Model Boundary Conditions

The boundaries for the Willamette River modeling project are Bonneville Dam (RM 145) and Beaver Army Terminal (RM 54) on the Columbia River and river mile 35 (near Canby Ferry) on the Willamette River. Figure 13 shows the location of the boundaries and some large scale basins in the region.

Figure13.  Model Boundaries on the Columbia and Willamette Rivers

Columbia River

Columbia River Water Level at Beaver Army Terminal

The Beaver Army Terminal station (USGS: 14246900) has been recording continuous water level data from October 1991 to present.In 1997, 1998 and 1999 there were several gaps in the data.To fill those gaps a correlation was developed between the Beaver Army Terminal site, the Vancouver, WA site (USGS: 14144700), the Longview, WA site (USACOE: LOP), the site below the Bonneville Dam (USGS: 14128870) and the tidal influences on the Columbia River.The correlation used (R2=0.8301) is:

Where:

Hourly is the tidal influence from 12.4 hour tidal cycle as:

Daily is the daily tidal cycle estimated as:

Monthly is the monthly tidal cycle estimated as: 

Annually is the influence of any annual tidal fluctuations as: 

The water level elevation at Beaver Army Terminal for the summers modeled is shown in Figure 14.Although the coefficient of determination was 0.83 the correlated water level information fit well with the existing data for the station.

Figure14.  Columbia River Water Level Elevation at Beaver Army Terminal, m NGVD, RM 53.8

Columbia River Flow at Bonneville Dam

Flows at Bonneville Dam were obtained from the USACOE from January 1993 to 2000. A plot showing flow data below the Bonneville Dam for all modeled summers is shown in Figure 15.

Figure15.Columbia River Flow measured below the Bonneville Dam, m3/s, RM 144.5


Water Quality

Water Quality monitoring in the Columbia River is predominately conducted by DEQ. However, USGS has monitored temperature and some water quality parameters such as conductivity, dissolved oxygen, and pH hourly. Continuous temperature data from 1997 to 2000 were obtained from a dissolved gas study carried out by USGS and the USACOE.DEQ (STORET) data is collected at a frequency of monthly to a few times a year.

Temperature data in the Columbia River at Bonneville dam and at Beaver Army Terminal were taken from USGS monitoring and STORET grab samples.Water quality data in the Columbia River at the Bonneville Dam were obtained from grab samples taken by DEQ (STORET) at Warrendale, Oregon because data at the Bonneville Dam were not available.Water quality input files for the model at the Beaver Army Terminal were generated by combining DEQ (STORET) data and continuous USGS data. Figure 16 and Figure 17 show temperature in the Columbia River at Beaver Army Terminal (RM 53.8) and at Bonneville Dam (RM 144.5) respectively. Table 3 shows a list of water quality parameters available for the Columbia River at Bonneville Dam and Beaver Army Terminal.

Figure 18 through Figure 20 shows plots of the water quality constituents used for the downstream boundary condition at Beaver Army Terminal.Figure 21 through Figure 23 show similar water quality plots for the Columbia River at the Bonneville Dam.The procedure used for developing the water quality files from data can be found in Appendix J:Water Quality file development procedures.

Figure16.Columbia River at Beaver Army Terminal (RM 53.8) water temperature, oC


 
 

Figure17.Columbia River at Bonneville Dam (RM 144.5) water temperature, oC



 
 

Parameters
Columbia River at Bonneville Dam
Columbia River at Beaver Army Terminal
ALKALINAS CACO3
X
X
Alkalinity
X
X
Calcium Hardness
X
X
CHLRPHYLA
X
X
Conductivity, mS/cm
X
X
D ORG CC
X
X
Dissolved Oxygen
X
X
D.O Saturation
X
X
ENTCOCCI
X
X
Fecal Coliforms ./100 mL
X
X
NH3+NH4-N DISS
X
X
NH3+NH4-N TOTAL
X
X
NO2&NO3N-DISS
X
X
NO2&NO3N-TOTAL
X
X
NO2-NDISS
X
X
NO2-NTOTAL
X
X
PH
X
X
PHOS-DIS
X
PHOS-DISORTHO
X
X
PHOSPHORSED,SUSP
X
X
PHOS-TORTHO
X
X
PHOS-TOT
X
X
S ORG CC
X
X
SUSP SEDCONC
X
X
SUSP SEDDISCHARG
X
SUSP SEDPARTSIZE
X
X
T ALKCACO3
X
X
Temperature
X
X
Turbidity
X
X
UN-IONZDNH3-N
X
X
UN-IONZDNH3-NH3
X
X
Table 3.Water Quality parameters available for the Columbia River Boundary Conditions
 
