Chapter 4
PHYSIOGRAPHY
 

INTRODUCTION
The mainstem river systems are the pathways through which surface-water ultimately is transferred out of the river basin. Water enters in the streams and floodplains via surface-water runoff and groundwater flow from the upland areas. The physiography of the uplands is a controlling factor in the flow of water to the floodplains. The objective of the analysis and description in this chapter is to show the interrelationships among the primary physical characteristics: landscape morphology and terrain heterogeneity, soils (including parent material, distribution, and topography), drainage density, wetland extent and land cover. These characteristics determine storage or runoff potential. Thus they contribute to the dynamics of flood processes. Understanding and classifying the terrain provides useful information to assist in extending the results of local analyses or evaluating such extensions.

TERRAIN ANALYSIS
An initial activity of the river-basin analysis was to develop a terrain classification system that characterizes the hydrologic response of different parts of the basin as well as the basin as a whole. Hydrologic response is affected by many factors including land use and management practices; hillslope gradient, aspect, and variance; drainage patterns and density; surficial deposits, soil texture, permeability, water storage capacity, soil hydrologic groups; and land cover. The terrain classes integrate these factors to characterize the hydrologic response of an area. These terrain classes are used as a means to extend limited research and model results on hydrology to other similar areas of the basin and to evaluate alternative land management and structural means to control surface water runoff and stream flow.

The Upper Mississippi River Basin (figure 4.1) is physiographically, ecologically, and climatically diverse. Physiographic regions include the Rocky Mountains on the western border, the rolling Rocky Mountain outwash areas of the Great Plains, the relatively level glaciated plains in the northern middle parts of the basin, the rolling till prairies and loess uplands on the east, and the dissected plateaus of the Ozarks on the south. Vegetation types include the intensively managed corn belt, the pasture and woodlands of the Ozarks, and the predominantly grassland or wheat areas in the Great Plains.

dem.gif

Preliminary geographic analysis shows that this region can be divided into as many as 70 terrains on the basis of slope variance and aspect (figures 4.2). Further analysis is needed to determine if each of these 70 terrains is unique with respect to hydrologic response. Other land resource maps (Soil Conservation Service, 1981, Agricultural Handbook 296) subdivide the basin into about 44 Major Land Resource Areas (MLRA's). The MLRA's are separated by differences in soils, topography, climate, water resources, potential natural vegetation, and land use.

FIG_4_2.GIF

Recommendation 4.1: Develop a regionalization scheme (based on existing or new data, whichever are appropriate), to prepare hydrologic response units (HRU). These HRU's are used for developing broad scale hydrologic models and for evaluating the effectiveness of different potential nonstructural actions in uplands for reducing flooding.

The State Soil Geographic Database (STATSGO) of the Soil Conservation Service is an inventory of soil resources of the United States at a scale of 1:250,000. Map units in STATSGO are sets of polygons that have similar composition of soil components. Each polygon may have up to 21 different soil components. Slope is one of the attributes for each component. The distribution of slopes within each map unit in STATSGO was analyzed in terms of 5-percent slope intervals from 0 to 70 percent with additional categories for slopes greater than 70 percent, and for water. The distribution was represented as a set of 16 proportions for each map unit. A multidimensional distance between the map units was computed, and a clustering algorithm was used to group the map units into a set of 9 groups. The frequency distributions of slopes within the groups are shown as histograms in figure 4.3. These slope groups are plotted on a regional map to show the areal distribution of slope groups in the basin (figure 4.4). The colors in figure 4.4 depict slope groups sa1 to sa9. Group sa9 is water. Slope groups range from sa1 which has the flattest slope gradients, to sa8 which has the steepest gradients.

f4_3.gif

The mauve color shows the nearly level closed-drainage late-Wisconsin glacial terrain of central Iowa, Minnesota, east central Illinois, and the Dakotas. The linear blue areas within the mauve in northwestern Iowa and southwestern Minnesota correspond to glacial moraines. The area west and south of the Missouri River in North and South Dakota and Nebraska has a wider distribution of slopes corresponding to the rolling nature of this erosional, dissected, and nonglaciated area. The east-west color patterns in the western part of the basin demonstrate the general nature of slopes and the drainage east from the Rocky Mountains. The dissected sloping loess areas along the Missouri and Mississippi Rivers are predominantly in classes sa3 and sa6. This open-drainage terrain is characterized by blue stringers representing the many streams in the area. The dissected area of broad interfluves covered by loess deposits in southern Iowa and northern Missouri is uniquely separated as slope group sa4. This is roughly the Grand River drainage basin.

