OVERVIEW

We began work on the project on July 1, 1998. Because the original schedule for the project assumed a starting date of September 1, 1996, we have revised the schedule appearing in the project proposal to reflect the actual starting date. The revised schedule is presented as Appendix A of this report.

Work has been simultaneously progressing on several tasks, as outlined in the project schedule, and these various project tasks are on or ahead of schedule. Major project components underway include:

Selection of study watersheds

Delineation of overland flow elements (OFEs)

Selection of field study sample points

Collection of field data

GIS analyses

Summaries of approaches and progress in each of these task areas are presented below. Some of our procedures are briefly described; they will be discussed in full detail in the final report.

SELECTION OF STUDY WATERSHEDS

Location of Study Watersheds

Watersheds being used to develop the monitoring protocol have been selected within four large fifth-order and sixth-order Utah river basins (Figure 1). These four river basins were chosen because they are areas where watershed approaches are being undertaken toward restoration of aquatic resources. Two of the basins, Little Bear River and Otter Creek, have been designated as U.S. Department of Agriculture Hydrologic Unit Areas (HUAs) established as part of the State of Utah's effort to comply with the 1987 Federal Clean Water Act, Section 319. A Coordinated Resource Management Plan has been implemented for the Chalk Creek basin, and Section 319 funds have been used for stream stabilization and water quality incentive programs. The Beaver River has been targeted for EPA Section 319 non-point source (NPS) funding and stakeholders in the basin are beginning development of a CRMP. Because results of this investigation could aid the ongoing watershed protection programs in these basins, we adopted these basins as localities for selecting study watersheds. Additionally, three of these basins, Chalk Creek, Little Bear River, and Otter Creek, have been areas where the investigators have had or have ongoing studies.

Identifying Second-Order Watersheds within the River Basins

The geographic unit delineated for detailed study for this investigation is the second-order watershed. A stream with no tributaries is considered first order. A second-order stream begins at the confluence of two first order streams and ends at the confluence with another second- or higher-order stream. Multiple first order streams may be tributary to a second-order stream segment. A stream network and its stream ordering may be identified using topographic maps, digital elevation models, aerial photography, or field mapping. For this investigation, we have used a GIS approach to extract the drainage network and to define second-order watersheds. Specifically, we have used routines available within the ArcView GIS from Environmental Systems Research Institute (ESRI) to derive land slope, flow direction, contributing area, stream network, stream ordering, and drainage divides from digital elevation model (DEM) data. We defined the heads of first order streams using a constant or threshold support area. Use of this method produced stream networks similar to the blue line networks depicted on 1:24,000 scale USGS topographic maps for the study basins. The "blue line" network on U.S. Geological Survey topographic maps are frequently used as the basis for stream cataloging systems such as the USEPA's Reach Files. Unfortunately, topographic map blue lines are cartographic generalizations that portray the stream network using the cartographer's subjective judgement. Thus, the DEM-derived and blue line networks were not entirely consistent, as would be expected since different criteria were used to generate each one. However, there is a substantial overlap in the second-order watersheds defined by these methods; the exact extent of this overlap will be explored as a later aspect of the project.

Using this GIS-based approach, all second-order watersheds within the four HUA/319 basins were identified (Figure 2). Where available, USGS 30-meter DEM data were used to define the drainage network. Unfortunately, three of the basins, Beaver River, Chalk Creek, and Otter Creek, did not have complete 30-meter DEM data sets available. Therefore, the drainage networks for parts of these basins were extracted from 3-arc-second USGS DEM data sets, which have lesser accuracy. Although statistical information was compiled for second-order watersheds in these areas and incorporated into the overall basin statistical summaries, accuracy of information from these areas is substantially poorer than from areas having 30-meter DEM data coverage. The number of second-order watersheds ranged from 64 in the Little Bear River basin to 188 in the Beaver River basin. The second-order watersheds in these four basins have an average drainage area of about 4 km²; the number of second-order watersheds a basin contains correlates with basin size (Table 1). Second-order watersheds in these four basins range in size from less than 1 km² to 25 km². The larger second-order watersheds typically contain many first-order tributary streams, whereas the smaller watersheds usually contain only two first-order tributaries. Between 57 and 69 percent of the drainage area in the four basins is incorporated into these second-order watersheds. In the remainder of the basin, zero-order (unchannelized hollows) or first-order watersheds drain directly into third- or higher-order streams.
 
 Selecting Second-Order Watersheds for Detailed Study

For the purposes of developing the monitoring protocol, we are specifically selecting second-order watersheds on the basis of land use conditions and data availability. Within each basin, we are attempting to select one watershed that is relatively undisturbed and another watershed with a greater disturbance level. In each of the river basins, a different type of dominant disturbance has been selected:

Beaver River basin: Fire

Chalk Creek basin: Grazing

Little Bear River basin: Roads

Otter Creek basin: Vegetation management

The selection and "pairing" of study watersheds based upon disparate disturbance levels is intended to accentuate the dissimilarities in linkages among watershed sub-components subjected to differing levels of disturbance. However, we will not know until after completion of field studies and data analyses how significant the differences in disturbance levels between watersheds actually are. This watershed pairing based upon disturbance dissimilarity is a space-for-time substitution intended in part to allow development of the protocol within the two-year investigation period.