 


Figure 18.Columbia River at Beaver Army Terminal Boundary Condition, RM 54
 


Figure 19.Columbia River at Beaver Army Terminal Boundary Condition, RM 54 (Part 2)
 


Figure 20.Columbia River at Beaver Army Terminal Boundary Condition, RM 54 (Part 3)
 
 


Figure 21.Columbia River at Bonneville Dam Boundary Condition, RM 145
 
 
 


Figure 22.Columbia River at Bonneville Dam Boundary Condition, RM 145 (Part 2)
 
 
 


Figure 23.Columbia River at Bonneville Dam Boundary Condition, RM 145 (Part 3)

Willamette River

Flow data

There is currently no gage station on the Willamette River at Canby Ferry so flow routing models were investigated to develop the flow at Canby Ferry to serve as the upstream boundary condition.USGS developed a flow routing model to estimate daily flows in the Willamette River at Portland. This approach is based on measured daily flows upstream of the Portland gage station (USGS: 14211720). The equation used to estimate flows is:
Mean daily Q at Willamette River at Salem with 1 day lag
+(Mean daily Q at South Yamhill) x 2 with 1 day lag
+(Mean daily Q at Pudding near Woodburn) x 2 with 1 day lag
+(Mean daily Q at Tualatin River-West Linn ) x 1.5
+(Mean daily Q at Clackamas River-Estacada) x 1.5
+Mean daily Q at Johnson Creek-Milwaukie
=Mean daily Q at Willamette River-Portland

The most upstream sites are lagged by one day to account for travel time and some of the lower sites are increased by a factor to account for ungaged flows. The flows from these stations were added together to obtain daily flows in the Willamette River at the Portland gage station (USGS: 14211720).

The U.S Army Corps of Engineers also developed the Lower Columbia River UNET Model (Knutson, 2000), a routing method to estimate daily flows in the Willamette River at Portland. The flow in the Willamette River at Portland was obtained using the following equation:

 
Mean daily Q at Molalla River-Canby
+Mean daily Q at South Yamhill-Whiteson
+Mean daily Q at Pudding-Aurora
+Mean daily Q at Tualatin River-West Linn
+Mean daily Q at Clackamas River-Estacada
+Mean daily Q at W.R at Salem
+Mean daily Q at Johnson Creek-Sycamore
+Mean daily Q Ungaged between Salem and Oregon City
=Mean daily Q at Willamette River-Portland

In order to utilize this flow routing equation for a larger period of time the USACOE developed several flow correlations between stations in the Willamette basin to fill gaps found in the data. These correlations are shown in Table 4.


 

Correlation Number
Correlation
Time Period
1
Molalla River at Canby = 0.48 * Clackamas at Estacada
prior 1929, 1959-1964, 1978-to present
2
Pudding River at Aurora = 0.60 * Yamhill at Whiteson
prior 1929, 1964-1993
3
South Yamhill at Whiteson = 1.7 * Pudding River at Aurora
prior 1940, 1991-1994
4
Johnson Creek at Sycamore = 0.024 * Clackamas at Estacada
prior 1940
5
Ungaged Flow = (ungaged drainage area/Aurora Creek drainage area)*Pudding River at Aurora 
Table 4.USACOE correlations to estimate flow at Willamette River at Portland


In order to determine which flow routing model was more appropriate for developing the boundary condition the flow in the Willamette River at Portland was estimated using both models for 1993 and 1994. The results were compared as shown in Figure 24, with daily data measured at the Willamette River at Portland gage station (USGS: 14211720) before it was discontinued in 1994. The difference between the estimated flows and the data were plotted in Figure 25.Error statistics were calculated, as shown in Table 5, to determine which approach gives a better representation of the flows at Portland.Based on the AME and RMS errors listed in the table the USACOE flow routing model provided a better approximation of the flow and was therefore used in developing the model boundary condition.


