authors: Charles Anderson, JOR Engineering, Inc. Al Kean, Minnesota Board of Water and Soil Resources May 2004 Executive Summary Flooding is a major problem within much of the Red River Basin. This problem is primarily related to geology, topography, weather, and land use. The Flood Damage Reduction Work Group in Minnesota seeks to provide Project Teams, Watershed Districts and others with science-based and consensus-based tools to enable more effective flood damage reduction within the basin. A fundamental premise of this technical paper is that flood damage reduction (FDR) along the main stem of the Red River and the lower reaches of its major tributaries (glacial lakebed region) is substantially dependent on the types and locations of FDR and related measures implemented upstream. Flooding in the glacial lakebed region of the basin is substantially affected by runoff timing and volume from upstream areas. Runoff timing and volume are, in turn, substantially affected by the topography, soils, precipitation and land use within different regions of the basin, as well as by the types and locations of FDR and natural resource enhancement (NRE) measures that may be implemented. A basin-wide FDR framework will better enable a coordinated approach to integrate various FDR and associated NRE measures that are most effective for achieving the overall goals envisioned by the Red River Basin Mediation Agreement adopted in December 1998. The goal of this framework is to implement various types of FDR measures individually, or in concert, at locations for which they are best suited to achieve FDR benefits locally and in the watershed, while also contributing to reduction of main stem flooding risk. This framework includes FDR measures that are also NRE measures, and promotes multi-purpose projects. This technical paper presents critical concepts about runoff timing and volume in relation to flood peaks on the main stem of the Red River, and facts about variations in topography, soils, precipitation and evaporation within the Minnesota portion of the basin, as foundations for defining the expected peak flow reduction effects of implementing various FDR measures within different areas of the basin. Available geologic, topographic, meteorologic and historical flood data, as well as computed runoff travel times, are used to illustrate these concepts and to define RRB FDR Framework Final.doc 1

early, middle, and late runoff areas within the basin relative to the downstream limit of the Red River Basin in Minnesota at the U.S./Canada border. A wide array of alternative FDR measures are identified, categorized and discussed, including pros, cons, and general recommendations for the best areas in which to implement these measures to optimize overall FDR benefits. A summary table is presented for the identified array of FDR measures with ratings of potential for peak flow reduction on the main stem when these measures are implemented in early, middle, or late runoff areas relative to the main stem. This technical paper stresses the importance of using multiple types of FDR measures in a strategic manner to achieve local, watershed, and main stem flood damage reduction. It presents a framework for creating policies and trends that will help to achieve basin-wide FDR goals, as well as NRE goals outlined in the Red River Basin Mediation Agreement. This technical paper includes a multi-measure example for the Red River Basin, utilizing various types of flood volume reduction and temporary storage measures to reduce local, watershed and main stem flood peaks, and to provide NRE benefits. For this example, it is estimated that the collective effects could reduce the 100-year peak flood flow at the U.S./Canada border by approximately 20%. A spreadsheet method is provided to assess and document the expected peak flow reductions on the Red River main stem at the U.S./Canada border of flood volume reduction and temporary storage measures implemented upstream. This method uses ratios of implemented storage (at a project location) to ideal storage (effect on main stem peak flood volume and flow) for different types of flood volume reduction and temporary storage measures located in early, middle and late areas relative to the main stem. These effectiveness ratios are based on flood routing and other experience of Technical and Scientific Advisory Committee (TSAC) members, including TSAC Technical Paper No. 10, Basin Strategy Hydrologic Analysis, and other previous studies. This method could be used to track progress toward achieving long-term FDR and NRE goals. It is intended that this technical paper be used in conjunction with other TSAC technical papers and the User s Guide to Natural Resource Efforts in the Red River Basin, published in 2001, to give decision makers additional tools to assess and achieve basin-wide FDR and NRE goals. RRB FDR Framework Final.doc 2

Red River Basin Flood Damage Reduction Framework Red River Basin Flood Damage Reduction Work Group Technical and Scientific Advisory Committee Technical Paper No. 11 Table of Contents Page Purpose and Overview... 1 Background... 1 Need for a Basin-Wide FDR Framework... 2 Critical Factors and Types of Information Considered... 3 Topography and Geologic Landforms of the Basin... 4 Climate... 8 Streamflow Characteristics... 12 Defining and Quantifying the Flooding Problem... 14 Hydrographs... 16 Runoff Volume and Timing... 17 Historic Floods... 19 Runoff Travel Time... 22 Definition of Early, Middle, and Late Areas Relative to the Red River Main Stem... 23 Flood Damage Reduction Measures... 24 RRB FDR Framework Final.doc i

Table of Contents (Continued) Page 1) Reduce Flood Volume... 25 a) Construction or Restoration of Depressional Wetlands b) Cropland BMPs c) Conversion of Cropland to Perennial Grassland d) Conversion of Land Use to Forest e) Other Beneficial Uses of Stored Water 2) Increase Conveyance Capacity... 26 a) Channelization b) Agricultural Drainage c) Diversions d) Setting Back Existing Levees e) Increasing Road Crossing Capacity 3) Increase Temporary Flood Storage... 29 a) Impoundments 1) On-Channel Impoundments 2) Off-Channel Impoundments b) Restored or Created Wetlands c) Drainage d) Culvert Sizing e) Setting Back Existing Levees f) Overtopping Levees 4) Protection/Avoidance... 33 a) Urban Levees b) Farmstead Levees c) Agricultural Levees d) Evacuation of the Floodplain e) Floodproofing f) Flood Warning and Emergency Response Planning Summary Table of Flood Damage Reduction Measures and Effects... 35 A Basin FDR Framework... 37 Multi-Measure Example... 38 Temporary Storage and Flood Volume Reduction Components... 39 a) Wetland Restoration / Creation b) Culvert Sizing c) Overtopping Levees d) Impoundments RRB FDR Framework Final.doc ii

Table of Contents (Continued) Page Summary for Temporary Storage and Flood Volume Reduction Measures... 41 Non-Storage Components... 42 a) Runoff Volume Reduction b) Protection / Avoidance Effects of Implementing Example Measures... 43 Flood Storage Assessment and Tracking Method... 44 Summary of FDR Framework... 44 References... 45 Table 1. Expected Peak Flow Reduction Effects on the Red River Main Stem of FDR Measures Applied in Early, Middle, and Late Areas Upstream... 36 Figures 1. Shaded Relief Map of the Red River Basin in Minnesota... 5 2. Major Landforms of the Red River Basin in Minnesota... 6 3. Profiles of the Red River... 7 4. Normal Annual Precipitation... 8 5. Average Annual Runoff... 8 6. Average Annual Temperature... 9 7. Average Annual Lake Evaporation... 9 8. Average Temperature, Precipitation, and Evaporation at Roseau and Wheaton, MN... 10 9. Average Date of Last Snow Cover... 11 RRB FDR Framework Final.doc iii

Table of Contents (Continued) Figures (Continued) Page 10. 100-Year Snow Water Equivalent... 11 11. Red River Mean Daily Discharge at Grand Forks, ND... 12 12. Red River Annual Mean Discharge at Grand Forks, ND... 12 13. Annual Peak Flows at Grand Forks, ND... 13 14. Flood Flows at Cities Along the Red River... 14 15. Flooded Areas Along the Main Stem of the Red River in 1997... 15 16. Approximate 100-Year Flood Hydrograph at Emerson, Manitoba... 16 17. Hydrographs of the 1997 Red River Flood... 17 18. Example of Tributary Contribution to Main Stem Flood... 18 19. Timing of Tributary Inflows (E=Early, M=Middle, L=Late Relative to the Main Stem)... 18 20. Tributary Contributions to the 1997 Flood Hydrograph at Emerson... 19 21. Comparison of a Tributary Contribution to Main Stem Flood Hydrograph in 1966 and 1965... 20 22. Area-Weighted Contribution to Main Stem Flooding... 21 23. Computed Runoff Travel Time to the U.S./Canadian Border... 22 24. Early, Middle, and Late Runoff Timing Zones in the Red River Basin... 23 25. Volume of Ideal Storage and/or Flood Volume Reduction Required to Reduce the 100-Year Peak Flow at Emerson... 37 26. Summary of Example for Temporary Storage and Flood Volume Reduction Measures and Estimated Effectiveness to Reduce 100-Year Flood Peak at Emerson... 41 27. Estimate of Modified 100-Year Flood Hydrograph at Emerson... 43 RRB FDR Framework Final.doc iv