Additional criteria are also being used to select the study watersheds. Watersheds in close proximity to one another were sought to reduce the potential affects of regional variations in climate, vegetation, geology, and soils. Significant variations in these four parameters could add confounding complexities that mask the effects of the disturbances. Road accessibility to the watersheds and landowner permission for access are also essential for conducting field components of the study. Finally, because of the short time frame of the study, we have attempted to select watersheds with a maximum number of available data of the following types:

30-meter DEM,

Soil mapping,

Climate data for WEPP modeling, and

Streamflow data on a third- or fourth-order stream downstream from the study watersheds.

We have essentially completed selection of the study watersheds, and the project's PIs have made site visits to all eight watersheds. We are in the process of finalizing access arrangements for several of the watersheds.

DELINEATION OF OVERLAND FLOW ELEMENTS

We are currently in the process of subdividing each watershed into hillslopes and overland flow elements (OFEs). Hillslopes are areas of unchannelized overland flow that contribute water and sediment to a watershed's stream network. Our primary tool for defining a watershed's hillslopes and drainage network has been GIS analysis of DEM data. The top of a hillslope represents a drainage divide, and the bottom of a hillslope drains into the stream network. A hillslope is subdivided into one or more overland flow elements that represent areas of fairly homogeneous soils, vegetation, and topography. OFEs allow representation of hillslope non-uniformity, which is essential to depicting hillslope runoff and sediment production. Division of hillslopes into OFEs is a somewhat arbitrary process, as one is attempting to subdivide a surface into discrete pieces based upon continuously varying parameters. To define OFEs, we are using soils mapping, land slope derived from DEM data, and vegetation cover derived from aerial photography.

Delineation of hillslopes and OFEs has been completed for the two watersheds in the Little Bear River basin. Little Bear River basin study watershed no. 1 (LBR1) will be used throughout the remainder of this document to illustrate some of the methods we are employing in the investigation. Figure 3 illustrates the hillslopes and OFEs for the LBR1 watershed. We are currently at various stages of collecting data and creating GIS coverages to delineate hillslopes and OFEs in the other six watersheds, a task that will be completed in the second quarter of the project.

SELECTION OF FIELD STUDY SAMPLE POINTS

To develop the monitoring protocol for assessing linkages among upland, riparian, and stream conditions, we are in the process of collecting data at sample points in upland areas and for the stream channels and riparian zone comprising the drainage network of each watershed. A brief synopsis of the procedures used for selecting these points is presented here.

Upland Data Collection Points

Sample points for collecting data in upland areas are being selected using the following steps:

1.  OFEs are grouped into several classes in each watershed. Classes are being derived using cluster analysis statistical procedures. These classes are created based upon similarity of OFE slope, vegetation, and soils. Soil, vegetation, and slope parameters for each OFE are derived by overlaying the OFE coverage and parameter coverage to obtain an average value. For the two watersheds for which this step has been completed, three OFE classes were derived.

2.  For each OFE class, three individual OFEs are randomly selected using an area-weighted random number generation strategy. This process has resulted in sample points being located in nine OFEs in each of the two watersheds completed thus far.

3.  Within each selected OFE, five points are chosen using a spatial random grid point generator that is available within the ArcView GIS. These points can be printed out overlaying an aerial photograph so they may be located in the field. Additionally, a table of geographic coordinates is generated which can be used in conjunction with a global positioning system to locate the sample point in the field. Figure 4 illustrates 45 upland data collection points within the LBR1 watershed, and the OFEs from which they were selected.

Stream Network Data Collection Points

Collection of stream and riparian zone data, which has begun in the Little Bear River watersheds, is being undertaken at several site types:

At locations where the sampled OFEs drain into the channel network.

At three random locations along the network in each watershed. Random locations along the network are being selected using ArcView to generate spatially random points, similar to the procedure for selecting points in upland areas.

At the watershed outlet.

At other selected locations chosen to evaluate specific channel/riparian conditions.

FIELD DATA COLLECTION

Phase I field data collection has just begun and will continue until the middle of 1999. We are beginning field data collection in the Little Bear River watersheds and will work toward the southern watersheds during the winter months. Further information on the procedures being used and status of field data collection will be provided in upcoming quarterly reports.

GIS ANALYSES

Because the core of our monitoring approach entails the collection and interpretation of spatially distributed data, we are utilizing a GIS-based approach to aid us in much of our effort. As one example that we have presented above, data collection points have been randomly selected using an internal GIS random grid point generator.

GIS development of various data components is progressing much more rapidly than anticipated. We have acquired most of the existing spatial data necessary for project purposes, including soils, vegetation, and DEM data. For data not already in computer format (soils, for example), we are in the process of entering such data. Figure 5 illustrates some of the GIS data that we are using or have generated for the LBR1 watershed.

---

  Back to Previous Page