Loading...
HomeMy WebLinkAbout2000 Characterization of Surface WaterDave R Nelson - M- Are p ort nov 17 2000.doc Pa e Characterization of Surface -Water Resources, Makah Indian Reservation and Vicinity, Northwest Olympic Peninsula, Washington Bureau of Reclamation Water Supply, Use, and Conservation Technical Service Center D -8520 Denver, CO November 17, 2000 Thomas W. Perry, IV Hydrologist ttyerryaldo.usbr.eov 303 -445 -2567 Dave R Nelson - MAKAHre ort nov 17 2000.doc Page 2 Characterization of Surface -Water Resources, Makah Indian Reservation and Vicinity, Northwest Olympic Peninsula, Washington Bureau of Reclamation Water Supply, Use, and Conservation Introduction — In recent years, increased use of local ground and surface water supplies by the Makah Indian Tribe has resulted in water shortages, consequent water quality problems, and environmental problems related to insufficient streamflows for anadromous fish. Currently the Tribe has limited water storage facilities and water quality problems that require shifting from surface to ground water resources during each year. This results in a challenging water resource management situation regarding supply availability, quality concerns, and water treatment difficulties. This report provides a comprehensive appraisal level assessment of surface -water availability for watersheds that are of reasonable interest in providing water to the Makah Reservation. The broad objective of this study has been to collect and analyze data to give an initial estimate of water quantity in areas accessible to the Reservation, and to evaluate water availability for an estimated target demand anticipated to exist annually during the summer months of June, July, and August. Description of Study Area — The study area, which includes the Makah Indian Reservation, encompasses Cape Flattery, at the northwestern tip of the Olympic Peninsula. This area is characterized by a somewhat mild, marine climate, where air temperatures at sea level reach a mean monthly temperature of about 40° F in January, and 60° F in July. Daily variation in temperature is rarely greater . than about 10° F. There is, however, a pronounced seasonal variation in precipitation. The average sea- level monthly precipitation at Neah Bay reaches a yearly maximum of about 16 inches in December, and a yearly minimum of just over 2 inches in July and August. Average annual precipitation at Neah Bay, is about 100 inches. In this region, precipitation occurring below 1500 feet in elevation falls as rain, or may be present as snowfall in the winter, which melts in a few days. Higher elevation precipitation accumulates as a snow pack in the winter that melts during the summer. Hence, streamflow originating at lower elevations has a runoff distribution similar to the annual precipitation.distribution. Streamflow originating at higher elevations shows two characteristic seasonal peaks in flow; one occurring from heavy precipitation in the mid- winter, and the other occurring from snowmelt in the spring on into the summer. Vegetative cover found throughout this region of the Olympic Peninsula, consists of dense conifer stands. Along the lower elevation maritime region of Pacific coast, vegetation is characterized as rain forest, which includes a dense understory of bushy shrubs, flowering plants, ferns, and green mosses. Numerous wetlands and marshes commonly have rushes, willows, sedges, reeds, cattails, and many other water loving plants. In areas outside of Olympic National Park, extensive clear- cutting and harvest of conifers is common. On lands in the vicinity of the Makah Indian Reservation, some of these activities involve reforestation of clear - cut areas for later harvest. Soil cover and eeoloav - Soil cover in this region has evolved in response to geologic influences. Soil cover, in turn, influences the character of vegetative cover. Principal soils units, as classified by the Soil Conservation Service, include three major soils groups, as follows: 1) Soils on hills ; The principal soil unit in this group is the Ozette - Kydaka soils series. These units are found on slopes ranging from 0 to 35 percent, and support vegetation of mainly conifers and shrubs. Elevation is from 100 to 1800 feet, where the average annual precipitation is 75 to 120 inches. In this grouping, Ozette soils are deep and moderately well drained. These soils are formed in loess and in glacial till derived dominantly from sandstone and siltstone. The surface is covered with a mat of organic material, and the 2 Dave R Nelson - MAKAHre ort nov 17 2000.doo Pa e3 surface layer is silt loam. Kydaka soils are primarily in basins (drainages). These soils are moderately deep and poorly- drained. They are formed in loess and glacial drift over compact glacial till. The surface is covered with a mat of organic material and the surface layer is a silty clay loam. 1) Soils on