Red River Basin Flood Damage Reduction Framework Red River Basin Flood Damage Reduction Work Group Technical and Scientific Advisory Committee Technical Paper No. 11 Purpose and Overview The purpose of this framework is to better enable a coordinated approach to flood damage reduction (FDR) within the Red River Basin that addresses local, watershed, and main stem FDR goals, so that efforts to reduce flood damages in one area will also consider the potential for negative or positive impacts in other areas. This basin level framework recognizes that FDR goals for the main stem of the Red River, and the lower reaches of its major tributaries, necessarily involve measures strategically implemented farther upstream in the tributary watersheds and subwatersheds, as well as local FDR measures. This technical paper provides evaluations of available geologic, topographic, and hydrologic information to help define what types of FDR measures, applied where, can help achieve both local and downstream FDR goals. It is recognized that a basin-wide FDR framework must consider many local, watershed, and basin-wide needs, opportunities, and constraints, as well as geologic, land use, and climatic variables. This technical paper includes critical concepts and provides general guidance for a wide variety of FDR measures for different areas of the Red River Basin in Minnesota. It also includes a multi-measure example for the Red River Basin upstream of the U.S./Canada border, utilizing various types of flood volume reduction and temporary storage measures to reduce local, watershed and main stem flood peaks, and to provide NRE benefits. It is intended that this technical paper be used in conjunction with other Technical and Scientific Advisory Committee (TSAC) technical papers and the User s Guide to Natural Resource Efforts in the Red River Basin published in 2001 to further develop and implement the FDR and NRE goals of the Red River Basin Mediation Agreement adopted in December 1998. The primary audience is expected to be Project Teams and others involved with watershed planning and development of FDR and natural resource enhancement (NRE) projects within the Red River Basin in Minnesota. Background Many floods have occurred in the Red River Basin in recorded history. The earliest accounts are found in journal entries of trappers and explorers beginning in 1824. Major floods have occurred on the main stem and tributaries both in the spring, due to snowmelt and rain, and in the summer, RRB FDR Framework Final.doc 1

due to more localized heavy rains. Damages due to flooding have, on occasion, been catastrophic. Although flooding has been a natural occurrence in the basin since glacial Lake Agassiz receded, the potential for damage increased with settlement and subsequent industrial, urban, and agricultural development. Flooding occurs when water from upstream enters an area at a rate exceeding the channel capacity to carry that water downstream. Flood damages occur when flood levels rise high enough, or remain long enough, to cause adverse impacts. Flood damages in the Red River Basin have included severe structural damage to private and public facilities and infrastructure, extensive crop loss, major environmental degradation, and loss of life. The Red River of the North is one of only a few major rivers in North America that flow north. This increases its spring flood potential, because snow in the southern headwaters of the basin often melts before snow in the northern areas, causing peak flows from downstream tributaries to coincide with the flood crest on the Red River. The northward flow of the river also results in more ice jam problems than most southward flowing rivers experience. In addition, the Red River is located within the broad, flat bottom of glacial Lake Agassiz, which has only a mild northward slope. As a result, the main stem and tributary rivers in the glacial lake plain area of the basin frequently overflow onto broad floodplains. Need for a Basin-Wide FDR Framework There are many alternative measures that can be implemented to reduce flood damages. These include structural measures such as levees, channel modifications, and various types of floodwater impoundments, as well as nonstructural measures such as limiting floodplain development, changing floodplain use, and changing upstream land use to reduce runoff volumes and rates. A flood damage reduction measure is typically selected by those who benefit from, and pay for, its implementation. The choice may also be influenced by public policy, which can control some measures through regulation, or may support others by offering technical assistance or financial incentives. Affected people and policymakers base decisions on their understanding of the flood situation and priorities. Typically, there is good understanding of local flooding problems, because firsthand knowledge and experience are substantial at the local level. What may not be well understood, or adequately considered at the local and watershed levels, are the effects of a particular FDR measure at the watershed or Red River Basin level. A basin-wide FDR framework is needed to ensure a coordinated approach to achieve outcomes that are effective at the local, watershed, and basin levels. This framework should encourage the implementation of measures that reduce local, watershed, and main stem flood damage potential. It should discourage the implementation of measures that, while achieving local flood damage reduction, would increase flood damage potential at the watershed or basin level. RRB FDR Framework Final.doc 2

Individual watershed district overall plans address coordination of FDR measures to achieve local and watershed FDR goals. This basin-wide FDR framework provides additional guidance to improve consideration of watershed and main stem needs when selecting FDR measures at the local level. A basin-wide coordinated approach may utilize a variety of FDR and related NRE measures that, collectively, comprise a basin-wide FDR framework. This variety of measures may include small, dispersed measures, such as wetland restorations, watershed-wide culvert sizing, increased perennial vegetation and agricultural best management practices, as well as local protection / avoidance, increased conveyance capacity, and strategically located larger impoundments. Implementation of a basin-wide FDR framework requires long-term commitment. This requires policies and trends that achieve basin-wide FDR goals over the long term. Critical Factors and Types of Information Considered Development of a basin-wide FDR framework requires an understanding of critical factors that affect flooding, including runoff timing and volume. When implementing individual projects, it is necessary to know how water from any given area will affect downstream flooding. Will the peak runoff arrive ahead of, coincident with, or after, downstream flood peaks? Will the flood damage reduction measure significantly change the volume of floodwater from a specific area? Some areas contribute more water to downstream flood peaks than others. This generally relates to the area s location within the basin, topography, soils, watershed size, land use, distribution of precipitation, and other physical features that affect the rate and volume of runoff. On the other hand, meteorologic variability makes each flood event unique. Therefore, an area s contribution will vary from flood to flood. The greatest floods on the Red River occur when the spatial and temporal patterns of meteorologic events coincide with the physical features of the basin in a way that causes high volumes of runoff with coincident peak flood flows from many areas of the basin. This technical paper uses the following types of information and analyses to provide guidance for defining where different types of FDR measures work best to achieve basin-level FDR goals. 1) Definition of the effects of watershed topography, soils, location, size, and current land use on the timing and volume of runoff from different areas of the basin. This includes evaluation of runoff travel time to the U.S./Canadian border based on topography and land use. 2) Evaluation of timing and volume of historical flood events for major watersheds and the main stem of the Red River. These flood events reflect all of the physical characteristics of the basin, in combination with the temporal and spatial meteorologic characteristics of these flood events. RRB FDR Framework Final.doc 3

3) Discussion of many different types of FDR measures, including definitions, characteristics, and effects on flooding at, downstream, and upstream from the implementation site. 4) Identification of the effects of different FDR measures implemented in different areas of the basin on main stem flood damage reduction. Topography and Geologic Landforms of the Basin The Red River Basin upstream from Winnipeg has a drainage area of about 45,000 square miles. This includes 17,806 square miles in Minnesota, 20,820 square miles in North Dakota, 573 square miles in South Dakota, and the balance (about 5,800 square miles) in Manitoba. The Red River Basin in Minnesota is shown on the shaded relief map on Figure 1. Also shown are topographic cross-sections of the basin from west to east. To allow for direct comparison, the horizontal and vertical scales are consistent for all cross-sections. RRB FDR Framework Final.doc 4

Figure 1. Shaded Relief Map of the Red River Basin in Minnesota RRB FDR Framework Final.doc 5

Figure 2. Major Landforms of the Red River Basin in Minnesota RRB FDR Framework Final.doc 6

The Red River Basin has four general landform regions characterized by elevation, topography, soils, and stream characteristics, as indicated on Figure 2. The following descriptions of these landforms are from the highest to the lowest in elevation. 1) Glacial Moraine This headwaters region is characterized by rolling hills, lakes, depressional wetlands, and variable soils associated with glacial ground moraine. 2) Lake-Washed Till Plain This region of the basin is characterized by land of gradual slope with large areas of non-depressional wetlands and poorly developed stream networks. Surface soils include large areas of peat lands. 3) Beach Ridge Areas The beach ridge areas of glacial Lake Agassiz are characterized by sandy soils, multiple levels of beach ridges, relatively steep slopes, and incised rivers with relatively narrow floodplains. Wetland areas often exist on the upstream side of beach ridges. 4) Glacial Lake Plain The lowest elevations are within the lake plain of glacial Lake Agassiz. The land within the glacial lake plain region is extremely flat with very low surface and river channel gradients. This flat area originally included large areas of wetlands. The soils are dominated by relatively impervious lacustrine silts and clays. The highest land in the Minnesota portion of the Red River Basin is located in Clearwater County in the Wild Rice River watershed at an elevation of 2,010 feet above sea level. The lowest land, located in Kittson County along the Red River near the Canadian border, is 750 feet above sea level. The most flood-prone areas generally are those with the least slope and those downstream from areas of steep slopes. The gradient of the Red River ranges from a little over 1 foot per mile north of Breckenridge, to about 0.5 foot per mile in the vicinity of Grand Forks, to about 0.2 foot per mile at the Canadian border. Profiles of the Red River from Wahpeton to the Canadian border are shown on Figure 3. Figure 3. Profiles of the Red River Figure 3. Profiles of the Red River (data from U.S. Army Corps of Engineers, et al.) RRB FDR Framework Final.doc 7

Climate The Red River Basin is large, with significant climate differences from west to east and from south to north. Within the Minnesota portion of the basin, the average annual precipitation, as shown on Figure 4, varies from 20 inches in the west to 25 inches in the east. Similarly, the average annual runoff, shown on Figure 5, varies from 1 inch in the west to 5 inches in the east. Figure 4. Normal Annual Precipitation Figure 5. Average Annual Runoff RRB FDR Framework Final.doc 8

Significant temperature differences also exist. Average annual temperatures, shown on Figure 6, range from 44 degrees in the south to 36 degrees in the north. This creates a similar pattern in evaporation potential, depicted by lake evaporation shown on Figure 7, ranging from 31 inches in the south to 23 inches in the northeast. Figure 6. Average Annual Temperature (State Climatology Office, DNR Waters) Figure 7. Average Annual Lake Evaporation RRB FDR Framework Final.doc 9

Seasonal variability is also very important. The charts on Figure 8 show mean monthly temperatures and seasonal precipitation and evaporation patterns at Roseau, near the northern end of the basin, and at Wheaton, near the southern end. The differences in temperature and evaporation from south to north are modest but significant. The monthly evaporation shown is the expected loss from shallow lakes and reservoirs. The difference between evaporation and precipitation develops an annual deficit in shallow water bodies that is restored by runoff from upland areas. The average annual deficit is about 9 inches in the south and about 4½ inches in the north. This may help to explain why, for example, wetlands are considered more useful for flood control in southern areas of the basin than in northern areas. Figure 8. Average Temperature, Precipitation, and Evaporation at Roseau and Wheaton, MN RRB FDR Framework Final.doc 10