Figure24.Willamette River Flow at Portland comparison between data, the USGS model and the USACOE model


Figure25.Flow difference between data and routing models in Willamette River at Portland

 
Model comparison
Number of Days
Std Deviation

(m3/s)
AME (m3/s)
RMS (m3/s)
USACOE vs. Data
356
-37
52
75
USGS vs. Data
357
-58
64
120
Table 5.Flow Routing Model Comparison with data Statistics

 


Willamette River flow near Canby Ferry was obtained using part of the USACOE flow routing model described above. The flow approximation was used:

Mean daily Q at Molalla River-Canby
+Mean daily Q at South Yamhill
+Mean daily Q at Pudding-Aurora
+Mean daily Q at W.R at Salem
+Mean daily Q Ungaged between Salem and Oregon City
=Mean daily Q near Canby Ferry
Table 6 shows the gage station used in developing the flow boundary condition at Canby Ferry and the extent of the data at each station.

 
 
Site ID
Address
RM
Drainage Area mi2
Datum ft NGVD
Min Date
Max Date
Flow Count
USGS14191000
Willamette River at Salem
84.2
7280
106
01/01/1948
09/13/2000
19248
USGS14192500
South Yamhill River near Willamina
133
236
05/01/1934
09/30/1993
21338
USGS14193000
Willamina Creek near Willamina
64.7
315
06/01/1934
09/30/1991
20941
USGS14194000
South Yamhill River near Whiteson
502
82
10/01/1940
09/30/1991
18627
USGS14194150
South Yamhill River at McMinnville
5.6
528
50
10/01/1994
09/12/2000
2174
USGS14201340
Pudding River near Woodburn
314
130
10/01/1997
09/12/2000
1075
USGS14202000
Pudding River at Aurora
479
72
06/21/1993
09/30/1997
1563
USGS14210000
Clackamas River at Estacada
23.1
671
287
11/28/1992
09/30/1999
2498
USGS14211720
Willamette River at Portland
12.8
11100
2
07/06/1994
09/30/1999
1548
Table 6.USGS gage stations used in developing the boundary condition and the extent of their data

 


Due to data gaps at some of the gage stations correlations were developed between stations where data existed and the correlations in Table 4 were utilized to fill these gaps for the summers of 1993, 1994 and 1997 through 1999.The equation for developing the boundary condition flow at Canby Ferry used the daily flow data from the USGS gage station on the Willamette River at Salem for all five summers.The Molalla River flow was calculated using correlation 1 in Table 4 and the daily flow data from the Clackamas River at Estacada gage station.

For the summers of 1993 and 1994 the Pudding River flow was obtained from the Pudding River at Aurora gage station and for the summers of 1997 through 1999 the Pudding River at Woodburn gage station was used since the Aurora gage station was no longer operational.A correlation could not be developed between the two gages on the Pudding River because none of the data overlapped in time.The South Yamhill River flow at Whiteson was calculated using correlation 3 in Table 4 and the Pudding River flow at Aurora for 1993 and 1994.The South Yamhill River at McMinneville flow data was used for the flow at Whiteson for the summers of 1997 through 1999 since the stations were close together.The ungaged flow to the Willamette River between Salem and Oregon City was calculated using correlation 5 in Table 4 and the Pudding River flow (Aurora or Woodburn gage station data).