In very general terms, the entire Upper Mississippi River Basin is characterized by the following landscapes:

SOILS IN THE BASIN
Soils are an important component in the physiography because they affect rates of water runoff to streams, absorption of water into the groundwater system, and vegetal land cover.

Surficial Materials and Soil Parent Materials
Surficial materials are consolidated earth materials which overlie bedrock and are the parent materials in which the surface and buried pedogenic soils have formed. Almost all of the soils in the Upper Mississippi River Basin have formed in parent materials that have been transported from their original source-rock areas. Figure 4.5 shows the distribution of surficial materials in the basin, the maximum extent of Pleistocene continental glaciation, and the extent of the last, late Wisconsinan-age glaciation. The late Wisconsisan glaciated area encompasses more than 50 percent of the basin, including the immature landscapes of the closed drainage or pothole region. In the upper Mississippi and lower Missouri River areas, these surficial materials are mostly comprised of wind-blown loess, glacial till, or glacial meltwater deposits, except for areas covered by weathered-rock and colluvial materials in the western parts of South and North Dakota, southwestern Wisconsin, and the extreme southern part of the basin. In the glaciated part of the area, surficial materials range in thickness from less than 3 feet to more than 400 feet (Soller, 1993); most of the surficial materials are calcareous or weekly calcareous.

FIG_4_5.GIF
The surficial materials of the region have been classified on the basis of inferred infiltration capacity through the surface soil profile and water storage capacity within the surficial material (Soller, 1993). Materials with relatively low infiltration/storage capacities include tills and weathered-rock materials. Clay loam and loamy tills (Gray and others, 1991; Hallberg and others, 1991; Soller, 1993; Whitfield and others, 1993) with thick, clayey soil profiles are present in the southern part of the glaciated area south of the limit of the late Wisconsinan (glaciation). Loamy tills and sandy loamy tills with moderately developed soil profiles underlie areas north of the late Wisconsinan glacial limit. These include areas of sandy loam, and sand and gravel moraine deposits in central Iowa (Hallberg and others, 1991), and large parts of Minnesota and northwestern Wisconsin (Goebel and others, 1983) that have closed drainage with high infiltration rates. The areas underlain by till also include large tracts underlain by surface peat and other permeable wetlands deposits in northern Minnesota. Clay and silt deposits, laid down on the bottoms of glacial lakes, are extensive in North and South Dakota. Clayey residuum and colluvium is present in southwestern Wisconsin (Hallberg and others, 1991; Lineback and others, 1983); silty to sandy clay or silt residual weather-rock materials (Whitfield and others, 1993) underlie areas south and west of the maximum glacial limit. Materials with relatively high infiltration/storage capacities include coarse-grained, stratified glacial meltwater deposits that underlie valleys in the region, and relatively thick (more than 20 feet) wind-blown loess deposits that cover the other surficial materials in the uplands in the southern part of the region.

Distribution of Soils
The distribution of soils in the basin is presented by soil order in figures 4.6 and 4.7. Soil order is the most general class in Soil Taxonomy (Soil Survey Staff, 1975). A map of soil order groups (figure 4.6) was produced by calculating the proportions of soil orders in STATSGO map units, computing a multidimensional distance between the map units, and using a clustering algorithm to group the map units into a set of 11 soil order groups.

orders.gif

Figure 4.7 is a set of histograms showing the distribution of soil orders that make up each soil order group. The statistical clustering technique is used because the soil patterns are complex at this regional scale, but details on the distributions of soil properties are important to hydrologic analysis.

