[Next Section] [Previous Section] [Table of Contents]

Background

Although each basin has specific climate, vegetation, geomorphic and population characteristics, there are some similarities throughout the region.

All basins in this study are in the Puget Sound-Willamette Lowland, west of the Cascade Range in Oregon and Washington. The Tualatin and Luckiamute Rivers drain the Coast Range and the west side of the Willamette River Basin. Johnson Creek drains a small portion of the east Willamette Basin through the eastern and southern suburbs of Portland. The Portland metropolitan area is currently one of the fastest growing regions in the United States, and further development in the area is expected (Wilson, 1997). The Newaukum River, south of Puget Sound, drains the lower flanks of the Cascade Range, and then flows into the Chehalis River.

Climate in the study area is maritime temperate. Winters are wet and cloudy, but summers are dry and clear. About 80 percent of annual precipitation falls from October through March (Figure 4). The area above 400 m and below 1,200 m is the transient snow zone, where temperature is the determinant of precipitation type (Harr, 1986). Snowfall, even at the highest altitudes of the study basins, is sporadic through the winter and is not concentrated in season-long snowpacks (Evans and Fibich, 1987; Green, 1982; Green, 1984, Knezevich, 1982b). Precipitation generally increases with altitude. For example, in the Tualatin Basin, Forest Grove at 67 m above sea level averages 1,143 mm of precipitation a year, while Timber at an elevation of 297 m averages 1,549 mm (Hart and Newcomb, 1965) (Table 2).

This climate type is also called Mediterranean, and results in rivers supplied primarily by rainfall rather than snowmelt (Shaw, 1988). This produces the typical Pacific Northwest streamflow pattern of highest flows during the winter and lowest flows during the summer.

In addition, the Oregon State Climatologist describes the area climate as cyclical, with approximately 20 year cycles of alternating wet and dry years (Taylor, 1998). The region was affected by drought through the late 1980s and early 1990s (Figure 5).

Table 2. Description of precipitation data, records from water years 1949-96 (Evans and Fibich, 1987; Gerig, 1985; Green, 1982; Green, 1984; Knezevich, 1982a; Knezevich, 1982b, WRCC, 1997).

Vegetation

Before European settlement and development, western Oregon and Washington were largely covered with Douglas-fir forests, typically with little overland flow because of the density of vegetation and hydraulic conductivity of the soil (Fredriksen and Harr, 1979). Other upland vegetation includes western hemlock (Tsuga heterophylla), red alder (Alnus rubra), western red cedar (Thuja plicata), western hazel, noble fir (Abies procera), vine maple (Acer circinatum), rhododendron (Rhododendron), and ferns. At lower elevations, native vegetation includes Douglas-fir, Oregon-grape (Mahonia aquifolia), rose (Rosaceae), black cottonwood (Populus trichocarpa), willows (Salix sessilifolia), ash (Fraxinus latifola), and blackberry (Evans and Fibich, 1987; Gerig, 1985; Green, 1984; Green, 1982, Knezevich, 1982a, Knezevich, 1982b).

Several researchers have found that clearcutting Pacific Northwest forested watersheds may result in changes in peak flows under some circumstances (Cheng, 1989; Fredriksen and Harr, 1979; Harr, 1986; Jones and Grant, 1996; Thomas and Megahan, 1998; Ziemer, 1981).

The increase of peak flows found by these researchers is not consistent. Jones and Grant (1996) found increases of up to 50% for small basins and up to 100% in large basins. Looking at the same data, but using a different methodology, Thomas and Megahan (1998) found increases of up to 90% for the smallest peak events in clearcut small basins, but no conclusive increases in the large basins. Cheng (1989) found 21% increases in peak flows after 30% of a small basin was clearcut. Ziemer (1981) found that when the ground was saturated, logging had no effect on large peak flows in a small basin.

Each of the four basins in this study was logged to various degrees through the period of flow records. The effect of logging on streamflows is beyond the scope of this study, and cannot be directly inferred from previous studies.

Ziemer and Lisle (1997) cautioned that observed changes in peak flows in small basins cannot be extrapolated to large basins. The response to precipitation in small basins (<100 km²) is primarily governed by slope, while the response of large basins is largely regulated by the channel network. Increases in peak flows from individual small tributary streams are likely to be attenuated by storage capacities of large basins. Unless the peak flows from tributaries are synchronized, large basins can absorb the increased flows from precipitation events. However, as the lag time between precipitation and tributary peak flows shortens, the possibility of synchronicity increases (Figure 6).

Figure 7 shows how the four basins in this study reacted to storms during three weeks in the fall of 1995. These hydrographs are typical of how these basins react to precipitation input. The Tualatin had a single peak, while the other streams each had at least two peaks. The Tualatin peak occurred after the peaks of the other three streams, and was lower than the highest peak of each of the other basins (streamflow is adjusted for basin size). The highest peak in Johnson Creek occurred earlier than in the other three basins.