A very significant climatic factor that affects spring flooding is the timing of the snowmelt. Figure 9 is a map showing the average date of last snow cover for Minnesota. The south to north snowmelt trend tends to build greater peak flows during spring floods. A characteristic pattern in snow pack volume of available moisture also exists. Figure 10 illustrates the regional snow water equivalent pattern, which increases from southwest to northeast. A more local analysis may indicate a relationship to other variables such as land cover. Figure 9. Average Date of Last Snow Cover (State Climatology Office, DNR Waters) Figure 10. 100-Year Snow Water Equivalent (U.S. Weather Bureau Technical Paper No. 50, 1964) RRB FDR Framework Final.doc 11

Streamflow Characteristics Many tributary streams within the Red River Basin have only intermittent flows. The Red River itself is a perennial stream, but has high seasonal variability and extended periods of very low flow. Figure 11 shows the mean daily discharge of the Red River at Grand Forks. Figure 12 shows the annual mean discharge. Figure 11. Red River Mean Daily Discharge at Grand Forks, ND (USGS gage data) Figure 12. Red River Annual Mean Discharge at Grand Forks, ND (USGS) RRB FDR Framework Final.doc 12

Figure 13 shows a chart of annual peak flows at Grand Forks reported by the U.S. Geological Survey. Figure 13. Annual Peak Flows at Grand Forks, ND RRB FDR Framework Final.doc 13

Defining and Quantifying the Flooding Problem Red River main stem floods are characterized by long durations and widespread flooded areas. Flood damages to roads, farmsteads and urban areas are primarily related to peak stage and velocity of flooding. Agricultural damages are primarily related to time of year, peak flow, and flood duration. Peak flow defines the depth, velocity, and extent of the flooding, while time of year and duration of inundation influence the degree of agricultural damage caused by the flooding. Flood flows associated with a range of flood return periods are shown on Figure 14 as a plot of flow rate (in cubic feet per second or cfs ) vs. drainage area at points along the Red River. Note that peak flood flows increase relatively little between Grand Forks and Emerson. This is primarily due to the effect of floodplain storage. Substantial floodplain storage also limits the increase in peak flood flows between Fargo-Moorhead and Halstad. The extent of the floodplain along the Red River main stem is illustrated by the map on Figure 15, which shows flooded areas along the main stem during the 1997 spring flood. Figure 14. Flood Flows at Cities Along the Red River RRB FDR Framework Final.doc 14

Figure 15. Flooded Areas Along the Main Stem of the Red River in 1997 RRB FDR Framework Final.doc 15

All communities along the Red River are subject to varying levels of flooding risk. At many of these communities, levee systems have been installed that provide varying degrees of flood protection, generally from 50- to 100-year flood levels. The profiles on Figure 3 show levee elevations at a number of communities along the Red River. Emergency flood fight activities are often required to augment the protection provided by levees. Many farmsteads are also protected by levees (ring dikes). Agricultural levees along portions of the Minnesota side of the Red River provide spring flood protection equivalent to about a 5-year event. These levees provide a greater degree of protection against a summer flood event, probably about 25-year protection. The long durations of spring floods commonly cause crop production losses due to delayed planting. Understanding the relationship between flood damage reduction accomplished by these urban, farmstead and agricultural levees, and the additional reduction of flood damage, or risk, resulting from peak flow reduction measures and strategies, is critical to developing an FDR framework. The spring flood of 1997 may be typical of a 100-year flood in the basin. The basin average available moisture in the snow pack and spring rains was about 8 inches, which generated about 4 inches of runoff. The peak flow on the Red River at Emerson, Manitoba, was 129,000 cfs. The total runoff volume at Emerson was about 6,500,000 acre-feet. Hydrographs Flood flows are often depicted as a line chart called a hydrograph. Figure 16 is a hydrograph approximating the 100-year flood flow at Emerson, Manitoba. A discharge hydrograph is simply a plot of flow (typically in cfs) on the vertical axis vs. time (typically in days) on the horizontal axis. The peak of the hydrograph represents the highest flow rate at the applicable location. Adding up the average flows in cfs for each day would yield the total volume of the flood in cfs-days. That volume is graphically represented as the area under the hydrograph. A commonly used unit for flood volume is acre-foot, which is the Figure 16. Approximate 100-Year Flood Hydrograph at Emerson, Manitoba RRB FDR Framework Final.doc 16

volume equivalent of 1 foot of water over 1 acre. Volume in cfs-days can be converted to volume in acre-feet by multiplying by 1.98 (i.e., approximately 2). In this illustration, each grid block is 100,000 cfs-days, which is about 200,000 acre-feet. The shape of a tributary s hydrograph indicates its flood characteristics. Tributaries with high peaks and short durations are often referred to as flashy. Hydrographs showing low peaks and long durations are typical of tributaries having substantial temporary storage. Storage may be provided by lakes, reservoirs, and wetlands, or by extensive floodplain areas. Hydrographs with sustained base flows may reflect significant groundwater contributions. Hydrographs of the Red River 1997 spring flood at Wahpeton, Fargo, Halstad, Grand Forks, and Emerson are shown on Figure 17. Figure 17. Hydrographs of the 1997 Red River Flood (data from USGS, et al.) Runoff Volume and Timing The relationship between watershed runoff and downstream flooding is defined by volume and timing. This is illustrated by the example hydrographs on Figure 18. The large hydrograph represents the flow at a downstream location on the Red River main stem. The small hydrograph within represents the contribution of a subwatershed area to the flow on the main stem. The volume from the subwatershed area that arrives during the peak flood period on the main stem RRB FDR Framework Final.doc 17

directly contributes to the peak flow and amount of flooding. Knowledge of relative runoff volume from tributaries is important for defining and prioritizing runoff reduction and storage strategies at the basin and watershed levels. This knowledge must be used in combination with knowledge of the opportunities and limitations to manage runoff within individual watersheds. Key variables affecting the volume of runoff from a watershed include drainage area, precipitation, soils, topography, and land use. Timing of the subwatershed hydrograph peak relative to the main stem peak is also very important. The peak contributions from some areas, typically in the lower part of the basin, tend to arrive early, ahead of the main stem peak. The peak contributions from other areas, typically in the upper part of the basin, tend to arrive late, after the main stem peak. The peak contributions from the remaining areas, typically in the central part of the basin, tend to arrive in the middle of the flood on the main stem, coincident with the main stem peak. Therefore, tributary subwatershed areas, and their flow contributions to the main stem hydrograph, can be referred to as early, middle, and late. Figure 19 illustrates this concept using example hydrographs. Everything else being equal, the middle areas contribute the most to downstream main stem flood peaks. The relationship between tributary and main stem flooding is also easiest to understand in the middle areas. Activities that decrease the peak flow from these areas will decrease peak flows on the main stem. Conversely, activities that increase the peak flow from these areas will increase peak flows on the main stem. Figure 18. Example of Tributary Contribution to Main Stem Flood (McCombs) The flow contribution from an early subwatershed area to the main stem flood peak comes from the falling limb of that area s hydrograph. Therefore, activities that reduce local peak flows likely will not reduce main stem peak flows. For example, water stored during an area s flood peak and released immediately thereafter might add flow during the main stem peak. Conversely, conveyance improvement projects may move more of an area s water out ahead of the main stem flood and thereby reduce main stem flood peaks. Figure 19. Timing of Tributary Inflows (E=Early, M=Middle, L=Late, Relative to the Main Stem) RRB FDR Framework Final.doc 18

The flow contribution from a late subwatershed area to the main stem flood peak comes from the rising limb of that area s hydrograph. Therefore, activities that result in earlier outflows from that area generally should be avoided. Flood storage projects in late areas will tend to reduce main stem flood peaks, particularly those that store the initial portion of the area s runoff. Historic Floods Much has been learned about the timing and volume of floods by analyzing flow data from historic flood events. The McCombs report (May 1984) analyzed data from 10 flood years including 1948, 1950, 1965, 1966, 1969, 1970, 1974, 1975, 1978, and 1979. The study determined the relative contributions of each gaged watershed area to the flood flows that occurred on the Red River by routing the tributary hydrographs to points along the main stem. A follow-up study by Houston Engineering (January 1999) analyzed the 1997 flood. Figure 20 is an illustration from the study showing a stacked hydrograph of the 1997 flood on the Red River at Emerson. The contribution from each of the tributary areas is illustrated as a layer of flow within the hydrograph. Figure 20. Tributary Contributions to the 1997 Flood Hydrograph at Emerson (Houston Engineering, Inc.) RRB FDR Framework Final.doc 19

Each gaged watershed s contribution to these 10 large main stem floods was also illustrated in a series of hydrographs similar to those of the Clearwater River shown on Figure 21. Those watersheds that deliver the highest volume of runoff during the main stem peak are considered to be the most significant contributors and would likely fall in the middle timing area of the Red River Basin (i.e., in relation to the main stem). Figure 21. Comparison of a Tributary Contribution to Main Stem Flood Hydrograph in 1966 and 1965 (McCombs) RRB FDR Framework Final.doc 20

Another way of looking at the data is the map of relative contribution from subwatershed areas as shown on Figure 22. This map shows the area-weighted average contribution to historic peak flood periods. Figure 22. Area-Weighted Contribution to Main Stem Flooding (McCombs) RRB FDR Framework Final.doc 21

Runoff Travel Time Another way to help identify early, middle, and late tributary areas relative to the Red River main stem and at other locations within the basin is to consider computed runoff travel times based on topography and land use. Figure 23 shows the relative travel time of runoff within the basin to the Canadian border. (The darker the shading, the longer the travel time.) This map was based on runoff computations using a 90-meter digital elevation model (DEM) for the basin, available land use information, and estimated retardance coefficients for various reaches of overland flow, streams, and rivers within the basin. The map only depicts time in transit. Therefore, the detention effect of lakes and other storage is not shown. Figure 23. Computed Runoff Travel Time to the U.S./Canadian Border (Minnesota Department of Natural Resources, Waters) RRB FDR Framework Final.doc 22