The remaining gap that existed in the data was from 05/01/1993 to 06/21/1993 at the Pudding River at Aurora gage station.Because of this gap the South Yamhill River flow and the ungaged basin flow could not be calculated.In order to fill the gap flow data from the South Yamhill basin was utilized.The goal of the analysis was to develop a daily flow time series for the South Yamhill River at Whiteson for 05/01/1993 to 06/21/1993.The first step in the analysis was to combine data upstream on the South Yamhill.Willamina Creek near Willamina (USGS: 14193000) has flow data from 1934 to 1991 and the South Yamhill near Willamina (USGS: 14192500) had flow data from 1934 to 1993. Table 6 lists the stations utilized in this analysis and the extent of the data at each station.A correlation was developed between these two gages station to complete the flow record at Willamina Creek for 1991 to 1993.The flow relationship used was 0.3926(Q, South Yamhill near Willamina, cfs)=Q, Willamina Creek near Willamina, cfs (R2=0.9459).The correlation was then used to complete the data set for Willamina Creek and then the flows from the two gage stations were added together.The flow below Willamina was then correlated with the South Yamhill River flow at Whiteson (USGS: 14194000).The correlation developed was: 1.8986(South Yamhill below Willamina)=South Yamhill at Whiteson (R2=0.8078).This correlation was then used to extend the flow record at Whiteson through June 21, 1993.The extended flow record at Whiteson was then used with correlation 2 inTable 4 to calculate the flow in the Pudding River at Aurora.Correlation 5 in Table 4 was then used to calculate the ungaged basin flow.Since several correlations were used to develop the necessary flows on the South Yamhill River a comparison was made between the Willamette River at Canby Ferry flow with the South Yamhill Correlations developed above and with using the Pudding River at Aurora data when it was available. Figure 26 shows the comparison plot with flow difference statistics.Figure 27 shows the Willamette River flows at Canby Ferry for the summer periods modeled.


Figure26.Willamette River at Canby Ferry flow comparison, Summer 1993


 
 

Figure27.Willamette River flow near Canby Ferry (RM 35), m3/s


Water Quality

Temperature and water quality data at Canby Ferry were obtained from DEQ (STORET data) and from two monitoring studies conducted by Montgomery Watson (1997 and 1999) for the City of Tigard and the Tualatin Valley Water District. These monitoring studies measured temperature and water quality parameters in the Willamette River at Wilsonville (RM 41).DEQ STORET water quality data consists of grab samples taken once a month to a couple of times a year.Water quality data monitored by Montgomery Watson were taken about once a week from 1994 to 1999.
Water quality data measured at Wilsonville and DEQ data measured at Canby Ferry were combined to generate the input files for the model.Figure 28 shows water temperature in the Willamette River at Canby Ferry for the period modeled. Table 7 shows a list of water quality parameters available for the Willamette River at Canby Ferry.Figure 29 through Figure 31 show the water quality constituents used for the boundary condition in the model. The procedure used for developing the water quality files from data can be found inAppendix J:Water Quality file development procedures.

Figure28.Willamette River at Canby Ferry water temperature, oC

Parameters
Willamette River at Canby Ferry
Alkalinity
X
Ammonia-Nitrogen mg/L
X
BOD5 DAY
X
Calcium Hardness
X
CHLRPHYLA
X
COLORPT-CO
X
Conductivity, mS/cm
X
D ORG CC
X
Dissolved Oxygen
X
D.O Saturation
X
E.COLI
X
ENTCOCCI
X
Fecal Coliforms ./100 mL
X
NH3+NH4-N TOTAL
X
Nitrate-N mg/L
X
NO2&NO3N-DISS
X
NO2&NO3N-TOTAL
X
PH
X
PHOS-DISORTHO
X
PHOS-TOT
X
SUSP SEDCONC
X
T ALKCACO3
X
T ORG CC
X
Temperature
X
Total Coliform
X
Total Hardness
X
Turbidity
X
UN-IONZDNH3-N
X
UN-IONZDNH3-NH3
X
Table 7.Water Quality parameters available for the Willamette River Boundary Condition
 


Figure 29.Willamette River at Canby Ferry Boundary Condition, RM 35




Figure 30.Willamette River at Canby Ferry Boundary Condition, RM 35 (Part 2)
 
 
 


Figure 31.Willamette River at Canby Ferry Boundary Condition, RM 35 (Part 3)