Discharge in the lower Columbia River basin, 1928-65

STEWART L. UDALL, Secretary

William T. Pecora, Director

Free on application to the U.S. Geological Survey, Washington, D.C. 20242 Abstract ----------------------------------· Introduction -------------------------------· Purpose and scope ---------------------· Description of the study area -----------· Acknowledgments ----------------------· Estimation of discharge ----·----------------· Summation method ---------------------·

10. Estimated "adjusted" flows-Columbia River at Longview, Wash ------------------------·
11. Estimated "observed" flows-Columbia River at mouth -----------------------------------
12. Estimated "adjusted" flows--Columbia River at mouth ---------------------------------· Gaged areas -----------------------· Ungaged areas --------------·------· Precipitation and evaporation -------· Time of travel ---------------------·
13. Correlation for inflow between The Dalles, Oregon, and Vancouver, Wash -----------------
14. Correlation for Willamette River at mouth ---·-------------------------------------------·
15. Correlation for inflow below Vancouver, Wash -----------------------------------------·
16. Comparison of estimated and measured flows-Columbia River at Vancouver, Wash _________
17. Estimated "observed" flow8-'Columbia River at Vancouver, Wash ------------------·------·--
18. Estimated "adjusted" flows-Columbia River at Vancouver, Wash ----·--·-----------------·
19. E:stimated "observed" flows-Willamette River at mouth --------------·-----------------·-·
20. Estimated "adjusted" flows-Willamette River at mouth --------------------------------·
21. gstimated "observed" flows-Columbia River at S.t. Helens, Oreg --·--------------·-------·
22. Estimated "adjusted" flows-Columbia River at St. Helens, Oreg --------·--------·-----·----
23. E"Btimated "observed" flows-Columbia River at Longview, Wash -----------------------·

TABLES

Adjusted monthly mean flows -----------· Accuracy of estimates -~----------------· Major inflow segments ___, _ Precipitation, evaporation, and time of travel ------------------------·

Comparison of estimated and ·measured flows ------------------------ Comparison of estimates by correlation and summation methods -----· Known deficiencies in methods ------·

m DISCHARGE IN THE LOWER COLUMBIA RIVER BASIN, 1928-65

BY HOLLIS M. 0REM

Estimates of monthly and annual mean discharge for five ungaged sites in the lower Columbia River are presented for water years 1928-65. These sites are Columbia River at Vancouver, Wash., Willamette River at mouth, Columbia River at St. Helens, Oreg., Columbia River at Longview, Wash., and Columbia River at mouth. Two tables of estimates are compiled for each site. One table lists estimates of "observed" flow, the flows that would be measured by a gaging station. The other table lists "adjusted" flows, the "observed" flows adjusted for storage changes in major reservoirs and for diversion at Grand Coulee Dam. Two methods are used in making these estimates. One method involves a detailed summation of flows from all gaged and ungaged areas, allowance for precipitation on and evaporation from the channel water surface, and adjustment of flow to allow for travel time from gaging stations at the Dalles, Oreg. (Columbia River main stem), and at Salem Oreg. (Willamette River). The other method is the same as the summation method except that flows from individual gaged and ungaged areas below The Dalles and Salem are not summed in detail. Instead, the entire drainage area below The Dalles is divided into three major inflow segments and the flow from each segment is obtained from a correlation with streams that are representative of the flow from that segment. Estimates of monthly mean flows are considered to be accurate within 2 or 3 percent.

INTRODUCTION The need for historical streamflow data for the lower reaches of the Columbia and Willamette Rivers is heightened by prospects for increased industrialization in those areas and by water-quality standards that are being set. Because of tidal effects in the lower reaches of these streams, no continuous records have been collected there until the last few years and then, only on the Columbia River at Vancouver, Wash. Because actual historical streamflow data are lacking, the obvious alternative is to estimate such records using the best available data. Throughout the Columbia River basin the Water Management Subcommittee of the Columbia Basin Inter-Agency Committee has used the period July 1f28 to June 1958 for planning purposes as this period includes the critical low-wa~r periods a~ well as high-water years such as 1948 and 195~. The estimates presented in this report includ ~ that period.