FIG_4_7.GIF

Mollisols are prairie soils that are formed under grasslands that have dark colored surface layers and high base status (high natural fertility). Mollisols cover over 50 percent of the basin and are the dominant soil order in Montana, Colorado, North Dakota, South Dakota, Nebraska, Kansas, Minnesota, Iowa, and Illinois, ranging from a high of 90 percent in North Dakota to a low of 9 percent in Wisconsin.

Alfisols are the dominant soil order in Wisconsin, Indiana, and Missouri. In the more humid eastern part of the basin, Alfisols formed under forest vegetation. In the extreme western part of the basin, Alfisols formed under a mix of deciduous, savanna, short and mid grasses, and forbs vegetation.

Hydric Soils
Hydric soils are important in the study because they indicate the areas of pre-settlement wetlands. Hydric soils, as their name implies, are soils that have formed under wet conditions. These conditions must have been wet enough for the soils to become anaerobic in the upper part (National Technical Committee for Hydric Soils, 1991). Presence of hydric soils is considered an indication that an area may be or was formerly a wetland.

Figure 4.8 is a map of hydric soils as a percentage of the land surface, produced from the STATSGO database. Hydric soils in the basin occur as depressions (potholes) in young, late Wisconsinan-age glacial, closed drainage landscapes of the Dakotas, Minnesota, Iowa, and Illinois; in depressional areas on wide interfluve areas of southern Iowa and Illinois and northern Missouri; and in backland and backswamp areas on floodplains.

hydric.gif

The area of hydric soils as a percentage of the total area of a state ranges from about 1 percent in Montana and Wyoming to over 35 percent of the area in Minnesota. The total area covered by hydric soils is about 26 million hectares (62 million acres) for the 13-state region. The area of hydric soils is considered roughly equivalent to the original area of wetlands before the region was settled.

The highest concentration of hydric soils in the basin is in the floodplains of the Missouri and Mississippi Rivers and in the glacial pothole and depressional (closed drainage) areas of central Iowa, southern Minnesota, the northern half of Indiana, and central and northeastern Illinois. A large portion of the hydric soils in this, the major corn belt region of the basin, represents former wetlands (closed depressions) that have been drained for agricultural use.

The Missouri River has a relatively wide floodplain through Nebraska and Iowa. The pattern of hydric soils in the floodplain is such that the highest percentage of hydric soils is located away from the river and next to the uplands. This pattern conforms to landforms in the floodplain: natural levees next to the stream where sediment is deposited as a result of rapid loss in water velocity, backland areas that generally slope to backland depressions, and occasional oxbow lakes (Strahler, 1969; Ruhe, 1975).

Topography
The basin is characterized by two distinct kinds of landscapes: (1) open systems where the drainage net grades from small streams to a larger trunk streams, and (2) closed systems where the drainage is trapped within a common depository and where surface flow, if it occurs, is mostly in ill-defined drainageways to trunk streams (Ruhe and Walker, 1968). Within these landscapes, hillslope geometry (linearity, convexity, or concavity along the slope length and width) is an important factor affecting water movement. Convergent slopes, which are concave in both slope length and width directions, are areas of maximum accumulation of runoff water. These convergent slopes are potential storage areas of runoff water. In contrast, divergent slopes, which are convex in both directions, are areas of maximum runoff (Pennock and others, 1987).

In open systems, surface water runoff generally flows to a stream and out of the system. Water may accumulate in convergent areas of open landscapes such as head slopes (convergent back slope) and convergent foot and toe slope positions. These wetter areas may support hydric soils if there is a water-restricting layer that crops out on the hill slope.

Closed landscapes are generally related to the areas of glacial drift in the drainage basin. Closed landscapes lack well defined stream outlets: thus, water, sediment, and other materials from the surrounding area are trapped in potholes or other depressions. Trapped or ponded water must either evaporate or recharge the ground water. During large storms, the smaller depressions may fill and any excess water may overflow in undefined surface drainage to other depressions or eventually to a stream. Constructed open ditch drainage systems change closed landscapes so that they function more like open landscapes with respect to both surface and ground water hydrology. Before agricultural drainage, closed landscapes were considered noncontributing with respect to surface water runoff, although they might contribute during storms large enough to cause the depressions to "fill and spill."