Each of the basins is gaged by the USGS (Table 3). Data from three of the gages are available from 1942 or 1943. Data for the Tualatin gage are available from 1929. Data from these gages are in the "excellent" category, indicating that the reported values are within 5% of their actual values (Wellman and others, 1993; Williams and others, 1984). Information in this study refers only to the portion of the basin upstream from the gage listed in Table 3, unless otherwise noted. All peak flows have been log transformed for this study. Peak events for each basin are plotted in Figure 8. Peaks are plotted against volumes in Figure 9. All streamflow data are in Appendix A.

Table 3. Basin gage information: gage site, gage number, size of gaged basin, elevation of basin, average basin slope, period of record, one year recurrence interval, average discharge (USGS, 1998a, b, c, d).

The Tualatin and Newaukum basins have diversions upstream from the gaging stations used in this study. A canal was completed in 1871 to divert water from the Tualatin River to Lake Oswego. The canal carries 1.66 x10^-3 m³/s/km² or approximately 7% of the total streamflow of the Tualatin, and has been consistent through the period of flow record, so should not affect flow patterns. In 1975, a dam was completed on Scoggins Creek, an upland tributary of the Tualatin River. Average streamflow on Scoggins Creek below the dam is approximately 4.2x10^-6 m³/s/km² (USGS, 1998d), affecting less than 0.01% of the average annual flow of the Tualatin. Approximately 9x10^-7 m³/s/km² is diverted from the Newaukum River, or less than 0.01% of the average annual flow, for municipal use (USGS, 1998c). All four of the streams in this study have unquantified diversions for agriculture during summer months. Streamflows from these months, June through September, were not a part of this study.

Population

Population data have been used by other researchers to approximate urbanization (Changnon and others, 1996, Ferguson and Suckling, 1990, Lazaro, 1976, Stankowski, 1972). For this study, decadal population estimates were made for all four basins (Center for Population Research and Census, personal communication; Office of Financial Management, 1997; Office of Financial Management, 1990; US Census Bureau, 1998; Woodward-Clyde, 1995).

Both the Tualatin and Johnson Creek basins have had more than a 1000% increase in population since 1940, though their absolute populations are the largest and smallest, respectively, in the study (Figure 10).

Population density is a more important measure of urbanization than raw population figures. Although Johnson Creek has the smallest population, it has the highest population density (Figure 11).

The Tualatin Basin is bounded to the north, south and east by Columbia River basalt, flood basalts erupted in the Miocene and uplifted, and to the west by the Coast Range, consisting largely of marine sediments and volcanics (Schlicker and Deacon, 1967). The deepest part of the Tualatin basin, near Hillsboro, is filled with as much as 450 meters of silt (Hart and Newcomb, 1965). Through the 72 km reach across the valley floor filled with these fine sediments, the channel drops only 6 m, creating a slow, sluggishly moving stream (Schlicker and Deacon, 1967).

Table 4 summarizes soil characteristics of the basin. The Coast Range mountainous areas are generally steep with well-drained soils. Much of the floodplain and terrace areas are natural wetlands that have been drained for farming or urban development. (Gerig, 1985; Green, 1982). Even with controls on wetland destruction, 10% of the wetlands in the Tualatin basin were destroyed between 1975 and 1988 (Division of State Lands, 1989).

The upland areas remain largely covered with Douglas-fir; elevations below 60 m are mostly agricultural or urban (Green, 1982). Extensive agricultural land is found mostly in the southern portion of the basin.

Farming areas include rolling hills, with well drained to moderately drained soils derived from colluvium; former terraces with well drained to moderately drained soils; and flood plains with well drained to poorly drained soils. Urban areas include hills, terraces, and flood plains.

Extensive urbanization is found generally in the eastern section of the basin, on old terraces and upland areas. Approximately half of the Tualatin Basin is in the Portland metropolitan Urban Growth Boundary, which encourages medium to high density development (Metro, 1996).

The EPA has compiled an index of watershed indicators for the Tualatin basin. On a scale of 1 (best water quality) to 6, the Tualatin rated a 4. Among the indicators of concern were a moderate level of wetland loss (38% from 1780 to 1980), and having at least 7% of land area covered with more than 25% impervious surfaces (EPA 1997).

Both Washington and Clackamas Counties, which cover most of the Tualatin watershed, have zoning restrictions against building in the 100 year flood plain. Both have provisions to protect existing wetlands from development; but under some circumstances allow transfer of building rights from wetland property so that nonwetland parcels can be developed more densely than originally zoned (Washington and Clackamas Co. Planning Depts., personal communication).

The Luckiamute Basin is bounded to the north, south, and east by uplifted Columbia River basalt, and to the west by the Coast Range, which consists largely of marine sediments and volcanics.