This report presents two tables of discharge estimates for each of five ungaged sites in the lower Columbia River basin for the period 1928-65 (see tables 3-12 at end of this report). One table lists "observed" monthly mean discharge; that is, the flows that would have been measured at a gaging station at the site. The other table lists "adjusted" monthly streamflow, which is the "observed" flo"V adjusted for storage changes in major upstream reservoirs and for diversions at Grand Coulee Dam. The "adjusted" flows represent f, "noreservoir" condition. Because these runoff figures more nearly represent natural flmv conditions, they are convenient for studies involving different sets of reservoirs or operations. The report also describes methods ur~d to estimate these monthly and annual strean1flows at the four sites on the Columbia River below Bonneville Dam and for Willamette River at mouth. The four Columbia River sites are at Vancouver, Wash., below Lewis River at St. Helens, Oreg., below Cowlitz River at Longview, Wash., and at mouth. No attemr~ has been made to estimate daily flows.

The lower Columbia River basin as used in this report includes that part of the basin

Figure 1.-Map of lower Columbia River basin. downstream from the gaging stations on Columbia River at the Dalles, Oreg., and on Willamette River at Salem, Oreg. (see fig. 1). The drainage area at the Dalles is about 237, 000 square miles and the drainage area at the mouth of the Columbia River is about 258,000 square miles. The incremental area between these points is, therefore, 21,000 square miles, lying almost entirely west of the Cascade Range. Only about 4,600 square miles of the area below the Dalles and Salem is ungaged, and about 400 square miles of that is river channel water surface or low sand islands which lie largely in the extreme lower reaches where the river becomes several miles wide. However, the water from the 21,000 square miles is an important part of the flow at the mouth of the Columbia River. Average runoff from this area is 3.62 cfs per sq mi (cubic feet per second per square mile) whereas average runoff from the area above The Dalles is only 0.824 cfs per sq mi. Most of the study area is between 500 and 5,000 feet elevation and is heavily forested; but some lower areas, such as the Willamette River valley, are largely under cultivation with interspersed forests. There is only one impoundment (Bonneville Dam) below The Dalles on the main stem of the Columbia River, but this is a Hrun-of-river" reservoir with inconsequential monthly storage changes. In determining "adjusted" monthly flows it was necessary to adjust for (1) storage changes in nine major reservoirs above The Dalles, (2) pumpage (diversions) at Grand Coulee Dam, and (3) storage changes in seven reservoirs above Salem in the Willamette River basin and three reservoirs in the Lewis River basin.

The author is indebted to Vail Schermerhorn of the U.S. Weather Bureau, David M. Rockwood and Carlton Jencks of the Corps of Engineers, and William Haushild of the U.S. Geological Survey foJ'l information on time of travel used in routing flows. Basic data f.or gaged areas are from the Oregon and Washington district Offices of the U.S. Geological Survey. ESTIMATION OF DISCHARGE Two methods were used for estbnating monthly mean flows at the five ungagoc sites. The first and basic method invplved summation of flows for all gaged areas with estimates .of flow for all ungaged areas (Orem, 1961). The second was a shortcut method tr.~t involved dividing the area between the Dalles, Oreg., and the mouth of the river intc three major segments and estimating the flow for each segment by correlation procedure:~. The second method was used for some of the early years before many of the gages were in~talled on intervening areas l>etween The Dall~s and the mouth of the river.

The summation method consisted of starting with gaged flows of the Columbia River at the Dalles, Oreg., and adding tri"'•Itary flows as they entered the river. Table 1, which illustrates the method, is for the 15-year period covering water years 1943-57. The table was prepared about 1960 when average ftc w for that standard period was needed for Columbia River at Vancouver, Wash., and Columbia River at mouth. That table is shown in this report in preference to using a specific month as an example because (1) average values for ungaged areas were estimated in this table, and (2) the tabulation is for a 15-year period and, therefore, the computation is not complicated with travel-time adjustments. To estimate flows for individual months, travel-time adjustments are used when flow at the end of the month is materially different from fow at the beginning of the month. These adjustments will be explained in more detail later in the report.