foothills The principal soil unit in this group is the Palix- Ilwaco soils series. These units are found on slopes ranging from 15 to 90 percent and support vegetation of mainly conifers and shrubs. Elevation range for this series is from sea level to 1800 feet. Average annual precipitation is 70 to 120 inches. Palix soils are on foothills. These soils are deep and are formed in material weathered from siltstone and very fine sandstone. The surface is covered with a layer of organic material, with the surface layer being loam. Ilwaco soils are also on foothills. These are very deep soils formed in material weathered from sandstone and loess. The surface is covered with a layer of organic material and the surface layer and subsoil are silt loam. 1) Soils on mountains The principal soil unit in this group is the Snahopish- Solleks -Makah soils series. These are moderately deep and very deep, well- drained, steep to extremely steep soils on mountainsides. Slope is 35 to 90 percent. Supported vegetation is mainly conifers and shrubs. Elevation is 300 to 2000 feet, with average annual precipitation of 90 to 140 inches. Snahopish soils are very deep soils developed on mountainsides and formed of material derived from sandstone. The surface is covered with a mat of organic debris, with the surface layer being gravelly loam. Sollecks soils are similar, but are moderately deep. Makah soils are also developed on mountainsides. These soils are very deep and are formed in the residuum and colluvium derived from basalt. The surface is covered with a mat of organic material. The surface layer-and subsoil are gravelly loam. - The most influential geologic event contributing to soils development and in shaping the present landscape, has been glaciation during the Pleistocene. Extensive glaciation of the northern end of the Olympic Peninsula occurred about 15,000 years ago just as the last advance of Pleistocene continental glaciers was coming to an end. This extensive glaciation covered Vancouver Island, and the Pacific Ranges in Canada. Movement of the ice southward caused one glacial lobe to move into the Puget Sound Lowland, while another lobe moved westward along the Strait of Juan de Fuca and around the northern end of the Olympic Peninsula. This ice penetrated upland into the Olympic Mountains, and covered most of the northwestern end of the Olympic Peninsula. For the eastern lobe in the Puget Sound Lowland, ice was constrained by the Olympic Mountains on the west, and the Cascade Range on the east. This glaciation declined rapidly after about 13,500 years ago. An extensive area of glacial till and morainal material from this glaciation is found as surficial deposits in much of the Sooes, Ozette, and Dickey River drainages, as well as in most of the glacially dominated upland valleys of the region. Stream alluvium in many valleys is derived from this glacial material. ' Geologic composition of the northwestern Olympic Peninsula is quite varied. The central core of the Olympic Mountains, and the region to the west of the Olympics, is comprised of a subduction complex that includes marine sandstone and other sedimentary rocks. These rocks are Eocene (35.4 ,to 56.5 million years) in age. Surrounding the Olympics is a sequence of marine volcanic and sedimentary rocks of Miocene to Paleocene ( -16.3 to 60 million years) in age. These units are separated from the Olympic subduction complex by the Hurricane Ridge thrust fault. This thrust fault is present in the study area and may be observed separating the lowland valley of the Sooes River from the Makah Peaks to the northeast. Hence, the Olympic core subduction complex which is to the south and southeast of this fault, is being thrust under these surrounding units which lie to the northeast of the fault. 3 Dave. R Nelson - MAKAHre ort nov 17 2000.doc Page 4 Drainaee and basin moroholoev - Watersheds comprising the drainages of the main streams found in the study area, have a barbed stream- drainage pattern which may vary to a somewhat deranged pattern where smaller drainages are incorporated into larger drainages. Drainage patterns occasionally show influences from lineaments, or other faulting and fracturing which may be identified with such linear features. This can be seen in the Hoko and Sekiu river drainages. The linear northeastward course of the Hoko River across the Sekiu- Stolzenburg Mountains highland, and the linear eastward, course of the North Fork Sekiu River and the Sekiu River below the North Fork confluence, may be due to lineaments. Other streams may show similar features, however, upland areas of some drainages tend demonstrate anomalous characteristics in drainage integration as the headwater - tributary watersheds of some streams have been seemingly captured from, neighboring watersheds. Some of the derangement in drainage patterns may be in response to glacial drift and outwash which