SURFACE AND SOIL PROFILE STORAGE OF WATER
In general, the surface water storage capacity is a function of slope and slope variance, that is, the amount and type of terrain variability there is over the average slope at a landscape scale. This can be determined from digital elevation data for specific storage capacities and flow directions. Surface water storage capacities are an attribute of the soils data as is the subsurface storage capacity. As improved digital elevation data and more detailed digital soils data become available more site specific digital analyses will become feasible. Current data are more suited to regional analysis at scales ranging from landscape to larger areas. Estimates based on existing data are quite revealing.

In the nine states comprising the detailed SAST study area, the potential for water storage both above the ground surface and in the soil is broadly estimated at 420 billion cubic meters, using information in the STATSGO soil database (table 4.1). This volume is equivalent to a water depth of 0.26 meters (10 inches) spread uniformly over the 1,631,000 square kilometer (630,000 square miles) area of the 9 states. This depth represents the average depth of surface ponding plus the average available water capacity of the soil, which is the difference between the field capacity and the wilting point of the soil. The volume of available water capacity can be divided by the land area to give an intuitive impression of the amount of water expressed as a depth (cubic meters / square meters = meters).

The combined surface and soil-water storage potential may be compared with accumulated precipitation depths for the June-September 1993 flood period, which ranged from 15 to 30 inches (figure 3.3b) and were about 4 to 6 inches for some individual storm events within that period (Wahl and others, 1993). The accumulated precipitation for the 8-month antecedent period, October 1992 through May 1993, also ranged from about 15 to 30 inches (figure 3.3). The equivalent depth of the water-year 1993 runoff of the Mississippi River at Thebes, Illinois spread over the 9-state area, is about 9.6 inches (table 3.1); the equivalent depth for the June-September 1993 flood season is 4.5 inches.

Although the combined surface and soil water-storage capacity is large, it is effective in reducing peak flood runoff rates only if water enters it at a sufficiently high flow rate at the time the flood peak is occurring. For a given inflow rate, the reduction in the outflow rate equals the rate at which water enters into storage. During intense rains, the rate at which the soil can absorb water may be insufficient to effect a significant reduction in the outflow. Antecedent moisture conditions also are important. If the storage is filled before the peak occurs, water will not be able to enter storage at the critical time, and no reduction in peak flow will be achieved. During June and July 1993, many of the areas that received heavy rainfalls and experienced severe flooding had depleted buffering capacity because the soils and wetlands were wet and unable to store additional water.

Subsoil Water Storage
The ability of a soil to store water varies with texture, density, organic matter content, and initial moisture content of the soil. Figure 4.9 is a map of the available water capacity of soils, produced from the STATSGO data base. The available water capacities by state are given in Table 4.1. Available water capacity is defined as the difference between the water storage at field capacity (saturation followed by gravitational drainage) and at wilting point (plants cannot extract more water). Operationally, this is calculated as the water retention difference using laboratory methods. Total water storage at soil saturation would be somewhat greater since water stored between field capacity and saturation is included.

graphic

Infiltration rates are a particularly important factor in the potential water storage in soils. Most medium textured soils, such as those in the basin, have average infiltration rates of 1.25 to 2.5 cm/hr (0.5 to 1.0 in/hr) (Donahue and others, 1977). Storm rainfall intensities greater than those averages are not uncommon. These intensities will exceed the infiltration rates of many soils and water will run off into streams in open drainage systems or into depressions in closed drainage systems, before the soil becomes saturated. The utility of the subsurface soil water storage capacity for flood reduction, therefore, is often limited by infiltration rates. In closed drainage systems, the detention of water in the depressions offers increased opportunity for infiltration to occur, but this infiltration may not be effective for reducing flood peak flows unless the depression is spilling water to the stream and the infiltration is synchronized with the flood peak.