For most of the 1928-65 period streamflow records are available for all but abcnt 1.6 percent of the Columbia River drainage area. This small ungaged area accounts for 4.6 percent of the flow at the mouth of the Col'lmbia River. The gaged areas are listed in u.ble 1. The "Remarks" column in the table inc.icates that even in the period 1943-57 there were a few months or years of missing record for several of the gaged areas and that flovr data for these short periods were estimated by correlation with records of a nearby stream.

Estimates of ungaged average inflow to the Columbia River for five areas are given in table

1. Because the ungaged areas and flows are small compared to the flow of the Columbia River, estimates of monthly flow at the ungaged main-stem sites cannot be materially in error because of erroneous estimates of ungaged inflow. Nevertheless, con~iderable care and effort went into- these estimates. Average runoff rates for the areas were determined in terms of the average runoff rate for a nearby gaged area based on a runoff map developed by Hulsing and Kallio (1964) of the Geological Survey. Their map of isopleths pf runoff was developed from a st:.:dy of all records of discharge in Washington and Oregon and an isohyetal map (lines of equal precipitation) prepared by the Portland District, Corps of Engineers.

Downstream from the gaging stations in Oregon at The Dalles and at Salem there are 400 square miles of river channel and islands. Increments of flow were included for precipitation directly on this surface and decrements were included for evaporation from the surface. Because these components are relatively small and tend to cancel each other, they were included in the estimates of monthly flow for only the two lower ungaged sites. About half the 400 square miles is upstream from the ungaged main-stem site below the Cowlitz River. Therefore, precipitation on and evaporation from the 400 square miles of water surface were taken into account f.or flow of Columbia River at mouth and from 200 square miles for Columbia River below Cowlitz River. Monthly flow increments were obtained by averaging the amount of precipitation recorded at Weather Bureau stations in Oregon at Bonneville, Salem, Portland, and Astoria and converting that average to equivalent cubic feet per second. Monthly flow decrements for evaporation were obtained using average monthly values derived from U.S. Weather Bureau Technical Papers 13 and 37 (U.S. Weather Bureau, 1950, 1959). Paper 13 shows that annual evaporation averages 24 inches for the area downstream from The Dalles and Salem. Paper 37 indicates a distribution by months of this 24 inches on the basis of records for Seattle, Wash., the most repre~~ntative point for which year-round records are available.

When flow changed materially during the month, the estimates for the five ungaged sites required adjustment for time cf travel of the water. Adjustments were made only for time of travel from The Dalles, Oreg., to Vancouver, Wash.; from The Dalles to mouth of the Columbia River; and frorr: Salem, Oreg., to mouth of the Willamette River. Travel times and changes of flow for other tributaries were neglected.

mn!h) --========================::J~oo25~

mn;h) --========================B2~ 25

Travel times were based on data and advice from Weather Bureau, Corps of Engineers, and Geological Survey personnel, plus some observations of velocities at Hc0d River, Vancouver, and a few other points. Although the flows at the lower end of the reach were the basis for the assignment of travel time through the reach, the flows at the upper end of the reach were tl'e basis for the change in discharge used in adjusting the "observed" flows at the ungagen site for travel time. For example, in May 1948 the "observed" monthly mean flow at Vancouver was 493,500 cf.s (cubic feet per second). F-:-om the above schedule the travel time from The Dalles to Vancouver was, therefore, taken as 1 day. The daily mean flow at The Dalles was 249,000 cfs on April 30, 1948, and 999,000 cfs on May 31, 1948. The change in flow at The Dalles during May was, therefore, + 750,000 cfs and the needed adjustment to monthly mean flow at Vancouver was negative, computed as follows: 750,000 cfs x 1 day 31 days The "observed" mean flow of 493,500 cfs at Vancouver was adjusted to 469,300 cfs. If the travel time had been 2 days, then 2 days rather than 1 day would have been used in the numerator of the above equation. The adjustment of -24,200 cfs shown above is about the maximum needed both in amount and in percentage of flow and is about 5 percent of the "observed" flow at Vancouver. For simplicity, the assumed travel times to Vancouver were used also for the ungaged sites below Lewis River and below Cowlitz River, even though it is obvious that somewhat longer travel times to those sites should be used., A good verification of the applicability and validity of the assumed travel time from The Dalles to Vancouver was possible because the Geological Survey measured flows at Vancouver during the 1964, 1965, and 1966 water years. Table 2 shows a comparison of estimated "observed" flows and measured flows at Vancouver for those three years. In 28 of the 36 months of record the adjustment for time of travel improved the agreement of the estimated "observed" flows with the measured flows. The increase in discharge during the spring rise in certain years was so large that application of the time-of-travel adjustment resulted in larger flows at Longview than at the mouth, but the difference never exceeded about 2 percent and it is probable that the greater upstream flows could actually occur under certain conditions.