has accumulated in valleys and lowland areas where channel choking may have occurred. An example of such derangement is seen in the headwaters of the North Fork Sekiu River, or in the headwaters of the Hoko River. Prior to extensive glaciation, integration of these upland watersheds into the Sooes River drainage, for the Sekiu, or into the Dickey River drainage, for the Hoko, seems likely. Land uses — Primary utilization of land relates to forestry and harvest of forest products. As such, logging practices are of primary - importance in consideration of hydrologic impacts. The development of logging roads, and clear- cutting large areas of forest, causes an increase in the sediment loading of streams, a deterioration of water quality, and an alteration in the discharge character of streams. Other forest clearing practices, such as forest burning, can produce similar yet more extreme results. Recovery of clear-cut areas by regrowth; can mitigate these concerns. However, large areas of cleared lands that are kept as open fields or meadows can result in an altered character in streamflow. Although the logging history of the region is not well documented, logging was well underway by the mid - 1930s due to the development of logging railroads in the area. These developments gave way to an extensive network of logging roads and clear- cutting practices. From 1940 to 1973, for instance, approximately 25 percent of the land in the Ozette River drainage basin had been clear-cut. As the logging industry expanded in the late 1970s and early 1980s, clear- cutting remained pervasive on lands of the northwestern Olympic Peninsula exclusive of Olympic National Forest. A review of aerial photography taken in 1995 shows that within the Makah Indian Reservation boundaries, a history of recent clear- cutting is evident. Water use — Water uses for Neah Bay and the Makah Indian Reservation primarily consist of domestic and commercial uses. Water meeting these needs is presently supplied from several groundwater wells, and direct surface water diversions from one surface -water storage reservoir and a supplemental stream infiltration gallery. The wells are located in stream alluvium adjacent to streams, such as the Waatch River, and are most likely in hydraulic connection with the stream. Within the channel of the Waatch River, an infiltration gallery in the alluvial streambed supplies much of the water used during summer months when only poor quality water is available from Educket Creek reservoir. For Educket Creek reservoir, inflow during summer months is insufficient to flush water in storage, and thereby eliminate stagnation of stored water. During higher periods of precipitation, particularly the winter season, inflow is sufficient for flushing and maintaining good quality water, and storage is satisfactory for meeting demands. Limitations in meeting demands during high precipitation times of year are presently due to the poor water quality of the supply, and the inability of the water treatment plant to treat highly turbid water at a sufficient rate. In the summer, conservation methods are sometimes required to alleviate shortages during peak summer season demands. 4 Dave R Nelson - MAKAHre ort noy 17 2000.doc Page 5 Assessments of future demands as indicated in the Project Summary (1996,1999), indicate the target rate for the supply to meet future summer season demands must meet a 12 hour flow requirement of about cfs, or about 900 gallons per minute. Meeting this flow requirement is equivalent to minimum surface -water flow rate of about 124 acre -feet per month because streamflows less than this rate will not meet the 12 hr flow requirement. -Therefore, achieving this target rate will require two objectives to be satisfied. First, a sufficient supply must be found, and second, capacity of the water treatment plant must be increased to accommodate this demand. The rate of supply must exceed the target of approximately 124 acre -feet per month for the months of June, July, and August, which are the most critically dry months of the year. Diversion of water from the source of supply must also accommodate environmental considerations. To examine the feasibility of meeting these demands, the following sections explain the analysis relevant gaging- station records, and characterization of the surface -water supply. Flow records used in this analysis are essentially for recorded natural flows existing at several gaging stations on, or near, the Makah Indian Reservation. Stream eaeinE records - Assessment of the surface -water supply was initiated by examination of U. S. Geological Survey (USGS) stream gaging records and of National Oceanographic and Atmospheric Administration precipitation records for weather stations in the region. These records are of total monthly streamflow, in acre -feet, and of total monthly precipitation, in inches. Stream