Definition of Early, Middle, and Late Areas Relative to the Red River Main Stem Figure 24 identifies early, middle, and late runoff areas within the Red River basin relative to the main stem at the Canadian border. This generalized map was based on the evaluations of historical flood hydrographs, knowledge of more recent floods, and computed runoff travel times. This map can be used to help define which types of FDR measures to use in different areas of the basin to help reduce peak flows on the Red River main stem, while also achieving local and watershed FDR goals. The lines between early, middle, and late areas are not exact. For example, smaller late areas may exist within the identified middle area. Therefore, this map should be used in conjunction with local knowledge of runoff timing. RRB FDR Framework Final.doc Figure 24. Early, Middle, and Late Runoff Timing Zones in the Red River Basin 23

Flood Damage Reduction Measures Flood damage reduction measures can be grouped into the four general categories outlined below. These categories and measures are listed here and discussed in more detail in subsequent sections. 1) Reduce Flood Volume a) Restore or create wetlands (providing infiltration and evapotranspiration) b) Use cropland BMPs (increase infiltration and evapotranspiration) c) Convert cropland to prairie or other types of perennial grassland (e.g., Conservation Reserve Program (CRP) and Reinvest in Minnesota (RIM), to increase infiltration and evapotranspiration) d) Convert land use to forest (forested areas generally have the lowest runoff coefficients, due to high interception and evapotranspiration) e) Other beneficial uses of stored runoff 2) Increase Conveyance Capacity a) Channelization (increasing the flow capacity of existing channels or flowages) b) Drainage (creating new or improved conveyance capacity) c) Diversions (of flood waters around a current damage area) d) Setting back existing levees (to restore floodway capacity) e) Increasing road crossing capacity 3) Increase Temporary Flood Storage a) Impoundments (with or without a normal pool, to detain water in excess of downstream channel capacity) b) Restored or created wetlands (functioning as impoundments) c) Drainage (to lower surface water and groundwater levels, which increases infiltration and temporary storage in the upper soil horizons) d) Culvert sizing (to increase temporary storage by widespread metering of runoff close to its source) e) Setting back existing levees (to restore floodplain storage areas) f) Overtopping levees (to utilize diked floodplain storage capacity when critically needed) 4) Protection/Avoidance a) Urban levees b) Farmstead levees c) Agricultural levees d) Evacuation of the floodplain (removing people and flood-prone facilities and converting to more flood-compatible land uses) e) Floodproofing f) Flood warning and emergency response planning RRB FDR Framework Final.doc 24

Many projects will combine two or more of the above measures. Appropriate application of each type of measure for local, watershed and main stem flood damage reduction is highly dependent on its design and location. The most important factors are how flood flow volume and timing are affected. 1) Reduce Flood Volume Activities in this category reduce the volume of water that enters the downstream channel during a flood. The water may be conveyed to the atmosphere by evapotranspiration, conveyed to groundwater by infiltration, or retained until after the flood for other beneficial uses. Infiltrated water can resurface as base flow sometime after the runoff event, or percolate to deeper groundwater. Water stored during a flood for other purposes, such as stream flow maintenance or water supply, is gradually released or used during non-flood periods. a) Construction or Restoration of Depressional Wetlands Depressional areas within the landscape capture runoff and allow time for evaporation and infiltration to occur, which normally results in natural seasonal drawdown. This drawdown storage is replaced during subsequent runoff events which reduces the downstream flood volume. Most depressional areas only have enough capacity to retain an initial portion of the runoff associated with major flood events. Therefore, from the standpoint of timing to reduce main stem flood peaks, these will be most beneficial when located within late contributing areas. Depressional wetlands located in the southwest area of the Red River Basin will typically provide somewhat greater available retention capacity than those located in northeastern areas, because of the differing evaporation rates, as shown on Figure 7. Similarly, wetlands located in more permeable soil areas may lose more water to infiltration and, therefore, provide more available retention capacity than wetlands located in low-permeability soils. TSAC Technical Paper No. 12, Wetland Hydrology and Biodiversity in the Red River Basin, identified significant seasonal drawdown of wetlands during the summer, fall and/or winter, due to evaporation, transpiration, infiltration, and/or operation, for many of the 28 wetlands studied. Seasonal drawdown storage is also referred to as permanent storage, which is very valuable for flood damage reduction, because it completely removes volume from the spring runoff hydrograph. The amount occurring naturally depends on site conditions and climatic variables of precipitation and evaporation, as well as the size and runoff characteristics of the watershed above the wetland. Construction and operation of drawdown capability for restored wetlands can enable seasonal drawdown in wetlands with large watershed to wetland area ratios that might not otherwise have seasonal drawdown before the spring runoff. RRB FDR Framework Final.doc 25

b) Cropland BMPs Cropland management practices have been developed to conserve soil and water resources. These are collectively referred to as best management practices (BMPs). The most commonly used agricultural BMPs are forms of conservation tillage that leave the soil better protected by crop residues than other tillage methods. This may also increase infiltration, thereby reducing runoff. The reduction in runoff varies with the topography, amount of rain, and type of soil. Based on Natural Resources Conservation Service (NRCS) runoff estimating procedures, a reduction in runoff of about 5% to 8% typically may be expected with conservation tillage practices. This method is appropriate to apply in all areas of the basin, but requires large areas of application to substantially reduce peak flows in rivers. c) Conversion of Cropland to Perennial Grassland Perennial grassland including CRP, hay meadow, and well-managed pasture produce much less rainstorm runoff than cultivated cropland. A typical reduction in storm runoff is estimated to be about 50% compared to row-cropped lands with conventional tillage methods. However, the effects on snow accumulation and spring snowmelt runoff have not been well documented. Using NRCS runoff estimating procedures, it can be shown that the greatest reduction in storm runoff through conversion of cropland to perennial grassland is achieved on lighter soils such as sandy or silty loams. d) Conversion of Land Use to Forest Forestland produces much less rainstorm runoff than cultivated cropland. A typical reduction is estimated to be about 55%. Using NRCS runoff estimating procedures, it can be shown that the greatest reduction in storm runoff is achieved on lighter soils such as sandy or silty loams. The effects on snow accumulation and spring snowmelt runoff from forestland have not been well documented. e) Other Beneficial Uses of Stored Water Stored water can be used for domestic or industrial purposes, or for stream flow augmentation during drier periods of the year to improve fish habitat and provide other instream flow benefits. Use of this water results in drawdown of the reservoir, providing annual removal of water from the spring flood volume. 2) Increase Conveyance Capacity Conveyance is the ability to move water downstream. It includes both channel and floodway capacity. Conveyance improvement has usually been accomplished by increasing channel capacity. It may also be accomplished by removing channel and floodway obstructions. RRB FDR Framework Final.doc 26

Historically, increasing conveyance has been a predominant flood damage reduction measure in the Red River Basin. There is little doubt that conveyance improvements have provided local benefit. However, there has been much study and debate about the effects on downstream flood flows. Conveyance improvement projects can affect downstream peak flood flows, depending on location and scope. Accelerating early water will reduce downstream flood peaks, whereas accelerating late water will increase downstream flood peaks. The impact of conveyance improvement projects on downstream flows can be predicted by the same analysis as storage projects. They simply have the opposite effect. That is, conveyance projects generally reduce floodplain storage and accelerate flood flows, which may, or may not, increase flood peaks, depending on location and the characteristics of a specific flood. a) Channelization Channelization projects may include enlarging or realigning natural channels or creating channels in areas of diffused or overland flow. A number of major tributaries in the Red River Basin have been channelized. The State of Minnesota, district courts, and counties implemented many of these projects in the early 1900s to facilitate agricultural settlement of the area. The Federal Government has also been involved through the Army Corps of Engineers, the Natural Resources Conservation Service (formerly Soil Conservation Service), and the Farm Services Agency (formerly Agricultural Stabilization and Conservation Service). There are often serious environmental concerns related to altering large portions of natural channels. In recent years, channelization projects have primarily been limited to improvements of previously altered channels. Channelization projects are usually done to decrease local flooding. The hydrologic effect is a decrease of floodplain storage, acceleration of flow, and a corresponding increase in local peak flood flows on the channelized stream. If channelization is implemented on a tributary stream, the impacts on main stem flood peaks will depend on relative locations and timing. Projects located in early water areas relative to the main stem will tend to reduce main stem flood peaks, whereas projects located in late water areas will tend to increase main stem flood peaks. b) Agricultural Drainage The primary purpose of agricultural drainage projects in the Red River Basin is to remove excess surface water and soil moisture. This allows the ground to warm up faster in the spring, provides an aerated rooting zone for crop development, and minimizes drowning of crops by excess precipitation. A large portion of the basin, especially within the Lake Plain and Lake Washed Till Plain regions, requires drainage in order to be agriculturally productive. The need for outlets for field drainage led to the development of larger collector ditch systems. Many of these were established as legal drainage ditches, which are systems implemented by a drainage authority for the benefit of local landowners. The associated costs are assessed to the benefiting landowners. RRB FDR Framework Final.doc 27