In the correlation method the area be~ween The Dalles, Oreg., and the mouth of the river was broken into three major segmentr and separate correlations were used to estimate the inflow from each segment. These segments are Columbia River from The Dalles to Vancouver, Wash., Willamette River from falem, Oreg., to mouth (including Multnomah Channel), and Columbia River from Vancouyer to mouth. Good estimates of the inflow betweer The Dalles ana Vancouver were obtained by· correlating monthly inflow figures for 191:3---65 (obtained by the summation method) with the sum of flows in Hood and Wind Rivers., Separate correlations were made for each 1nonth of the year to adjust for seasonal variations in the relationship. Figure 2 shows the correlation developed for January. Similarly, the flows of Willamette River at the mouth correlated well with flows of Willamette River at Salem after both wer~ adjusted for storage changes upstream. Figure 3 shows the correlation for the month of October. Correlations for each month were used to obtain "adjusted" monthly mean flows at the mouth of the Willamette for the early years, and the adjustments were then subtracted to obtain the "observed" monthly mean flows. The "adjusted" inflow to Columbia River below Vancouver also correlated well with the sum of flows in Cowlitz and Lewis Rivers adjusted for upstream storage in Lewis River reservoirs. Figure 4 shows the correlation for the month of January. Here, too, the adjusted inflow obtained through correlation was converted to "observed" inflow. Discharges obtained by correlation for these three major segments of inflow were then used directly to obtain estimated flows at three of the five ungaged sites; however, estimated flows for the other two ungaged sites required further manipulation of data. Estimated flows for Columbia River below Lewis River were obtained by adding to the flow at Vanc-:mver, January 1943-65

Figure 2.-Correlation for inflow between The Dalles, Oreg., and Vancouver, YTash.

the flows of Willamette River at mouth, Lewis River, East Fork Lewis River, and an increment for Engaged area prorated among the months based on East Fork Lewis River. Estimated flows for Columbia River below Cowlitz River were obtained by a further addition of flows for Cowlitz and Kalama Rivers and another increment for ungaged area distributed according to Kalama River monthly flows.

The procedure outlined above did not result

in "observed" monthly mean flows comparable to those obtained by the summation method shown in table 1. Adjustment~ for precipitation on and evaporation from the 400 square miles of water surface and for time of travel because of changing discharge also had to be made. This was done in the SG.me way as for the summation method, with tr~ following exception; the evaporation correction applicable to flows at the mouth of the Columbia River was made an integral part of the correlation developed for the inflow segment from Van-

=>a::

U..ILL. 1- enu ~iii c=> ~(.)

Figure 3.-Correlation for Willamette River at mouth.

Figure 4.-Correlation for inflow below Vancouver, Wash.

couver, Wash., to the mouth. This procedure was used because the same annual evaporation of 24 inches and the same monthly distribution of that 24 inches were used each year. Precipitation corrections were not made in the same manner because the actual amount of

precipitation by months and years is known.

The tables of "observed" mont}' ly mean flow for water years 1928-65 for each of the five ungaged sites were developed uring both the summation and correlation methods. Tables of "adjusted" monthly mean flows were then compiled by adjusting the "observed" flows for upstream storage changes. The reservoirs for which adjustments were made are:

1. Nine major reservoirs in the Columbia River basin; namely, Kootenay Lake, Hungry Horse Reservoir, Flathead Lake, Noxon Rapids Reservoir, Pend Oreille Lake, Coeur d'Alene Lake, Franklin D. Roosevelt Lake, Chelan Lake, and Brownlee Reservoir.