gaging- station records for streams on, or in the vicinity of, the Makah Indian Reservation, are either incomplete as continuous records, or of apparently insufficient length as continuous records. However, because the climatic regime of the northwest Olympic Peninsula is so consistent, gaging - station records may be useable if an adequate sequence comprising a continuous discharge record can be reconstructed, and sampled. Meeting this objective is generally the case for streams that have been gaged in this region. Reconstruction of missing portions of these records was accomplished using statistical reconstruction through regression analysis. Regression analysis and statistical-reconstruction of missing portions of these records, was carefully accomplished. The process is completed by beginning with primary records of sufficient length to establish a basis for correlation with records of interest. Primary records may inherently require reconstruction of missing elements. This process was initiated by acquisition of all relevant precipitation records of sufficient length. Comprehensive databases providing precipitation records are available from the National Oceanographic and Atmospheric Administration, and from compact disc records of U. S. Weather Bureau data that are published by Hydrosphere. The precipitation records of interest essentially included Tatoosh Island, Neah Bay, Clallam, Forks, and Sappho. These records comprise a sequence documenting a continuous precipitation history of the region from about 1885 to the present. These records were carefully examined and missing monthly values were researched, and recovered from Weather Bureau monthly summaries showing values for missing entries had been previously estimated and published. Remaining missing values in these records were then statistically reconstructed by correlation with the record of one, or more, nearby precipitation stations. Comprehensive databases providing stream gaging - station records are accessible through the USGS, or are available as published by Hydrosphere. Stream gaging - station flow histories were initially reconstructed for selected stations of interest, using statistical correlation with records that had a flow history, although possibly incomplete, spanning more than about ten years. These records were cross - correlated by calendar month, using a digital method similar to that developed by Reid, Carroon, and Pyper (1969). This process was initiated with monthly tabulations of streamflow in acre -feet for the Hoko and Dickey rivers. Missing months that could not be recovered by correlation with a primary nearby gaging station having a sufficient record, were then statistically recovered by correlation with a nearby precipitation station. Hence, the same statistical process was used in the correlation of gaging- station records with precipitation stations, as was used in correlation with other primary gaging - station records. Because this analysis was carried out on a 5 Dave•R Nelson - MAKAHre ort nov 17 2000.doc Page 6 calendar month basis, the best correlation for each given month could be chosen from the cadre of correlation results that were available using nearby precipitation stations. The record for the Sooes River was then recovered by correlation with the statistically recovered records of the Hoko and Dickey rivers. These extended records then provide a primary gaging - station framework through which histories for other nearby gaging stations may be recovered. The time span for these three semi - regional gaging - station flow histories is from October, 1962, to September, 1998, inclusive. With three semi- regional gaging- station histories recovered, correlation and reconstruction of secondary gaging - station records could be accomplished for a 37 year history beginning in October, 1962, and ending in September, 1998. Several of these records were of insufficient length for the monthly -basis correlation used above. Therefore, these additional records were recovered using a serial correlation of the monthly sequential flow history for the secondary station in question, with one of the reconstructed primary semi - regional histories existing for a nearby gage. This process was carried out for the East Fork of the Dickey River, and the Waatch, Ozette, and Sail rivers. Because the serial correlations were very good, these reconstructions are considered excellent in consistency with their semi - regional counterparts. These streamflow histories were then evaluated to determine the distribution of streamflow that was indicated to occur month-to -month throughout the year. Because all records used are tabulated as water year records, the results of this evaluation are given for each calendar month from October consecutively through the following September. Results of this analysis are developed for each gage as the flow duration occurring for each