Infiltration generally increases with an increase in organic matter, stability of soil aggregates, and soil cover. These properties are supported by management practices such as residue management and no-till systems. Structural practices such as terracing trap water on the soil surface and allow for seepage of water into the soil. These management practices increase infiltration and thus increase the ability of the soil to store water from rainfall. Additionally, transpiration of plants and evaporation from the soil surface will lower the soil moisture content. Thus, soils with dense live plant covers have more storage capacity available for water retention than soils with sparse or no live coverage.

Surface Water Storage
In the closed drainage areas of the basin, glacial landscapes with depressions or potholes will pond water. During small storms, these depressions do not fill, and the landscape does not contribute direct flood runoff to streams. During larger storms, the depressions may fill, and surface water may flow from pothole to pothole through an ill- defined drainage network and eventually find an outlet to a stream. However, flow in this ill-defined drainage network is a relatively tortuous process in comparison with open drainage systems, and the water in the depressions is still unavailable for runoff into the stream. For this reason, drainage areas in pothole regions often are designated as indeterminate. The area that contributes to a runoff event generally increases with increasing antecedent moisture conditions and storm volumes as the smaller potholes fill and spill.

Figure 4.10 is a map of surface water ponding, produced from the STATSGO database. The depth of ponding is recorded for each soil series, typically to the nearest half-foot. The depth of ponding was weighted by the area of the soil series within each map unit, to calculate an average depth of ponding for the map unit.

graphic

Surface water storage in the basin is confined to the late-Wisconsinan glacial till area (the prairie pothole region) of North and South Dakota, Minnesota, north central Iowa, and east central Illinois. The rest of the basin consists of open systems with only minor areas that store water on the surface. In the basin, estimates show 10 times more storage capacity in the soil than above ground. Subsoil storage is important in all of the basin, but in open drained areas, subsoil storage is the major component and has relatively high capacities in Iowa, Missouri, Kansas, Montana, Colorado, and Nebraska relative to other states. Relatively high surface water storage is found in the Dakotas, Minnesota, Illinois, Indiana, and Wisconsin.

While these analyses based on STATSGO data are valuable for regional estimations, more detail is necessary for subregional and local planning and decision making. The SCS is conducting and coordinating an effort to obtain soils data at a much finer resolution. These data, the Soil Survey Geographic Data Base (SSURGO), are being collected in map and digital form.

Recommendation 4.2: Accelerate production of soils data of finer resolution than the STATSGO data, e.g., SSURGO data, and include those data as part of the clearinghouse. The SCS should maintain these data for distribution in a generally accepted format that can be easily converted to other generally accepted formats.
 

WETLANDS OF THE UPPER MISSISSIPPI RIVER BASIN
Wetlands in the upland areas of the basin are discussed with respect to changes in wetlands from presettlement to the present, to their value for biodiversity and for water quality, and to their potential use for flood redirection in the basin.

Presettlement vegetation consisted of deciduous hardwood forests in the eastern Ozark Plateau parts of the basin, tall-grass prairies in the central part, and mixed and short-grass prairies in the western part. In the prairie region, woodlands occurred in riparian zones and around upland wetlands.

The amount of presettlement wetlands in the basin is estimated at 58 million acres (Kusler, 1993; Table 4.2). Presently there are about 23 million acres of wetlands remaining in the basin (Kusler, 1993). The loss of 35 million acres of wetlands has mostly been a result of agricultural drainage (Kantrud, 1989; van der Valk, 1989), and channel modification and flood control (Funk and Robinson; 1974, Eckblad, 1988; Jahn and Anderson, 1986).

The wetlands in the northern Great Plains (prairie potholes) and sand hills are critical to breeding, migration, and wintering waterfowl (Tiner, 1984). The prairie pothole region occupies only 10 percent of the total continental waterfowl breeding range, but it produces over 50 percent of the continental waterfowl. Wetlands in the prairie potholes provide cover and nesting to a total of 15 waterfowl species, 57 species of nongame birds, and 39 mammals (Kantrud and others, 1989; van der Valk, 1989). Prairie potholes support a diverse community of invertebrates including 44 mollusk species in North Dakota alone. Drained wetlands (prairie potholes) can be restored to a close semblance of their natural state (Kusler and Kentula, 1990), and typically recover much of their former biotic function within a year of restoration. However, restored wetlands consistently have less biodiversity than natural wetlands within at least the first few years of restoration (LaGrange and Dinsmore, 1989; Delphy and Dinsmore, 1993; Hemesath and Dinsmore, 1993; Galatowitsch, 1993).