2. Willamette River basin reservoirs of Hills Creek, Lookout Point, Cottage Grove, Dorena, Fern Ridge, Cougar, and Detroit.

3. Lewis River reservoirs of Swift, Yale, and Merwin. Adjustments for upstream storage were computed using the system of adjusting that has been standard with the Water Management Subcommittee of the Columbia Basin Inter-Agency Committee. In that system, adjustments are made strictly on a calendar-month basis with no allowance for travel time. In other words, in any given month the adjusted flow fr<>r Columbia River at the mouth may include an adjustment for a storage change that occurred largely during the last day or two of a month and the effect would not be felt at the mouth until the following month. However, such events should not lead to consistently large errors. At times during the final days of the month one reservoir may store while another releases, and the resultant differences in flow would tend to cancel each other.

Because of the many approximations involved, it is difficult to place accuracy limits on the estimates of flow. The ungaged area between the mouth of the Columbia River and The Dalles, Oreg., is only 1.6 percent of the total Columbia River drainage area, and the ungaged flow is estimated to be less than 5 per cent of the total flow.. Therefore, fairly large errors in estimates of ungaged flows would not greatly affect estimates of total flow. However, in some months, the approximations used could easily result in errors of a few percent if errors in the several estimates were in the same direction. It is morfl likely that the errors would be compensating.

The only real check of the accuracy of estimates lies in a comparison of estimatd flow for Columbia River at Vancouver, Wash., for water years 1964-66 with measured flC''¥S for those years (table 2). The estimates for 1964 and 1965 were by the detailed sum.mation method whereas those for 1966 were by the correlation method. The mean deviation of the three estimated annual mean flows from the measured flows is 0.24 percent and maximum deviation is only 0.42 percent. Corresp>nding figures for the 36 monthly mean flows are 2.3 and 5.7 percent, respectively. Howeve:r, this does not mean that the estimated flows are in error by this much, as the measured fLows in this tide-affected reach of the riv may be a few percent in error. The comparison does show that the estimates are rear0nably close to the measured flow. It is the author's opinion that in general the estimates are accurate to within 2 or 3 percent.

The correlation, or shortcut, method has been used to. estimate fLows of the Columbia River at mouth for cu,rrent data releaser, Comparison of the 60 monthly estimates thus computed for water years 1961-65 with those computed by the summation method and shown in table 11 (p. 23) indicates that the z.verage difference is 1.7 percent and the maximum difference is 4.6 percent. Average difrerence for annual mean flows was 0.6 percent. Some of this difference is undoubtedly caused by the fact that provisional figures of disn.harge for index streams were necessarily u';<~d in the shortcut method but final published figures were used in the summation method to obtain the flows for 1961-65 shown in table 11. The shortcut method was used to eft;imate monthly discharge fr<>r 1928-40 but for those years it was possible to use publishe:d discharge for all index stations, thereby eliminating provisional data as a source of error.

:Known deficiencies in the results pre:~ented herein are as follows:

1. The 80,000 acre-feet of water supply for Portland, Oreg., diverted from the Sandy River basin as of 1955 had gr,own to about 110,000 acre-feet per year in 1965. Between one-half and two-thirds of this amount, or an average of about 100 cfs, was returned to the lower Willamette River in recent years, but no provision for this was made in the computation.
2. The increment for precipitation on and decrement for evaporation from the 400 square miles of river surface was inCluded in the tables only for Columbia River at mouth. Fifty percent of this correction was applied below the Cowlitz River. Actually, about 40 percent of the water surface is above St. Helens, Oreg., but these factors were not included in tables for the three upstream sites.
3. Although the compilations used to estimate flows for this report take into account all known surface runoff, they do not include ground-water inflow directly into the rivers below the gaging stations. This may be important particularly for the Willamette River below Salem, Oreg., where inflow is from extensive areas of permeable alluvium. The extent of the inflow cannot be evaluated until a gaging station has been installed at Portland.
4. Adjustments for changing discharge through use of travel timer are only approximate and are used only f,or flows passing the gaging stations on Columbia River at The Dalles, Oreg., 2nd Willamette River at Salem; they are neglected for tributary streams.
5. Adjustments for changes in r·eservoir storage, used to convert "observed" flows to "adjusted" flows, are made strictly on a calendar-month basis with no allowance tor travel time. This could result in occasional small errors.