month throughout the year as sampled in each of the reconstructed 37 year gaging- station histories. The regional analysis also included an evaluation of monthly- average and annual- average discharge by drainage area tributary to each gage for which the gaging histories had been statistically recovered. Results of this analysis show that for drainage areas less than about 40 square miles, during the summer months of June, July, and August, the respective average monthly and annual discharge for the 37 yr gaging history is in linear proportion to drainage area. For the averages occurring in other months, and for the annual average, this linear relationship is well evidenced for drainage areas of less than about 100 square miles. Consequently, streamflow histories for smaller ungaged watershed areas less than about 40 square miles, may be resealed from the gaging - station history of a nearby gage by using the ratio of the two areas as a scaling factor applied to the flow history of that nearby gage. This process is easily accomplished for any ungaged area of interest within the northwest Olympic Peninsula. Evaluation and characterization of water availabili The analysis of gaged and ungaged watersheds was limited to drainages in the vicinity of the Makah Reservation. This detailed analysis indicates the probable recurrence intervals of monthly flows existing in flows that are available to meet demands. Results that are derived for each watershed area are given as a table for each calendar month showing the percentage of time streamflow is expected to be less than, or equal to, specified monthly flows evidenced in the period of record, 1962 to 1998. This assessment is given for the Sooes River, Waatch River, Educket Creek, Sail River, Bullman Creek, headwaters of the North Fork Sekiu River, and the Ozette River. These results are included in the appendix to this report. The anticipated peak demand on the Reservation is expected to occur during the summer months of June, July, and August. If the surface -water supply is limited in meeting the peak demand, then another supplemental source of supply must be developed to meet any shortages. The supply available to meet demands must also include accommodation of possible environmental considerations. Therefore, if the demand exceeds a specified proportion of the surface -water flow needed to protect the instream environment, diversions of water to meet the demand will be restricted. Given the potential for supply limitations in meeting demands, the cumulative supply from several watersheds must then considered rather than dependence on a single watershed as the source of supply. If, however, the demand can be met by a specific proportion of the flow from any of the watersheds considered, and not impact the water needed for possible environmental considerations in any of the streams in question, then there are no supply 6 r. . • ° Dave.R Nelson - MAKAHre ort nov 17 2000.doc Pa e 7 limitations. Evaluation of the supply is, therefore, based on an examination of expected minimum flow available to meet demands in June through August, because this is the most significant limitation in meeting the expected summer demand. Results of the analysis are summarized in Table 1, below. These results are shown for selected percentages of the time flow was expected to be equal to, or less than, the specified duration values that were indicated in the reconstructed records. The derived duration values, or % time less than or;eoual to are raw values and are not specified relative to an adjusted plotting position in the duration plot. However, allaying risk must be examined with regard to recurrence of insufficient supply. Acceptable risk regarding shortages must relate the probable recurrence of shortage to the magnitude of the shortage. The lower quartile was chosen because the indication of supply insufficiency for many of the watersheds of interest, was evident at this recurrence probability. The chance of events being considered is, therefore, very nearly 1 in 4, or less. This means that for selected monthly flows shown in the flow duration analysis, there is a probable chance that one year in four, or about 25% of the time, the monthly total flow for the month in question will be at, or below, the monthly flow volumes indicated. Streams with a chance of insufficient flow 1 year in 4, were generally noted as having insufficient flow at a more frequent probable recurrence, say 1 year in 2, or.1 year in 3, during the ciritical summer season. These results were determined for the entire summer season, May through September. Only the Sooes and Ozette rivers seem to have sufficient flow for the range of probabilities less than 1 in 4, to meet the potential June through August demand of about 124 acre-feet per month. Tablet. Expected minimum monthly flow. May through September. Values are in acre -feet. Stream or watershed name % time 5 May June July August September Sooes R bw Miller Cr 24.3 2971 1151 1442 815 2147 watershed area: 32 sq mi 10.8 2238 533 1083 684 1934 5.4 1306 505 1074 325 1766 2.7 1195 418 833 257 1453 Waatch R bw Educket Cr 24.3 864 354 .443 270. 