Depressions and fringe wetlands are critical to preservation of water quality (Reddy and Patrick, 1975; Gambrell and Patrick, 1978; Johnston and others, 1990; van der Valk and others, 1979; Mausbach and Richardson, 1994). Precipitation, runoff, topography, pedology, and vegetation affect the extent to which a wetland can enhance water quality (Furness, 1983; Wigham and others, 1988). Wetlands where the soil becomes anaerobic can remove excess nitrates from the soil and water through denitrification processes, and wetlands reduce nitrate to nitrogen gas (Gambrell and Patrick, 1978). Gambrell and Patrick (1978) also show that wetlands help degrade agricultural pesticides to environmentally safer compounds. Wetlands also serve as sinks for phosphorous (Mitsch and others, 1977; Johnston and others, 1984). Wetlands serve as natural filters that prevent sediments from entering lakes and streams thus enhancing surface water quality (Kusler and Brooks, 1988).

Role of Wetlands in flood reduction
Pleistocene glaciation significantly modified the landscapes of the upper reaches of the Mississippi and Missouri Rivers. The morphology of these landscapes in the northern prairie region consists of end moraines, stagnation moraines, ground moraines, outwash plains, and lake plains. The upland areas contain numerous glacial depressions of various shapes, sizes, and depths that store runoff. Most wetlands in the prairie pothole region occur within depressions in end moraines and ground moraines. Local relief from hilltop to adjacent lowland may be 15 to 45 meters in end moraines and only a few meters in ground moraine (Winter, 1989).

Closed flow systems are still common in the basin, covering thousands of square miles. These landscapes do not normally contribute to stream flow by runoff, except during storms large enough to make the depressions fill and spill. Therefore, they do not fit the classical definition of "watershed" unless they are artificially drained. Winter (1989) states that, "Where it is not extremely flat, such as in morainal areas, a natural drainage network has not developed, and the many depressions are not connected by an integrated drainage system". More than l,000 square miles of the 1,760 square mile drainage area above the Jamestown, North Dakota, reservoir are still considered to be noncontributing to runoff (due to absence of artificial drainage) (Wiche and others, 1990). Future drainage would convert these closed systems to open systems.

The primary loss of water in these closed landscapes is through evapotranspiration and groundwater seepage. Average annual evaporation amounts in inches vary among reservoirs in the Midwest from north to south and from east to west: Milwaukee (29); Minneapolis (32); Bismarck (39); Havre (43); Kansas City (47) and North Platte (51) (Van der Leeden and others, 1991). Kantrud and Steward (1977) studied the water losses from 135 wetlands of different water regimes over a 6- year period. On the average, wetlands are either dry or their water depths are substantially reduced by the time of the November freeze-up. In an average year, therefore, the depressions are available for runoff storage the following spring.

In contrast, thousands of square miles of depressional areas have been drained in the basin. Wetlands as a percentage of surface area of many states have declined drastically since the initiation of European settlement in the 1780's (table 4.2).

Recommendation 4.3: The U.S. Fish and Wildlife Service should complete the National Wetlands Inventory. Agricultural wetlands should be included. This classification would improve the usefulness of the data set for evaluating areas of wetlands and for local and subregional entities to use in planning.