In designing drainage systems, an important consideration is the capacity to convey surface runoff flows. Crop damage occurs when water inundates crops for more than a brief period of time. Damage depends on the crop, its stage of development, and heat stress. Most crops can tolerate 24 to 48 hours of inundation. Therefore, most of the drainage systems have been designed to remove frequent floodwater (i.e., 2- to 10-year events) within a 24-hour period. The primary determining factors in choosing a design protection level have been cost and outlet capacity. Increasing the capacity of a drainage channel will reduce the frequency of adjacent land flooding, but will have a related increase in peak discharge rates immediately downstream. The impact on flood-prone areas farther downstream will depend on relative locations. Removing early water faster will decrease main stem flood peaks. Removing late water faster will increase main stem flood peaks. Determinations of the adequacy of an outlet should consider the effects on downstream flooding, both within and downstream from the drainage system, for various magnitudes of flood events. c) Diversions Diversion projects typically remove water from a flood-prone stream, convey it safely around a significant damage site, and return it to a downstream watercourse. A diversion is an alternative to channelization or protection measures, such as levees and floodwalls, when environmental impacts, cost, or other land use issues are better addressed by this measure. Diversions may include operable controls. Operation can consider upstream, downstream, and local flood conditions. This allows for optimizing benefits while minimizing potential adverse impacts. d) Setting Back Existing Levees Levees constructed along flood-prone waterways often restrict conveyance enough to cause a backwater effect by encroaching on the floodway. Moving the levees back farther away from the channel will restore a portion of the lost floodway capacity. Doing so with a primary purpose of increasing conveyance will primarily benefit lands upstream from the levee encroachment. The downstream effects in this situation may include an increase in peak flows, due to the reduction in upstream floodplain storage. However, this may be offset by increased floodplain storage within the setback levee reach. e) Increasing Road Crossing Capacity Road crossings typically restrict conveyance. Often, the road approaches block the floodway on both sides of the channel, and the bridge or culvert is smaller than the channel. In many cases, this is not a problem because the effect of the bridge constriction does not extend far upstream. However, in flatter areas of the Red River Basin, roads, bridges, and culverts can be a major factor affecting upstream flood elevations, due to backwater effects. RRB FDR Framework Final.doc 28

In areas where the upstream flood damage potential is high, some of the lost conveyance can be restored by increasing bridge or culvert size. Another method, which may be less expensive, is to lower the approach road grades to allow for overtopping at an elevation lower than a critical upstream flood damage point (provided traffic safety concerns are adequately addressed). Increasing road crossing capacity may increase downstream peak flows, due to reduction in upstream floodplain storage. 3) Increase Temporary Flood Storage Activities in this category are designed to store water during a flood and release it after the flood peak or, more beneficially, after downstream floodwaters have subsided. Another facet is to better manage existing storage. A huge amount of flood storage already exists in the Red River Basin. However, most of it is floodplain storage. The floodplain lands are often very productive for agriculture and include desirable areas for urban development. Therefore, this naturally occurring floodplain storage and associated overland flows often cause substantial damage to buildings, roads, bridges, fields, and crops. One approach for flood damage reduction is to find more acceptable places to store water. Another approach is to manage the floodplain storage in ways that reduce the associated damages and maximize the downstream benefits. a) Impoundments Impoundments are reservoirs constructed to temporarily store (impound) floodwater. These are often thought of only as dams on major rivers or tributary streams (i.e., on-channel ). However, there are many impoundments that are off-channel and connected to streams and rivers by diversions, or fed by collector ditches. Impoundments can be wet (having a permanent pool), dry (not having a permanent pool), or something in between when, after the flood pool is released, a conservation pool is released over a longer period of time for streamflow augmentation and/or vegetation management within the conservation pool area. The most important consideration from an overall flood control standpoint is the timing of the storage and release. The design and operating goal should be to store water that would otherwise contribute to downstream flood peaks and to avoid causing damages during the subsequent release of the stored floodwater. In relation to maximizing downstream benefits, impoundments are most effectively located in the middle and late areas of the basin. Impoundments located in a late area should be designed to store the early water on the rising limb of the local hydrograph to help reduce main stem peak flows. Impoundments located in a middle area should be designed to store the peak of the local hydrograph. Impoundments located in early areas of the basin may also be beneficial to the main stem, if they are designed to store the falling limb of the local hydrograph. This would usually require either a very high capacity storage site to store all of the floodwater, or high capacity gate RRB FDR Framework Final.doc 29

works that can pass the early flows and be closed to store the late flows. This also requires substantial flood monitoring and prediction capability. 1) On-Channel Impoundments On-channel impoundments can be a very effective means of controlling river flow and providing FDR benefits downstream. The valleys of rivers and streams can provide substantial floodwater storage relative to the size of dam needed. There can be serious environmental concerns associated with this type of impoundment. Consideration must be given to the potential effects on fish migration, riparian corridor interruption, wetland impacts, sedimentation, channel stability, and other potential issues. In general, on-channel impoundments with large, normally open control gates, and with no permanent pool (also referred to as normal, or conservation, pool), will have less adverse environmental impact than on-stream impoundments with substantial permanent pools. Therefore, the decision whether to include a permanent pool requires weighing the potential benefits of the pool against the potential impacts to the stream. On-channel sites on high quality natural streams should be avoided, if possible, in favor of less environmentally sensitive alternatives or sites. See TSAC Technical Paper No. 13, On- Channel Storage Assessment for further discussion and guidance regarding early coordination to identify levels of concern, alternative approaches, and efficient decisionmaking. 2) Off-Channel Impoundments Storage reservoirs may be constructed away from a natural or constructed channel, connected by inlet and outlet facilities. In areas of flat topography, the storage site may require dikes on all four sides to contain water. The inlet works for an off-channel storage facility may be designed to remove a percentage of the flow from the stream, or they may be designed to remove high flow only. Projects intended to remove a percentage of flow will have an inlet channel capacity in proportion to stream channel capacity. Those intended to remove only high flow will usually have an overflow weir as the inlet to the off-channel impoundment. The distinction is important from a main stem perspective. Those removing peak flows are more beneficial located in middle areas of the basin. Those removing a percentage of flow are more beneficial located in late areas of the basin. Gates may be included at the inlet and the outlet to control the amount of water being stored at any time and associated outflows. b) Restored or Created Wetlands Wetlands are restored or created for a variety of reasons, which may include providing temporary flood storage. The distinguishing characteristic of a flood storage wetland will usually be a RRB FDR Framework Final.doc 30

restricted capacity, or gate-controlled, outlet. The restriction may be a small pipe, orifice, or notched outlet. The controlled outlet may have a lift gate or stop log bay. In general, stop logs provide the best level control during non-flood periods because they function as an adjustable overflow weir. Compatibility with other wetland purposes may limit the amount of water level rise (bounce) that can be tolerated. For typical multipurpose wetlands, TSAC Technical Paper No. 1, An Overview of the Impacts of Water Level Dynamics ( bounce ) on Wetlands, suggested that bounce should be limited to about 2 feet during a 10-year, 24-hour summer rain storm event and most of the bounce water should be removed within 10 days to minimize impacts to vegetation and wildlife habitat. Spring flood bounce generally can be greater than indicated above without adverse impacts, provided the water levels are returned to normal by early in the growing season. Wetlands with temporary flood storage are most beneficial for main stem flood control when located in middle and late contributing areas of the basin. c) Drainage In areas where the ground is saturated, its ability to absorb and retain water may be enhanced by surface or subsurface drainage. From a flood control standpoint, this is done most effectively by installing seepage tile to lower the water table. If the tile is moderately sized, its subsurface drainage rate will be small compared to surface runoff, contributing little to downstream peak flows, and the increased soil storage capacity will reduce surface runoff. It is important to understand the context wherein this measure will be beneficial for downstream flood control. Everything else being equal, subsurface drainage measures will produce lower peak outflow rates than surface drainage measures, because of increased storage capacity in the soil and more limited flow capacities. However, the peak flow reduction benefits may be overshadowed if adding subsurface drainage is accompanied by change in land use that increases runoff potential or by loss of depressional storage. The most appropriate application for drainage is on existing agricultural land where a high water table restricts the crop rooting depth. The additional temporary storage capacity provided will be most beneficial for main stem peak flow reduction when implemented in middle and late areas relative to the main stem. d) Culvert Sizing Runoff, particularly from well-drained cropland areas, typically accumulates quickly, often developing flows greater than downstream channels can carry. This is especially true where surface ditches serve small drainage areas. These ditches, in order to be deep enough to provide adequate subsurface drainage, tend to have much more capacity than necessary to remove flood flows within acceptable time frames. The result often is a high rate of runoff from upstream areas and flooding farther downstream. High runoff rates provide little value to the cropland because RRB FDR Framework Final.doc 31

most crops are tolerant of short-term flooding. High flow rates are also related to soil erosion and associated water quality problems. Culvert sizing is a technique that can be used to control runoff rates. By appropriately sizing road and drainage system culverts throughout a subwatershed or watershed, the flow rates can be regulated to better match downstream channel capacities. Excess water is temporarily detained upstream from culverts. This approach can provide a more consistent level of protection throughout a drainage system. If properly done, it should also result in reduced road repairs and increased traffic safety. At some locations, road safety may dictate using two stage inlets, raising the road grade, or taking other measures to prevent overtopping. Culvert sizing provides relatively short-term storage. It is most effective in reducing main stem flooding if implemented in middle and late contributing areas of the basin. e) Setting Back Existing Levees Levees are constructed to protect floodplain areas from flooding. Doing so reduces the amount of floodplain storage that otherwise helps to naturally attenuate downstream flood peaks. Moving existing levees back, farther away from the channel, will restore a portion of the lost floodplain storage, benefiting downstream areas by helping to attenuate flood peaks. However, there may be a corresponding negative effect on downstream areas, due to increased floodway capacity that decreases backwater effects and may reduce upstream floodplain storage. f) Overtopping Levees Floodplain areas gradually fill on the rising limb of a flood hydrograph and gradually empty on the falling limb. Therefore, much of the floodplain storage capacity is used up before the flood peak, and much of that storage is released while flooding is still occurring downstream. Building levees in the floodplain that are designed to overtop during major flood events will save the floodplain storage for use during the peak period of a large flood. When the levees overtop, all the available storage will be used to reduce peak flows. Following the peak, stored water can be retained until its release will not increase downstream flood damages. Building overtopping levees in agricultural areas can protect cropland behind the levees from relatively minor, frequent floods up to about a 25-year event, while improving the timing and duration of floodplain storage for larger floods. Lands behind overtopping levees get protection from frequent floods at the cost of increased flood periods for larger, less frequent floods. This can result in a substantial net reduction in flood damages for lands behind the levees, as well as for lands downstream of the overtopping levees where peaks of the larger floods are reduced. For lands outside the levee-protected area, the net effect will usually be higher peak flood stages during lesser floods and lower peak flood stages during greater floods. This is due to decreased floodplain storage during lesser floods and improved timing of floodplain storage during greater RRB FDR Framework Final.doc 32

