Summary

Hydrologic responses to greater than average precipitation in the Johnson Creek basin in water years (WY) 1996-1997 resulted in the flooding of Johnson Creek, elevated groundwater levels, and, over the following several years, increased groundwater discharge to springs. Although the precipitation during distinct storm events resulted in a rise and fall of streamflow on the scale of hours to days, the back-to-back high annual precipitation in WY 1996-1997 led to a response of the groundwater system that lasted for months to years. This study was prompted by inquiry into the causes of floods from increased spring discharge and from an elevated water table, and by a need to better understand spatial and temporal trends in annual, low, and high streamflow of Johnson Creek. The U.S. Geological Survey began the cooperative study in 1997 with the city of Portland. In 2000 the cities of Gresham and Milwaukie and Clackamas County Water Environment Services joined the cooperative study, followed in 2002 by Multnomah County.

New information was collected for this study, and was analyzed in context with previous studies as well as long-term hydrologic data. New data consisted of a network of groundwater level measurements (including continuous water-level records), numerous streamflow measurements, and continuous records of streamflow and stream temperature.

The average annual total precipitation from WY 1911 to 2006 was about 37 in./yr. Precipitation WY 1996 was 54.7 in., followed by a record 58.7 in. in WY 1997. The 3-year moving average annual precipitation in WY 1997, WY 1998, and WY 1999 was the greatest since recordkeeping began. The large inflow (recharge) led to a large outflow (discharge) from the groundwater system to surrounding streams.

Results from water-table mapping and from a regional groundwater flow model of the Portland basin from previous USGS studies indicate that groundwater and surface-water flow divides are not necessarily coincident for the Johnson Creek basin, implying that in some areas of the Johnson Creek basin water that recharges the groundwater system flows out of the basin and discharges to major rivers such as the Sandy, Columbia, or Willamette Rivers. The greater relative discharge observed in the lower reaches of the Johnson Creek basin compared to the upper and middle reaches may be caused by groundwater flow out of the basin in the upper and middle reaches resulting in relatively low base flows in Johnson Creek compared to what might be expected from precipitation amounts in the basin.

Groundwater levels in wells open to the surficial aquifer in the Johnson Creek basin fluctuate generally in a sinusoidal annual pattern in which water levels rise in response to autumn and winter rains and decline during the relatively dry and warm spring and summer seasons. Toward the west, water-level fluctuations are damped in magnitude and delayed in time relative to precipitation. In the western area of the basin, water levels in the Westmoreland Park area wells, which are open to a former alluvial channel of either the Clackamas or Willamette Rivers, or both, respond quickly to precipitation and to the levels of the Willamette River and Johnson Creek.

Flooding of Crystal Springs Creek, first recorded in 1997, was caused by elevated groundwater levels and associated increase in groundwater discharge to the springs and the creek. Comparisons made between streamflow of Johnson Creek at Milwaukie and streamflow of Crystal Springs Creek indicated that the high flows of Crystal Springs Creek began in 1996. Comparison of streamflow and precipitation (annual and 3-year moving average) indicated that the relatively high flows of Crystal Springs Creek from 1996 to 1999 probably had not occurred since the channel modifications were made in 1936 during construction of Westmoreland Park, and possibly not since precipitation records began in 1911. A relation was developed between the groundwater level in a well near the springs and streamflow of Crystal Springs Creek in the Reed College area and at the mouth of the creek, enabling prediction of high and low streamflow of Crystal Springs Creek based on groundwater level.

The appearance of Holgate Lake and occasional flooding in the surrounding area was the result of the shallow water table. Flooding was related to high precipitation years. Groundwater monitoring that began in 1998 and lake-level monitoring that began in 1999 indicated that the lake levels were similar in height to groundwater levels in a nearby well and followed the same pattern of rise and fall. A relation was developed providing some capability to anticipate flooding from Holgate Lake based on early season (autumn to winter) precipitation and antecedent groundwater levels.