Investigators have documented the depressional storage in some of the closed flow systems within the northern prairie portion of the Upper Mississippi River Basin. Wiche and others (1990) used digital elevation models in 5 test sites consisting of 26.4 square miles in North Dakota. Total storage capacities were 809.8, 330.5, 326.3, 321.8, and 199.2 acre-feet per square mile. Equivalent depths of stored water were 15.2, 6.2, 6.1, 6.0, and 4.7 inches. Hubbard and Linder (1986) measured the water held in 213 wetlands representing 50 percent of the surface water of the study area in South Dakota. These wetlands held about 162 acre-feet (20 ha-m) of water; the equivalent depth of stored water was 3 inches for the pothole area of the basin. The authors concluded that immense quantities of runoff could be impounded by prairie wetlands, acting to limit flooding and to recharge groundwater supplies. Ludden and others (1983) studied wetland storage capacities in the Devil's Lake Basin of North Dakota. They found that the wetlands in the area store about 72 percent of the total runoff from a 2-year frequency flood and about 41 percent of the total runoff from a 100-year frequency event. The Devils Lake Basin itself is a large closed flow system that does not contribute to the Red River watershed.

Moore and Larson (1979) studied the effects of drainage projects on surface runoff from small depressional watersheds throughout the North central region of the United States. They were interested in the role played by prairie pothole depressional wetlands in regulating high rates of runoff (flood flows) from major summer storms and from spring snowmelt, before and after drainage for agriculture use. In 23 watersheds in Minnesota, they determined that the mean annual flood increases in proportion to watershed area and inversely with the percentage of lakes and wetlands within the watershed. Their data support the contention that artificial drainage increases the watershed runoff area and decreases the amount of depressional storage. On the other hand, Lindsley and Franzini (1972, p 626) note that, although there is no question about the value of watershed storage and land treatment for reduction of soil erosion and preservation of soil moisture, there is debate about their value for flood flow reduction. Also, Miller and Frink (1984) found that the year-to-year variability of flood flows in the Red River of the North basin may have masked any small effects of drainage of potholes of agricultural lands.

Theoretical consideration of the relations between inflows, outflows, and storage changes suggests that the effectiveness of wetlands and soil moisture storage in reducing flood peaks will be greatest for small floods with dry antecedent conditions, and least for floods like 1993 in which all available storage is full before the peak occurs.

Identification of Noncontributing closed landscapes
Noncontributing landscapes in the uplands store a considerable amount of water. Some scientists argue that closing formerly noncontributing landscapes can increase upland water retention with the effect of reducing flood peaks. Wilen (FWS, written commun., 1994) suggests identifying the formerly noncontributing landscapes and calculating the amount of retention that would be obtained by decreasing the artificial drainage in those areas. A rough estimate can be obtained by identifying the closed landscapes, identifying the systems that have been opened by artificial drainage, calculating the storage capacity under both conditions, and calculating the difference. Using GIS and detailed data produced by adding the agricultural wetland category to the NWI data would allow the calculation to be accomplished at the individual wetland scale as well as in aggregate. A refinement of this calculation could be made by incorporating the effect of water storage retention under various storm conditions and using the estimates of runoff conditions caused by agricultural management practices.

CONCLUSIONS
The terrain classification system is crucial to understanding the hydrologic response of the basin. The terrain classes will allow analysis of research results on hydrologic process within areas of similar terrain. This analysis is needed to understand how parts of the basin and the basin as a whole respond to various land use and land treatment activities. Further analysis of the STATSGO database is needed to develop thematic maps that show areas of similar hydrologic response. A multidimensional distance and clustering analysis similar to the slope analysis presented in this report is needed for soil properties related to hydrologic response. These properties include slope, permeability, hydrologic group, texture, and available water capacity. This analysis will support development of the terrain classification system by providing attribute information for the terrain classes.

The effect of management practices that increase infiltration of water into the soil needs further study with respect to hydrologic response of the basin. Questions to be answered are (1) does increased storage of surface water reduce peak flows and stages, and for which types of floods? (2) what is the effect of increased infiltration on groundwater flow and groundwater quality? and (3) for which terrain classes are these practices most effective?

The restoration of wetlands is important for reducing surface water flow in certain landscapes. Further study is needed to identify the terrain classes that are most affected by surface water storage.

Back to Chapter 3

Chapter 5: Floodplain Geomorphology

Back to the Table of Contents