636 - watershed area 9.96 sq mi 10.8 662 193 343 222 578 5.4 405 185 275 117 532 2.7 375 161 146 94 446 Educket Cr 24.3 249 102 128 78 183 watersed area 2.83 sq mi 10,8 191 56 99 64 167 5.4 117 53 79 34 153 2.7 108 46 42 27 129 Sail R nr Neah Bay 24.3 752 333 234 178 193 watershed area: 5.42 sq mi 10.8 656 248 182 137 119 5.4 567 233 163 112 99 2.7 496 172 119 66 86 Bullman Cr 24.3 325 133 167 102 239 watershed area: 3.74 sq mi 10.8 249 73 129 84 217 5.4 152 70 104 44 200 2.7 141 61 55 35 168 N Fk Sekiu R nr Sooes Pk 24.3 900 398 280 213 231 watershed area: 6.48 sq mi 10.8 784 297 217 164 140 5.4 678 278 195 134 118 2.7 593 206 142 79 103 7 Dave. R Nelson - MAKAHre ort nov 17 2000.doc Pa e 8 Ozette R at Ozette 24.3 8906 5699 6070 5173 7237 watershed area: 77.5 sq mi 10.8 7636 4686 5622 4324 6939 5.4 6023 4637 5204 3025 6273 2.7 5831 4485 3614 1811 5802 Dave R Nelson - MAKAHre ort nov 17 2000.doc P Discussion - Flows at the minimum duration level (percent time less than or equal to) of 2.7% are at the minimum that was either recorded, or reconstructed, for the 37 year period of record for these streams. The recurrence probability for such flows is therefore about 1 in 37. This is compared with the minimum duration level of 24.3 %, which is equivalent to a recurrence probability of about 1 in 4. Any examination of risk regarding anticipated shortages must be based on the indicated probable recurrence of flows available to meet the demand, and estimated the magnitude of, and acceptability of, any possibly occurring shortage. Without an adequate source of supply and the possibility to use a sufficiently sized reservoir for carryover storage, the margin of acceptable risk that can be tolerated in meeting shortages, is very narrow. Therefore, all of the examined watersheds must be considered together in meeting the anticipated demand expected to occur during the summer because the analysis indicates many of the individual watersheds may not be large enough to supply sufficient water. Within this consideration, the following are noted: 1. Only two streams seem to have adequate flows at the 2.7 % minimum duration level during the three month summer season of June, July; and August, to meet anticipated demands of 124 acre -feet per month. These streams are the Sooes River, and the Ozette River. 2. Educket Creek does not appear to have sufficient flow to meet, or make up a reliable portion of, the anticipated summer season demand. Data in the appendix show the indicated critical nature of summer seasonal flows for this stream. 3. Given a recurrence probability of 1 year in 4 during August, sufficient water may be available collectively in Waatch River, Sail River, and North Fork Sekiu River to meet anticipated summer demands. However, a conservative consideration of this supply with regard to environmental concerns seems to indicate that for recurrence probabilities of 1 year in 4 or lower, water from these streams, collectively, will apparently be marginally sufficient to meet demands. Water from these streams would be available to supplement water from another source, such as the Sooes River. Nevertheless, for recurrence probabilities less than 1 in 4, the sufficiency in supply available from these streams, diminishes markedly. In drier years, the flow from these watersheds may be insufficient to meet the anticipated demand and maintain environmental instream flows. These streams have suffered an extremely significant decline in anadromous fish populations, and generally would be adversely impacted by water resources developments of the kind necessary to meet the water supply needs of the Makah Indian Reservation. A further examination of the biological resources is necessary to quantify the magnitude of such impacts, and the quality of the present environment for sustaining the current resource. 4. Development of a water supply from the Ozette River does not appear to pose environmental concerns regarding diversion limitations and instream flow issues. The Ozette River has substantial flow during the summer season, that can meet anticipated demands. The Ozette River, however, should not be considered for large -scale water- resources development, even though this stream, to a much less degree than other streams in the region, has suffered a significant anadromous fish population decline within the last 50 years. Development of water from the Ozette River may require consideration of alternatives such as drawing water from Ozette Lake, as a means for diverting water from the river. This means of diversion, which is effectively the same as diverting from the Ozette River, may obviate turbidity problems that maybe encountered by diverting directly from the Ozette River. 