floods. The levee may restrict the floodway, causing higher water levels upstream, which also affects the amount of floodplain storage. This effect, if present, needs to be considered in the overall analysis. Overtopping levees are most effective when located at or near the area that needs protection, because proper timing of the storage is automatically provided. These levees will be most beneficial for the main stem when located in early and middle runoff timing areas of the basin. 4) Protection / Avoidance Protection and prevention measures are aimed at reducing flood damage potential. Ideally, their implementation would not adversely affect flooding in other areas. However, some measures, levees in particular, may affect flood flows and elevations by significantly constricting floodways and reducing floodplain storage. a) Urban Levees Urban levees have been built extensively in the Red River Basin. Typically, they were built as an emergency measure in response to an immediate flood threat. Over time, many of these emergency levees have been upgraded to provide a substantial level of permanent protection. Levees do not require cooperation from other areas in the basin and can be implemented relatively quickly. These advantages have often made levees the FDR method of choice. However, there are several reasons why urban levee construction should be carefully considered. Levees can create a false sense of security that may lead to unwise development within the protected floodplain area. Levees have the potential for catastrophic failure with rapid inundation of the protected area. Levees may raise flood elevations upstream, due to constriction of the floodway, or downstream, due to loss of floodplain storage. Urban levees tend to isolate the community from the river unless greenways, trails, and/or other river corridor access components are included in the design, which can substantially increase costs. Construction of permanent levees may make other flood control alternatives with more widespread benefits (such as temporary upstream storage, or reduced flood runoff) less feasible to construct, due to reduced incremental benefits within the levee-protected area. Levees may be appropriate to protect flood-prone urban areas at any location within the basin. However, this FDR measure is most appropriately used where other options that benefit more than just the levee-protected area are not available. When used in conjunction with temporary flood storage and/or flood runoff reduction, the height and cost of the levee can be reduced and the level of protection of the leveed area increased. b) Farmstead Levees Farmstead levees are similar to urban levees except that the area protected typically is small and located within a rural floodplain or floodway. Farmstead levees are often referred to as ring RRB FDR Framework Final.doc 33

dikes because they are normally constructed as a dike that completely surrounds the farmstead. c) Agricultural Levees Agricultural levees protect entire fields, or agricultural areas, and have been built extensively within the Red River Basin. In many cases, they have been implemented by simply installing flap gates on culverts under road embankments or ditch spoil banks. Levees can easily protect agricultural land from frequent smaller floods. This is especially useful where ditch overflows might cause serious field erosion. However, protecting agricultural land from flooding can substantially increase downstream flood flows, due to loss of floodplain storage and/or constriction of a floodway. The most important consideration in the application of agricultural levees is the overtopping elevation. In general, the levees will cause an increase in all flood peaks up to the overtopping elevation and a decrease in all flood peaks above that elevation. By carefully selecting the overtopping elevation, the balance of benefits may be optimized. Unfortunately, the benefits and damages may not be equally distributed. In other words, the lands damaged by the increase in peak flows and elevations for small floods may not be the same lands that benefit by the decrease in larger flood peaks. This calls for a comprehensive approach to managing agricultural levee construction in any given area. d) Evacuation of the Floodplain Some areas of the floodplain are so low that they flood frequently, or are very difficult to protect. In these areas, it may be more practical to eliminate uses of the floodplain that are incompatible with flooding. In urban areas, this means removing residential, commercial, and industrial development, along with the associated infrastructure, from at least portions of the floodplain. The evacuated areas may be converted to parkland or greenways, thereby enhancing the urban environment. In agricultural areas, this means removing farmsteads and converting existing cropland to flood tolerant agricultural or non-agricultural use. Establishing greenbelt corridors along flood-prone streams is an example of environmentally beneficial, non-agricultural use. Converting from annual crops to perennial grasslands or forests, or other water-tolerant species, such as wild rice, are examples of agricultural conversions to reduce flood damage potential. This measure may be appropriately applied in any flood-prone area of the basin where it is feasible. Its application reduces the need for other flood control measures. Therefore, a vision of reasonably foreseeable floodplain evacuation should be included in a comprehensive flood damage reduction strategy. RRB FDR Framework Final.doc 34

e) Floodproofing Floodproofing means making flood-prone property resistant to damage. It includes raising buildings and essential access routes above the flood level and using flood resistant materials or construction techniques. This measure may be appropriately applied in any flood-prone area of the basin where it is feasible, because it typically will not significantly affect downstream flood peaks. f) Flood Warning and Emergency Response Planning Flood warning and emergency response planning are an established flood damage reduction method in the Red River Basin. They begin with long- and short-term forecasts of flood potential and lead to sandbagging, earthen levee construction, or other emergency protection methods, and ultimately evacuation, if necessary. Flood forecasting has become more accurate with experience and improved monitoring and forecasting techniques. However, the ability to accurately predict floods, especially those larger than previously experienced, should not be depended on. Emergency response plans must include the ability to react to changing conditions. Most important is having a mechanism for determining when to fight the flood and when to evacuate. Summary Table of Flood Damage Reduction Measures and Effects In Table 1, flood damage reduction measures are rated in terms of appropriateness for local and downstream flood damage reduction, based on location in the watershed in relation to timing of runoff to the main stem. A plus sign (+) indicates application of a particular flood damage reduction measure would normally have a positive effect downstream on the main stem of the Red River or the lower reaches of its major tributaries (i.e., it would result in a reduction in downstream peak flows). A minus sign (-) indicates a likely negative effect on downstream flooding, and a zero (0) indicates a likely insignificant effect on downstream flooding. Double plus signs (++) and double negative signs (--) indicate more substantial positive or negative effects on downstream flooding. RRB FDR Framework Final.doc 35

Table 1. Expected Peak Flow Reduction Effects on the Red River Main Stem of FDR Measures Applied in Early, Middle, and Late Areas Upstream Flood Damage Reduction Measure Early* Upstream Area Middle* Upstream Area Late* Upstream Area 1) Reduce Flood Volume + ++ ++ a) Wetlands + + ++ b) Cropland BMPs + ++ ++ c) Conversion to grassland + ++ ++ d) Conversion to forest + ++ ++ e) Other beneficial uses of stored water + ++ ++ 2) Increase Conveyance Capacity + - - - a) Channelization + - - - b) Drainage + - - - c) Diversion + Variable - d) Setting back existing levees (to increase conveyance capacity) + - - - e) Increasing bridge capacity + - - 3) Increase Temporary Flood Storage Variable ++ + a) Gated impoundments + ++ ++ b) Ungated impoundments - + + c) Restored or created wetlands - + + d) Drainage - + ++ e) Culvert sizing - + + f) Setting back existing levees (to increase floodplain storage) + ++ + g) Overtopping levees ++ + Variable 4) Protection/Avoidance Variable Variable Variable a) Urban levees - - - b) Farmstead levees - - - c) Agricultural levees - - - d) Evacuation of the floodplain 0 0 0 e) Floodproofing 0 0 0 f) Warning and emergency response 0 0 0 * Location of FDR measure relative to the Red River main stem at the international border. RRB FDR Framework Final.doc 36

A Basin FDR Framework Flood control measures will never totally eliminate flood damage potential, or risk, as there will always be the possibility of the occurrence of a greater flood than FDR measures are designed to withstand or control. However, a basin-wide FDR framework should substantially reduce the risk of flooding and realize measurable flood damage reduction from local, watershed and basin perspectives, and should include a realistic appraisal of the remaining long-term flood risks. Flooding and related water management problems exist throughout the Red River Basin. Without a comprehensive basin-wide strategy, it is probable that individual solutions will work against, rather than toward, a basin-wide solution. Solutions to problems in upstream areas, if properly selected, can provide part of the solution for downstream areas as well. The types of upstream local solutions that have the best potential for also providing watershed and main stem benefits are flood volume reduction and increased temporary storage. Therefore, a basin-wide FDR framework should include a substantial focus on these types of flood damage reduction measures. However, because these types of measures alone cannot achieve the FDR and NRE goals envisioned in the Mediation Agreement, a multi-measure approach is needed. The relationship between storage volume and peak flow reduction on the main stem is illustrated by the hydrograph and curve in Figure 25, which presents a relationship between ideal storage (and/or flood volume reduction) and peak flow reduction for the 100-year flood event at Emerson. Ideal storage or flood volume reduction is a theoretical minimum, based on 100% effectiveness for reducing the peak flood flow on the main stem. In reality, much more temporary storage or flood volume reduction would need to be implemented upstream to achieve the effects of the associated ideal volume in the main stem flood hydrograph. Figure 25 indicates that for a 10%, 20% or 30% reduction of the 100-year main stem peak flow at Emerson, the ideal storage and/or flood volume reduction is about 55,000 acrefeet, 250,000 acre-feet, and 540,000 acre-feet, respectively. Figure 25. Volume of Ideal Storage and/or Flood Volume Reduction Required to Reduce the 100-Year Peak Flow at Emerson The amount of implemented volume required depends on the types of projects and their locations in the basin. It also depends on project-specific design and operation. The effectiveness of RRB FDR Framework Final.doc 37