REFERENCES

Hulsing, Harry, and Kallio, Nicholas, 1964, Magnitude and frequency of floods in ti'~ United States, Part 14, Pacific slope basins in ,O"..egon and lower Columbia River basin: U.S. Geol. Survey Water-Supply Paper 1689, fig. 3. Orem, H. M., 1961, Computation of total flow of the Columbia River at the mouth, in, Short papers in the geologic and and hydrologic sciences: U.S. Geol. Survey Prof. Paper 424-C, p. C148-C149. U.S. Weather Bureau, 1950, Mean monthly and annual ev•aporation from free water surface for the United Jtates, Alaska, Hawaii, and West Indies: U.S. Weather Bur. Tech. Paper 13. U.S. \Veather Bureau, 1959, Evaporat~on maps for the United States: U.S. Weather Bur. Tech. Paper 37. TABLES 1-12 Table !.--Outline of summation method for estimating flow of Columbia River at mouth

Station Columbia River at The Dalles Dam

Klickitat River near Pitt Hood River and conduit nr Hood River White Salmon River nr Underwood Little White Salmon ~above Lapham Creek near Willard Dec.l944-59 Wind River near Carson Ungaged area-The Dalles Dam to Bonneville Dam Columbia River at Bonneville Dam-subtotal Washougal River nr Washougal Sept.l944-59 Sandy River below Bull Run River nr Bull Run

Ungaged area-Bonneville Dam to Vancouver Columbia River at Vancouver-subtotal Willamette River at Salem Mill Creek at Salem ~outh Yamhill River near Whiteson Ungaged area-Salem to Wilsonville Mollalla River near Canby

t:;

Pudding River at Aurora Tualatin River near Willamette Clackamas River near Cazadero Johnson Creek at Sycamore Ungaged area above mouth of Willamette River Willamette River at mouth Columbia River below Willamette River-subtotal Lewis River at Ariel *East Fork Lewis River near Heisson *Kalama River below Italian Creek, near Kalama Cowlitz River at Castle Rock *Elochoman River near Cathltrnet *Youngs River near Astor~a · Precipitation on water surface Evaporation from water suiface Ungaged area-below Willamett.e River Columbia River at mouth-uncorrected Sum of change during period-Columbia + Willamette Correction for time of trave!'to mouth Columbia River at mouth-"observed" Adjustment for storage change in reservoirs Columbia River at mouth-"adjusted"

!/Cubic feet per second.

~/Cubic feet per second per square mile.

1943-46 correl.with Toutle R.

Avg.precip.for 4 stations. ~vg.figures used for months. 2.76 x sum of 4 starred rivers. ater ear 1964 Summation of upstream observed flows Correction for time of travel Estimated "observed" discharge Measured discharge

1966 Summation of upstream observed flows 1/ 111.7 Correction for time of travel Estimated "observed" discharge Measured discharge

l/Preliminary figures.

Table 2.--Comparison of estimated and measured flows - Columbia River at Vancouverz Wash.

Difference Percent deviation of estimate Avg. dev. = 25.9/12 = 2.2%

[Tnousanas of CUblC feet per secona]

Year

• .42 Water Oct. year 1928 1929 1930

Nov.

87.6 Jan.

88.6 Feb.

88.3 Mar.

86.0 May June July Sept. Apr.

98.3 Water Oct. ear 1928 1929 1930

Nov.

86.5 Table 4.--Estimated "adjusted" flows -Columbia River at Vancouverz Wash.

Dec.

98.0 Jan.

82.7 [Thousands of cubic feet per second] Feb.