5. Water required for instream flows and other environmental considerations will necessarily need to be quantified and the evaluation of available supply then reconsidered for each of these watersheds. Tenability of data analysis - The potential for shortages in meeting the anticipated demand of 124 acre -feet per month, for the summer 9 Dave R Nelson - MAKAHre 2a nov 17 2000.doc Pa a 10 season of June, July, and August, was examined using a flow duration analysis of the reconstructed flow histories for the selected watersheds of interest. The use of regression analysis to reconstruct missing data in these records, however, has drawbacks and limitations. Regression analysis uses a least squares procedure to fit a straight line through the scatter plot evidenced by two data sets, a primary data set, A, and a secondary data set, B. Evidence of a good correlation between corresponding values in A and B is indicated when the points that comprise the scatter plot lie close to, and may be closely approximated by, the fitted line. Missing values in B- are then computed from the equation for the line by using the. appropriate time associated value given in A. The explained variation in relating the secondary data set, B, with the variation in the primary data set, A, is usually very good when the correlation relationship is good. This means that the variability in B, for instance, or difference across the range from high to low values in B, is explained well when the correlation of B with A is good. Generally, however, when the correlation relationship declines, the explained variation declines. When there is no correlation between the data sets, , the line of relationship between A and B has simply one value equal to the average noted for the data group in B that forms the scatter plot with the concurrent data group in A. Therefore, as the explained variability declines, reconstructed values that are missing in B tend toward the average for the values originally existing in B that are, as a data group, concurrent with corresponding values -in A. This loss of information I egarding explained variability is particularly important in the assessment of extreme values, or those occurring at the high and low ends of the range in reconstructed data for B. Because seasonally related precipitation patterns become regionally inconsistent during the summer, the correlation between precipitation measured at a nearby weather station, and flows in adjacent watersheds, may not be well correlated for summer monthly flows. This problem was addressed by using the cadre of available correlations between recorded monthly streamflows at nearby stream gages, and recorded monthly precipitation at regionally close precipitation stations. By using this procedure, the best correlations could be examined, modified as necessary, and used in optimally reconstructing the missing portions of the records. Even so, the best explained variability in summer monthly flows was still generally marginal. This means information was unrecoverable regarding the magnitude of extreme values in the reconstructed flow histories for the summer months. Given this consideration, reconstructed flows that are indicated as less than the estimated average during each of the summer months may be higher than would have been measured if the stream in question had been gaged. -1n other words, actually gaged monthly summer flows less than the average given in the flow duration assessment, will probably be lower than the corresponding flow values given in the assessment. Confidence limits for the reconstructed values used in the assessment, were not determined. As a final consideration, regression analysis was used in an unconventional manner to develop extreme values in the reconstructed records. Conventional use of regression analysis limits reconstruction of missing values for the secondary data set to those that may be derived within the range of variation noted in the data group comprising the correlated data of the primary data set. In many instances for the present assessment, extreme values were, or may have been, reconstructed from values present in the primary data set that were outside the range of variation evidenced in correlation with the secondary data set. Therefore, the reconstruction of the missing extreme values under this circumstance was unrestricted, as the procedure used was, or may have been, in violation of the accepted convention limiting the range of data to be used in such data reconstructions. However, experience has shown that cautious utilization of this unconventional technique will not adversely affect the determination and use of estimated extreme values derived from the correlation analysis of such hydrologic records. Conclusion - Water uses on the Makah Indian Reservation are provided from