storage will also vary with the total amount of storage and runoff reduction provided in the basin relative to the size of the flood. Small amounts of storage relative to the flood volume will have small effects on the main stem flow, while the effectiveness will increase to approach 100% as the basin storage approaches the total volume of the flood. For these reasons, it is difficult to estimate the total amount of implemented volume required, and any estimate given will depend on assumptions that are made. Water management challenges in the Red River Basin are not only related to flooding. Quality of life is also affected by fish and wildlife resources and recreational opportunities, and future development may be limited by inadequate water quality and supply. Channel stability and maintenance of low flows are related challenges in the Red River Basin. The sustainability of the region is substantially dependent on sound water management that protects and enhances both man-made infrastructure and natural resources. Therefore, to the extent possible, FDR projects should also address these related water resource issues. Multi-purpose project opportunities include temporary storage and/or runoff reduction for flood damage reduction and: waterfowl and other wildlife habitat creation and management; peak flow reduction for improved channel stability downstream; low flow augmentation for fishery improvement (via surface water control and infiltration); low flow surface water and ground water augmentation for water supply (via surface water control and infiltration); and/or sediment control and nutrient removal. The following example illustrates the use of multiple FDR and related NRE measures and estimates the combined effect on Red River main stem flood peak reduction. Multi-Measure Example This multi-measure example uses various types of FDR and related NRE measures to demonstrate the combined effects on reducing downstream flood volumes and peak flows. The point of reference is Emerson, Manitoba (U.S./Canada border) at the downstream end of the Red River Basin in the U.S. This example suggests that temporary storage and flood volume reduction can be provided by various types of FDR and related NRE measures. Based on the experience and judgment of the TSAC, the amount estimated to be provided by each measure is considered reasonable within the U.S. portion of the Red River Basin over the long term. The mix of measures implemented will likely vary, but the total amount should remain reasonable. The effectiveness of each measure (i.e., the ratio of implemented to ideal storage or flood volume reduction) has also been estimated on the basis of what is considered reasonable to achieve, on average. Here again, there will be variability from project to project. RRB FDR Framework Final.doc 38

The example presented here is not meant to dictate specific measures or projects. This is better left to the project development process wherein NRE needs, as well as economic and environmental feasibility, funding availability and other factors can be considered along with local, watershed and basin-level FDR needs. The example flood is a 100-year spring flood. Some measures are more or less effective depending on the time of year and size of flood. Therefore, the measures chosen and their estimated effectiveness would likely be different for different floods. Temporary Storage and Flood Volume Reduction Components a) Wetland Restoration / Creation The example includes an increase in multi-purpose wetlands with a flood control component equivalent in area to 0.5% of the contributing Red River Basin drainage area of 36,400 square miles in the United States. (This does not include 3,800 square miles in the Devils Lake Basin.) A total of 190 square miles of flood control wetlands would be constructed. TSAC guidelines, as presented in Technical Paper No. 1, An Overview of the Impacts of Water Level Dynamics ( bounce ) on Wetlands, recommended 2 feet of bounce and a 10-day drawdown period for 10-year summer flood events. Two feet of bounce would provide 240,000 acre-feet of temporary storage. TSAC Technical Paper No. 12, Wetland Hydrology and Biodiversity in the Red River Basin, identified significant seasonal drawdown of wetlands during the summer, fall and/or winter, due to evaporation, transpiration, infiltration, and/or operation, for many of the 28 wetlands studied. Seasonal drawdown storage is also referred to as permanent storage, which is very valuable for flood damage reduction, because it completely removes volume from the spring runoff hydrograph. The amount occurring naturally depends on site conditions and climatic variables of precipitation and evaporation, as well as the size and runoff characteristics of the watershed above the wetland. Data from TSAC Technical Paper No. 12 indicates that, on average, about 8 inches will be available for spring runoff events. Construction and operation of drawdown capability for restored wetlands can enable seasonal drawdown in wetlands with large watershed to wetland area ratios that might not otherwise have seasonal drawdown. Seasonal drawdown of an average of 8 inches would provide about 80,000 acre-feet of permanent storage. For this example, the total estimated flood control volume provided by multi-purpose wetland restorations is 320,000 acre-feet. Wetlands are also expected to be restored primarily for NRE benefits. The total area envisioned is also equivalent to 0.5% of the contributing drainage area in the United States, or about 190 square miles. The primary FDR benefit of NRE wetlands is seasonal drawdown. It is estimated that these wetlands will also provide about 8 inches of seasonal drawdown storage, on average, which is equivalent to 80,000 acre-feet of flood volume reduction. RRB FDR Framework Final.doc 39

b) Culvert Sizing This example assumes that 15% of the contributing drainage area in the United States could be treated in middle and late areas relative to the Red River main stem, detaining an average of 1 inch of runoff from 5,670 square miles. The total estimated flood control volume provided by this measure is 53 acre-feet per square mile, or about 300,000 acre-feet. Culvert sizing at road crossings typically provides only short-term storage. Therefore, the estimated effectiveness for reducing the 100-year flood peak on the Red River main stem is relatively low. The effectiveness for local and watershed FDR is expected to be much higher. c) Overtopping Levees The example includes 200 square miles of agricultural diked storage area in the early and middle areas of the basin, with overtopping at an elevation designed to provide peak flow reduction. The levees envisioned would be designed to protect agricultural lands from minor floods and to provide critical peak storage during major flood events. An average of 2 feet of storage is assumed. The total estimated flood control volume provided by this measure is about 250,000 acre-feet. Note that the benefit is due to a change in timing of the floodplain storage rather than an increase in the total floodplain storage. d) Impoundments This example includes various sizes and types of flood control impoundments that would be predominantly gated to provide controlled timing of storage and release. The impoundments would be strategically located and operated to provide local, watershed and main stem flood control. The storage capacity would be equivalent to an average of 2 inches of runoff from 10% of the total contributing drainage area within the United States. The total estimated flood control volume provided by this measure is 400,000 acre-feet. This is potentially very high value storage for FDR. Its effectiveness depends on location, design, and operation. TSAC Technical Paper No. 10, Basin Strategy: Hydrologic Analysis, indicated that adding storage for 1 inch of runoff in subwatersheds can provide about a 10-year level of protection to intensively farmed lands. Additional storage capacity not only affords greater local protection, but also allows design and operation to better address main stem flood damage reduction. About 4 inches of runoff volume control at individual storage sites has been shown to be very effective for both local and main stem flood control. Construction of the above impoundments also affords an opportunity to provide a related benefit of seasonal streamflow augmentation. The amount envisioned is 120,000 acre-feet. This is equivalent to a total release of 200 cfs for 300 days in an average year. Allowance for evaporation from the reservoirs would require an estimated additional 40,000 acre-feet. The total estimated flood volume reduction provided by the addition of streamflow augmentation provisions to the impoundment component of this example is 160,000 acre-feet of seasonal drawdown storage. Seasonal drawdown storage is very beneficial because it reduces the total RRB FDR Framework Final.doc 40

spring flood volume. Summary for Temporary Storage and Flood Volume Reduction Measures The effectiveness of each component included in the above example on reducing the 100-year peak spring flood flow on the Red River main stem is estimated in the spreadsheet shown on Figure 26. The average effectiveness of the individual FDR measures was based on the hydrologic analyses and experience of TSAC members and the advice of others with related expertise. The estimated total implemented storage is 1,510,000 acre-feet upstream from Emerson, Manitoba. This volume is approximately equivalent to 0.8 inch of runoff from the contributing area of the Red River Basin in the United States. The corresponding ideal storage is estimated to be 285,000 acre-feet. This corresponds to about a 21% reduction in the main stem 100-year peak flow at Emerson. Figure 26. Summary of Example for Temporary Storage and Flood Volume Reduction Measures and Estimated Effectiveness to Reduce 100-Year Flood Peak at Emerson It should be noted that the effectiveness of each measure depends on the goals chosen. These include the size of the target flood, the amount of reduction, and the type of flood (summer or spring). In this example, we have chosen to use the 100-year flood as a reference. A 100-year flood can theoretically occur at any time. However, its most probable occurrence on the Red River is during spring runoff when the entire basin can have substantial accumulated precipitation in the form of snow and variable rain during the melt. Some measures, such as wetlands, tend to have a greater effect on smaller floods. Other measures, such as overtopping levees, may actually increase small flood peaks downstream, but are very effective at decreasing large flood peaks downstream. Flood control impoundments can be designed to reduce different floods by different proportions. RRB FDR Framework Final.doc 41

The amount of the peak flow reduction goal changes the significance of the timing and storage duration parameters. The smaller the reduction goal, the smaller the window of ideal timing. For the 100-year flood event, the period of time that flows exceed 80% of peak at Emerson is about 10 days. There are only 5 days above 90%, but 17 days above 70%. Clearly, timing is more critical for small degrees of reduction. On the other hand, duration of storage is more critical for larger degrees of peak flow reduction. It may follow in reasoning that timing is the most important factor in achieving short-term flood peak reduction and that storage duration and flood volume reduction will be more important in the long term. Some measures, such as seasonal drawdown of impoundments and wetlands, are typically effective only during spring floods. Other measures, such as land use change, are primarily effective for reducing summer runoff. In light of the above, planners should be cautious of focusing on a single type of FDR measure, or goals that are too specific. The most appropriate plan may be to maintain a balanced approach, considering the short- and long-term effects on all flood sizes and types. Non-Storage Components Although not included in the spring flood reduction estimates, there are additional FDR measures that could further reduce local, watershed, and basin-level flood damages, if applied to substantial areas of agricultural lands, or strategically used upstream from, or at, local flood damage sites. a) Runoff Volume Reduction This subcategory of flood volume reduction measures includes land use conversion to grassland or forest, and cropland BMPs. These measures apply primarily to summer conditions, because of uncertainties of the net effects on snow accumulation and spring runoff. However, summer floods can be the most devastating to agricultural lands and production. Flood volume reduction can be accomplished via agricultural set-aside and conservation easement programs targeted toward marginal lands in middle and late areas, and through promoting more perennial crops and expanded markets for these crops. These trends for land use can be affected by Federal, State, and local programs and policies. b) Protection / Avoidance This type of local FDR measures can complement watershed and basin-level efforts. However, it is important to recognize that these measures generally do not provide significant FDR benefits downstream, or upstream, from the implementation site. RRB FDR Framework Final.doc 42