Annual, low, and high streamflows were analyzed spatially and temporally using daily mean and annual peak flows of Johnson Creek, and where applicable, of a nearby basin. On an annual basis, streamflow varies across the basin according to several factors. Precipitation is greater in the higher elevation area of the basin. Depending on the direction of groundwater flow and variations in groundwater flow direction at depth, recharge within the surface-water contributing area of the Johnson Creek basin may discharge to Johnson Creek or may discharge to rivers outside of the basin. Some runoff is intercepted by the combined sewer system and is routed out of the basin. Runoff per unit area was greatest in the upper basin, where the average (from WY 1999 to 2006) was 25.3 in., compared to 18.5 and 12.1 in. in the middle and lower basin, respectively. Similarly, runoff as a fraction of precipitation also decreased, where 45, 37, and 28 percent of precipitation resulted in runoff from the upper, middle, and lower areas of the basin, respectively. About half of the increase in annual mean streamflow from the middle to the lower area of the basin during this period was from Crystal Springs Creek.

Low flows varied from year to year, primarily caused by fluctuations in groundwater discharge to Johnson Creek, although low flows were apparently affected by an unidentified water use between 1955 and 1977 as well. Seepage measurements indicated minimal gains in streamflow upstream of RM 5.5 and increased gains toward the mouth of Johnson Creek. Streamflow records indicated minimal increases in low flow between the Gresham and Sycamore sites from 1998 to 2006. Most of the increase in low flows between the Sycamore and Milwaukie sites is attributable to Crystal Springs Creek. Low flows may be exacerbated by minimal groundwater discharge to the stream in the upper and middle areas of the basin, where Johnson Creek captures little groundwater flow. In these areas, groundwater flow leaving the drainage basin is inferred from water-table mapping and previous modeling results. A comparison of low flows of Johnson Creek at Sycamore from 1941 to 2006 to low flows of the Little Sandy River indicated that water withdrawals from Johnson Creek during the low-flow period from 1955 to 1977 may have been the cause of a 50 percent decrease in summer flows during this period.

High flows result from runoff from the land surface and depend on precipitation volume and intensity. Although the magnitude of high flows masks groundwater discharge to the stream during rainy periods, interception by the combined sewer system, UIC systems, and a permeable landscape moderates high flows in the lower area of the Johnson Creek basin. The difference in high flow of Johnson Creek at the Sycamore and Milwaukie sites, in terms of the magnitude of peak streamflow and volume of storm runoff, is relatively small considering the increase in drainage basin size.

Temporal trends in flow typically associated with increasing urban development generally were absent in Johnson Creek. Annual flow as a fraction of annual precipitation, low flow (with the exception of unusually low flows from 1955 to 1977), annual peak flows, and the percentage of days in a given year when the daily mean streamflow is greater than the annual mean streamflow, exhibited no trend from 1941 to 2006. The absence of a temporal trend does not mean that the hydrology of the Johnson Creek basin has been unaffected by human occupation because much of the agricultural, residential, and urban development including the extensive channel-modification of the lower 11 mi of Johnson Creek predated data collection in the basin. Because of the management of stormwater in urban areas, most urbanization additionally was not associated with simple piping of stormwater to the creek. Drywells and other UIC systems have been in use for decades, taking advantage of the infiltration capacity of permeable subsurface deposits in the urban areas. Continued development of innovative onsite methods of stormwater disposal, particularly in upland areas, may lead to decreased peak flows and enhanced recharge to the groundwater system.

Summer temperature of streams in the Johnson Creek basin is affected by the cooling effects of groundwater discharge and by warming from ponded and open-water areas. Although the source of water for Crystal Springs Creek is spring flow, warming due to instream constructed ponds results in a temperature in Crystal Springs Creek which is greater than that in Johnson Creek. Because Crystal Springs Creek accounts for most of the flow in Johnson Creek downstream from its confluence, it warms Johnson Creek during summertime.

High flow of Johnson Creek may entrain sediment and sediment-borne contaminants. Several factors indicate that sediment yield of the basin may be sensitive to conditions in the upper area of the basin. These factors include generally higher precipitation, greater slopes, a dense network of roads and ditches associated with agricultural and rural residential land uses, and a relative abundance of sources of sediment and sediment-borne contaminants compared to the relatively low relief, urban areas of the lower basin.

Continued study of the Johnson Creek basin will provide insight into the response of the hydrology to land-use changes. In particular, runoff from newly acquired areas for urban development in the upper area of the basin may be particularly sensitive to changes in land use.