surface -water streams and ground -water wells that are located on the reservation. These ground -water wells are typically located in alluvium associated with a surface -water stream. During the summer months, surface water stored in Educket Creek Reservoir is of such poor quality that ground water is used to alleviate shortages. However, with existing facilities and limited ground -water reserves, ground water cannot be delivered in sufficient quantity to meet the entire present demand, and surface water for the supply is primarily developed from an infiltration 10 D ve R Nelson MAKAHre ort nov 17 2000.doc Pa a 11 gallery in the Waatch River. Presently, the demand for water in the summer is sufficiently great that the demand exceeds the delivery capacity of the water treatment plant. Summer season water uses during June, July, and August, on the Makah Indian Reservation are anticipated to require a peak delivery capacity of about 900 gallons per minute, or 124 acre -feet per month, which is in excess of the currently available rate of supply for this time of year. An assessment of statistically reconstructed stream gaging - station records, and of watershed discharges that had been estimated from these gaging - station records, indicates that individual watersheds of the Makah Indian Reservation, other than the Sooes and Ozette rivers, are likely to be insufficient in providing water at the anticipated demand rate of 124 acre -feet per month for the summer season. Collectively, however, a supply may be available from streamflow developed in the watersheds of the Waatch River, Sail River, and North Fork Sekiu River. This supply may be able to meet the anticipated demand during summer season, or be used to supplement water diverted from another source. However, during drier periods, available streamflow from these watersheds will likely be limited in meetin g the anticipated summer season demand while maintaining flows necessary for the environment. As an alternative, the Sooes River, and particularly the Ozette River, seem to be adequate to supply the anticipated summer demand even during the driest summers that may occur. Only the Ozette River appears to have sufficient flow to meet the anticipated demand without impacting adversely any environmental issues, or concerns. Water required for instream flows and other environmental considerations will necessarily need to be quantified and the evaluation of available supply then reconsidered for each of these watersheds. • Dave RNelson - MAKAHre ort nov 17 2000.doc Pa a 12 Biblioeranhy - The following technical reference was cited in this report: Reid, L K, Caroon, L. E., and Pyper, G. E., 1968, Extensions of streamflow records in Utah: Utah Department of Natural Resources Technical Publication no. 20. The following references were consulted in developing this report: Babcock, R. S., Suczek, C. A., and Engebretson, D. C., 1994, The Crescent "Terrane ", Olympic Peninsula and Southern Vancouver Island, in Raymond Lasmanis and Eric S. Cheney, convenors, Regional Geology of Washington State: Washington Department of Natural Resources Division of Geology and Earth Resources. Booth, Derek B., and Goldstein, Barry, 1994, Patterns and processes of landscape development by the Puget Lobe Ice Sheet, in Raymond Lasmanis and Eric S. Cheney, convenors, Regional Geology of Washington State: Washington Department of Natural Resources Division of Geology and Earth Resources. Bortleson, G. C., and Dion, N. P., 1979, Preferred and observed conditions for Sockeye salmon in Ozette Lake and its tributaries, Clallam County, Washington: U. S. Geological Survey Water - Resources Investigations 78 -64 (Tacoma, WA). Clallam County Profile, 1992: County Commissioners, and Department of Community Development, Clallam County, Washington. Dion, N. P., Waiters, K. L., and Nelson,. L. M., 1980, Water resources of the Makah Indian Reservation, Washington: U. S. Geological Survey Water - Resources Investigations 80 -15 (Tacoma, WA). Langbein, W. B., and Hardison, C. H., 1956, Extending streamflowdata: Amer. Soc. Civil Egr. Proceedings, v. 81, paper no. 826. Project Summery, 1996 (August), Community Water Source Improvements on the Makah Indian Reservation, Clallam County, Washington: Project PO -96 -684, 22p. Project Summery, 1999 (September), Community Water Source Improvements on the Makah Indian Reservation, Clallam County, Washington: Project PO -99 -751, 13p. Snavely, P. D., MacLeod, N. S., Niem, A. R., and Minasian, D. L., 1986, Geologic Map of the Cape Flattery Area, Northwestern Olympic Peninsula, Washington: U. S. Geological Survey Open -File Report 86 -344B (Menlo Park, CA) Soil Survey of Clallam County Area, Washington, 1987: United States Department of Agriculture, Soil Conservation Service. Tabor, R. W., and Cady, W. M., 1978, Geologic map of the Olympic Peninsula, Washington: U. S. Geological Survey Map I -994.