Effects of Implementing Example Measures Future flood hydrographs would be modified by volume reduction, temporary storage, and release from storage. Figure 27 shows the current estimated 100-year flood hydrograph at Emerson, along with an approximation of what the hydrograph would look like after the example measures were implemented. The 100-year peak flow is reduced by about 20%. The hydrograph illustrates the approximate net result of all implemented flow management measures. On the rising side and near the peak of the hydrograph in area A, the flow is reduced. This is accomplished by removing or delaying flood volume. This volume is greater than the ideal storage volume shown as area a1. The additional volume is incidental to providing peak reduction. It reflects the fact that some water will be removed earlier than necessary. This results in a delayed flood peak, which may be beneficial by providing additional time to prepare for a flood, or by allowing time for more ice jams to clear ahead of the peak flow. It also makes conveyance improvements in early and middle areas more feasible as a main stem FDR measure. Figure 27. Estimate of Modified 100-Year Flood Hydrograph at Emerson On the falling side of the flood hydrograph, in area B, flows are increased relative to the current hydrograph. This is caused by release of temporarily stored floodwater. Some will be RRB FDR Framework Final.doc 43

automatically released from ungated impoundments and wetlands. Some may also be intentionally released from gated impoundments during the flood in order to restore a portion of their capacity. Although these discharges do not increase the flood peak, they do delay the recession of water from flooded areas. When the flood has receded below the significant damage level, the remaining temporary flood storage can be released as shown in area C of the hydrograph. Water released in area C has been detained by gate controlled storage an average of 55 days. Releasing the water as soon as possible after flood waters have receded will restore storage capacity to handle summer storm events. Implementing protection measures to raise the significant damage threshold would allow earlier release. The remainder of the removed flood volume is essentially eliminated from the flood event. A portion of this water is beneficially used to restore drawdown levels in pools and wetlands. Another portion is retained to provide beneficial streamflow augmentation during low flow periods beyond area D on the hydrograph. Flood Storage Assessment and Tracking Method The spreadsheet shown on Figure 26 provides a computational method for assessing the Red River main stem peak flow reduction benefits of the example temporary storage and flood volume reduction measures located in early, middle, and late areas relative to the main stem. This method could be used to track main stem effects of various implemented FDR and related NRE measures and progress toward long-term FDR goals. The estimates for effectiveness of FDR measures can and should be validated or improved through future monitoring and research. Summary of FDR Framework The framework presented is a basin-wide approach. It puts priority on identifying and implementing FDR measures that provide local, watershed and Red River main stem benefits. It also integrates NRE measures and goals, recognizing that many NRE projects can also provide FDR benefits. Multi-purpose projects are recommended to the extent practical and feasible. The critical concepts presented in this framework may help create policies and trends to achieve basin-wide FDR and NRE goals. The multi-measure example presented defines and quantifies temporary storage and flood volume reduction measures that could reduce the peak of the 100- year main stem flood at the U.S./Canada border (Emerson gage) by approximately 20%, while providing substantial local and watershed level flood damage reduction and NRE benefits. RRB FDR Framework Final.doc 44

References Hershfield, David M. 1961. Rainfall Frequency Atlas of the United States for Durations from 30 Minutes to 24 Hours and Return Periods from 1 to 100 Years. Technical Paper No. 40. Cooperative Studies Section, Hydrologic Services Division, U.S. Weather Bureau. Prepared for Engineering Division, Soil Conservation Service, U.S. Department of Agriculture. Washington, D.C. Houston Engineering, Inc. 1999. Hydrologic Routing of the 1997 Red River Flood. Prepared for the Red River Watershed Management Board. Krenz, Gene and Jay Leitch. 1993. A river runs north: managing an international river. Red River Water Resources Council. 1998 Addendum prepared by the Red River Water Resources Council. Kuehnast, Earl L., Donald G. Baker and James A. Zandlo. 1982. Climate of Minnesota, Part XIII-Duration and Depth of Snow Cover. Technical Bulletin 333-1982. Agricultural Experiment Station, University of Minnesota. Maclay, R.W., T.C. Winter and L.E. Bidwell. 1972. Water Resources of the Red River of the North Drainage Basin in Minnesota. Water-Resources Investigations 1-72, U.S. Geological Survey. McCombs-Knutson Associates, Inc. 1984. Water Resources Engineering/Planning Program for the Red River of the North Basin in Minnesota. Prepared for The Lower Red River Watershed Management Board. Miller, John F. 1964. Two- to Ten-Day Precipitation for Return Periods of 2 to 100 Years in the Contiguous United States. Technical Paper No. 49. Cooperative Studies Section, Office of Hydrology, U.S. Weather Bureau. Prepared for Engineering Division, Soil Conservation Service, U.S. Department of Agriculture. Washington, D.C. Red River Basin Flood Damage Reduction Work Group. 2001. A User s Guide to Natural Resource Efforts in the Red River Basin. Technical and Scientific Advisory Committee. 1998. An Overview of the Impacts of Water Level Dynamics ( bounce ) on Wetlands. Technical Paper No. 1. Red River Basin Flood Damage Reduction Work Group Technical and Scientific Advisory Committee. Technical and Scientific Advisory Committee. Basin Strategy: Hydrologic Analysis. Technical Paper No. 10. Red River Basin Flood Damage Reduction Work Group Technical and Scientific Advisory Committee. RRB FDR Framework Final.doc 45

Technical and Scientific Advisory Committee. 1998. Integration of Flood Damage Reduction and Natural Resource Enhancement in the Red River Basin. Technical Paper No. 8. Red River Basin Flood Damage Reduction Work Group Technical and Scientific Advisory Committee. Technical and Scientific Advisory Committee. 1998. Siting and Design of Impoundments for Flood Control in the Red River Basin. Technical Paper No. 4. Red River Basin Flood Damage Reduction Work Group Technical and Scientific Advisory Committee. Technical and Scientific Advisory Committee. 1998. The Effectiveness of Agricultural Best Management Practices for Runoff Management in the Red River Basin of Minnesota. Technical Paper No. 3. Red River Basin Flood Damage Reduction Work Group Technical and Scientific Advisory Committee. Technical and Scientific Advisory Committee. 1998. Watershed Modeling of Various Flood Damage Reduction Strategies. Technical Paper No. 6. Red River Basin Flood Damage Reduction Work Group Technical and Scientific Advisory Committee. Technical and Scientific Advisory Committee. Wetland Hydrology and the Effect of Water Level Bounce on Natural Resource Wetland Diversity. Technical Paper No. 12. Red River Basin Flood Damage Reduction Work Group Technical and Scientific Advisory Committee. U.S. Army Corps of Engineers. 1988. Technical Resource Service, Red River of the North, Volume I, Timing Analysis. St. Paul District, Corps of Engineers, St. Paul, Minnesota. U.S. Army Corps of Engineers and Federal Emergency Management Agency, Region V and Region VIII. 2003. Regional Red River Flood Assessment Report, Wahpeton, North Dakota/Breckenridge, Minnesota to Emerson, Manitoba. U.S. Soil Conservation Service. Getting the Most Out of Your Raindrop, Hydrology Guide for Minnesota. U.S. Department of Agriculture. St. Paul, Minnesota. U.S. Weather Bureau. Evaporation Maps for the United States. Technical Paper No. 37. U.S. Department of Commerce. U.S. Weather Bureau. Frequency of Maximum Water Equivalent of March Snow Cover in North Central United States. Technical Paper No. 50. U.S. Department of Commerce. Wiche, Gregg J. and Tara Williams-Sether. 1997. Streamflow Characteristics of Streams in the Upper Red River of the North Basin, North Dakota, Minnesota, and South Dakota. U.S. Geological Survey Open-File Report 97-416, Bismarck, North Dakota (http://nd.water.usgs.gov/pubs/ofr/ofr97416/index.html). RRB FDR Framework Final.doc 46

Williams-Sether, Tara. 2000. High-Streamflow Statistics of Selected Streams in the Red River of the North Basin, North Dakota, Minnesota, South Dakota, and Manitoba. U.S. Geological Survey Open-File Report 00-344, Bismarck, North Dakota (http://nd.water.usgs.gov/pubs/ofr/ofr00344/index.html). Williams-Sether, Tara and Gregg J. Wiche. 1998. Streamflow Statistics of Selected Streams in the Lower Red River of the North Basin, North Dakota, Minnesota, and Manitoba. U.S. Geological Survey Open-File Report 98-21, Bismarck, North Dakota. (http://nd.water.usgs.gov/pubs/ofr/ofr9821/index.html). Web Links http://climate.umn.edu http://deli.dnr.state.mn.us/index.html http://www.lmic.state.mn.us http://www.mvp.usace.army.mil/fl_damage_reduct http://www.mvp-wc.usace.army.mil http://waterdata.usgs.gov/nwis RRB FDR Framework Final.doc 47