Table 5 summarizes soil characteristics of the basin. Upland areas are generally steep with well-drained soils. Farming areas include rolling hills, with well drained to moderately drained soils derived from colluvium; former terraces with well drained to moderately drained soils; and flood plains with well drained to poorly drained soils (Knezevich, 1982a, Knezevich, 1982b).

The upland areas remain largely covered with Douglas-fir. Terraces and lowlands are generally used for agriculture (Knezevich, 1982b). There are no true urban areas in the basin; the largest incorporated city is Falls City, with a 1996 population of 993 (US Census Bureau, 1998).

The Newaukum basin is bounded to the east by the uplands of the andesitic Cascade Range, and to the north and south by low foothills of the Cascade Range. To the west the basin is bounded by low foothills of the Coast Range, consisting of marine volcanics and sediments. The basin has been significantly affected by glacial activity. Table 6 summarizes other basin characteristics.

The uplands of the eastern headwater area have long, steep slopes and roughly parallel drainages. The lower slopes of the Cascades have slopes up to 90 percent and soil predominantly formed from basalt and andesite colluvium. High terraces and hillsides are covered with soil formed by highly weathered glacial drift deposits.

About 40% of the basin is high terraces and ridges; another 20% is floodplain and low terraces. Glacial outwash and reworking by streams has created broad, shallow slope areas with uneven conductivity and drainage characteristics. Perched water tables are common, and brief periods of annual flooding are typical of unmodified land (Evans and Fibich, 1987).

The broad lowlands of the valley are part of the Puget Trough and are surrounded by highly dissected hills with rounded ridges. (Dethier, 1988) The lowlands are typically deposited alluvium and glacial outwash, with soils less than 10,000 years old. This area has undergone multiple glaciations over the past 2 million years, and the drainage network is still adjusting to the basin changes created by glacial advance, cover, and retreat (Booth, 1988).

The upland areas remain largely covered with Douglas-fir. Elevations below 125 m are mostly agricultural or urban. Soils in the flood plains and lower terraces are typically subject to seasonal flooding, and much of the original drainage has been disrupted because of extensive tiling, which is needed for agriculture and urban activities (Evans and Fibich, 1987).

The urban area currently is tightly clustered on the broad flood plain of the river, but development in formerly agricultural areas is increasing (Lewis County Planning Department, personal communication).

The Johnson Creek watershed is bounded by the Boring Hills to the south and east, terraces of the Columbia River to north and the Willamette River to the west (Hogenson and Foxworthy, 1965). Table 7 summarizes soil and precipitation characteristics of this region of the basin.

The watershed includes two distinct provinces (Figure 12). The northern province has very permeable soils, derived from alluvium. Most precipitation on these soils infiltrates into the ground in nonurbanized areas. However, because of the proximity of this area to Portland and Gresham, and the relative flatness of the area, the remaining agricultural lands are being turned into urban areas. All of this land is within the urban growth boundary of cities. This province contributes less than one-quarter of the stream gage readings used in this study (Gude, 1994).

The southern province is characterized by impermeable soils formed from weathered basalt and andesite on moderate to steep slopes. Most of the precipitation that falls on this area contributes to Johnson Creek in the form of overland flow (Gude, 1994).

Figure 12. The northern portion of the Johnson Creek basin contributes little to overland flow because of permeable soils. The southern portion of the basin includes steep slopes (100 foot contour intervals) and weathered volcanic soils with low conductivity (Gude, 1994; Metro, 1996).

In 1994, residential, commercial, and industrial uses covered 46% of the gaged area of the basin. Based on current plans, 78% of the watershed will eventually be covered by these uses.

Clement (1984) found the hydrology of Johnson Creek had been changed by the urbanization in the upper basin, resulting in peak flow rates 30% higher in the 1980s than in the 1940s. A report for the Johnson Creek management program (Woodward-Clyde Consultants, 1995) estimated that current peak flows are 30% - 50% higher than in the basin’s natural state (Table 8). (The Woodward-Clyde (1995) report includes a comprehensive bibliography of previous research efforts in the Johnson Creek basin.) A peak flow rate (Q_p) of 1.92 m³/s before development would have been expected to recur every 100 years. With current development, the same flow rate is expected to recur every 25 years. As development continues, that flow rate will be expected to recur more often.

A tool now available for population analysis is a geographic information system (GIS). Information from Metro (1996) shows the pattern of development in the Johnson Creek watershed. Figure 13 is a graphic representation of the increasing building density in the basin, showing the tax lots in the basin with structures. In the northern portion of the basin, rain has traditionally infiltrated groundwater rather than contributing to the stream by overland flow. This is the area of highest growth, leading to ground cover conditions that disrupt the basin’s natural infiltration capacity.

1930

1950

1970

1996

Figure 13. General development pattern in the Johnson Creek basin. Each polygon is a tax lot with a building on it, based on 1998 assessor data. The X marks the position of the gage used in this study (Metro, 1996).

[Next Section] [Previous Section] [Table of Contents]