88.5 85.0 93.6 91.5 Mar.

79.1 Sept. May June July Aug. Apr.

-338 93.2 The ear 99.6 90.0 98.3

Mar.

69.2 96.8 4f.2 35.8 61.9

June

15o4 18.6 13.4 26.7 11.1

Sept.

3.65 4,5r 7.5.1 4.02 4.61

8.43 7.71 8.47 ll.6 8.03

8.13 7.9:i'

Tile year

41.8 31.9 33.7 33.8 41.2 Water year 1928 1929 1930

1961 1962 1963 1964 1965 Oct. 22.0 5.83 4.50

4.55 11.6 42.7 l1.5 9.50

6.65 9.62 17.8 7.79 5.56 Table 6.--Estimated "adjus:ted" flows - Willamette River at mouth

Nov.

66.7 31.5 48.5 52.6 21.2 Dec. 58.0 22.7 48.7

48.9 78.4 53.8 34.8 [Thousands of cubic feet per second] Jan. 69.5 45.7 21.7

43.9 41.3 19.8 Feb. 35.7 35.0 85.7

37.8 71.9 49.9 70.8 Mar. 58.1 37.5 29.0

58.3 34.6 47.1 25.3 Apr. 72.8 47.3 19.0

38.1 54.6 73.6 32.9 27.7 May 29.6 35.5 22.2

35.7 40.4 54.9 28.3 21.6 June 10.2 24.4 15.1

15.1 18.8 13.5 28.5 11.8 July Aug. 6.18 8.56 6.08

10.4 6.24 Sept. 4.87 5.83 4.12

4.10 5.8 3 6.88 4.34 5.n

5.Z5 6.46 4.E9 4.43 3.(6

4.46 5.r4 7-,07 4,21

5.70 6.fl)

5.69 5.19 5.44 4.63 3.53 The ear 35.7 23.7 21.6

41.0 32.4 33.7 34.0 41.2 Water Oct. year 1928 1929 1930

Nov.

95.0 Table I.--Estimated "observed" flows -Columbia River at St. Helens, Oreg.

Dec.

94.1 Jan.

96.1 [Thousands of cubic feet per second]

Feb.

67.2 The Mar. May June July Aug. Sept. Apr. year

l9

Water year 1928 1929 1930

1961 1962 1963 1964 1965 Oct.

us

96.6 Table 8.--Estimated "adjusted" flows - Columbia River at St. Helensz Oreg.

Nov.

Dec.

69.2 Jan.

ll9 [Thousands of cubic feet per second]

Feb.

97.5 May June July Apr. Aug. Mar.

28l-The Sept.

no

no

ll7 Water Oct. ;rear 1928 1929 1930

Nov.

96.7 Table 9.--Estimated "observed" flows -Columbia River at Longviewz Wash.

Dec.

96.2 Jan.

95.4 96.5 [Thousands of cubic feet per second] Sept. The Mar. July Aug. Feb. Apr. May June

Water year 1928 1929 1930

1961 1962 1963 1964 1965 Oct.

Nov.

llO.

86.8 76.0 93.8 ll3 87.3

83.4 Table 10.--Estimated "adjusted" flows -Columbia River at Longview, Wash.

Dec.

71.3 Jan.

89.0 [Thousands of cubic feet per second] The Feb. Mar. May June July Aug. Sept. Apr. ear

water Oct. year

The July Feb. Mar. Apr. May June Aug. Sept. z:e;:tr

77.3 Water Oct. year 1928 1929 1930

Nov.

84.0 Table 12.--Estimated "adjusted" flows -Columbia River at mouth

Dec.

72.5 Jan.

93.8 [Thousands of cubic feet per second] The Sept. May June July Aug. Feb. Mar. Apr. year

Where this page came from

This page was imported from U.S. Geological Survey. Published by the U.S. Geological Survey and, as a work of the United States government, in the public domain.

Nobody has written it yet — it is the source material at a new address, which is why search engines are asked to skip it and why no one earns from it. It is up for grabs: take it on, and it is yours to rewrite and to earn from.