Streamflow forecasting is a fundamental component of various water resources management systems, ranging from flood control and mitigation to long-term planning of irrigation and hydropower systems. In the context of floods, a probabilistic forecasting system is required for proper and effective decision-making. Therefore, the primary goal of this research is the development of an advanced ensemble-based streamflow forecasting framework to better quantify the predictive uncertainty and generate enhanced probabilistic forecasts. This research started by comprehensively evaluating the performances of various lumped conceptual models in data-poor watersheds and comparing various Bayesian Model Averaging (BMA) modifications for probabilistic streamflow simulation. Then, using the concept of BMA, two novel probabilistic post-processing approaches were developed to enhance streamflow forecasting performance. The combination of the entropy theory and the BMA method leads to an entropy-based Bayesian Model Averaging (En-BMA) approach for enhanced probabilistic streamflow and precipitation forecasting. Also, the integration of the Hydrologic Uncertainty Processor (HUP) and the BMA methods is proposed for probabilistic post-processing of multi-model streamflow forecasts. Results indicated that the MACHBV and GR4J models are highly competent in simulating hydrological processes within data-scarce watersheds, however, the presence of the lower skill hydrologic models is still beneficial for ensemble-based streamflow forecasting. The comprehensive verification of the BMA approach in terms of streamflow predictions has identified the merits of implementing some of the previously recommended modifications and showed the importance of possessing a mutually exclusive and collectively exhaustive ensemble. By targeting the remaining limitation of the BMA approach, the proposed En-BMA method can improve probabilistic streamflow forecasting, especially under high flow conditions. Also, the proposed HUP-BMA approach has taken advantage of both HUP and BMA methods to better quantify the hydrologic uncertainty. Moreover, the applicability of the modified En-BMA as a more robust post-processing approach for precipitation forecasting, compared to BMA, has been demonstrated. / Thesis / Doctor of Philosophy (PhD) / Possessing a reliable streamflow forecasting framework is of special importance in various fields of operational water resources management, non-structural flood mitigation in particular. Accurate and reliable streamflow forecasts lead to the best possible in-advanced flood control decisions which can significantly reduce its consequent loss of lives and properties. The main objective of this research is to develop an enhanced ensemble-based probabilistic streamflow forecasting approach through proper quantification of predictive uncertainty using an ensemble of streamflow forecasts. The key contributions are: (1) implementing multiple diverse forecasts with full coverage of future possibilities in the Bayesian ensemble-based forecasting method to produce more accurate and reliable forecasts; and (2) developing an ensemble-based Bayesian post-processing approach to enhance the hydrologic uncertainty quantification by taking the advantages of multiple forecasts and initial flow observation. The findings of this study are expected to benefit streamflow forecasting, flood control and mitigation, and water resources management and planning.

Doctor of Philosophy / Department of Biological & Agricultural Engineering / James K. Koelliker / Over the past decades, it has been observed that the streamflow characteristics in the Great Plains rivers have substantially changed. These changes have affected and will continue to affect the management decisions within the watershed. This study was undertaken to document the changes for some unregulated streams in Kansas, characterize the streams in terms of some hydrologic indices, and identify the probable factors influencing the changes. Fourteen unregulated streams with 60 or more years of daily discharge data geographically distributed across the state were used. The analysis focused on hydrologic indices judged to be relevant to the lotic ecosystem. The state was divided into four regions, representing roughly the northwest, southwest, northeast and southeast sections of the state. Log Pearson III method was used for computing flow probabilities, Mann-Kendall test in conjunction with Sen’s slope estimator was used for trend analysis, whereas the indicators of hydrologic alterations software was used to generate most hydrologic indices. Several factors believed to affect the streamflow were identified, and their influence was modeled over time. A multi-variate statistical model was run. Results show that there is substantial difference in the streamflow characteristics between the western and eastern regions. Many streamflow aspects have changed over time, and a number of them show significant and important change. Most streams in western Kansas have longer and more frequent dry periods. Potential recharge rate, land use, water use, soil and water conservation practices, and soil type were significant factors influencing the median to very low flow, but the effect varied among the regions. Results of this study could be useful to decision makers, water users, watershed stakeholders, and environmental conservation advocates in addressing problems and concerns related to stream and river management.

Published Article / Early indication of possible drought can help in developing suitable drought mitigation strategies and measures in advance. Therefore, drought forecasting plays an important role in the planning and management of water resource in such circumstances. In this study, a non-linear streamflow forecasting model was developed using Artificial Neural Network (ANN) modeling technique at the Melka Sedi stream gauging station, Ethiopia, with adequate lead times. The available data was divided into two independent sets using a split sampling tool of the neural network software. The first data set was used for training and the second data set, which is normally about one fourth of the total available data, was used for testing the model. A one year data was set aside for validating the ANN model. The streamflow predicted using the model on weekly time step compared favorably with the measured streamflow data (R2 = 75%) during the validation period. Application of the model in assessing appropriate agricultural water management strategies for a large-scale irrigation scheme in the Awash River Basin, Ethiopia, has already been considered for publication in a referred journal.

A finite element numerical model is developed for the modeling of coupled surface-water flow and ground-water flow. The mathematical treatment of subsurface flows follows the confined aquifer theory or the classical Dupuit approximation for unconfined aquifers whereas surface-water flows are treated with the kinematic wave approximation for open channel flow. A detailed discussion of the standard approaches to represent the coupling term is provided. In this work, a mathematical expression similar to Ohm's law is used to simulate the interacting term between the two major hydrological components. Contrary to the standard approach, the coupling term is incorporated through a boundary flux integral that arises naturally in the weak form of the governing equations rather than through a source term. It is found that in some cases, a branch cut needs to be introduced along the internal boundary representing the stream in order to define a simply connected domain, which is an essential requirement in the derivation of the weak form of the ground-water flow equation. The fast time scale characteristic of surface-water flows and the slow time scale characteristic of ground-water flows are clearly established, leading to the definition of three dimensionless parameters, namely, a Peclet number that inherits the disparity between both time scales, a flow number that relates the pumping rate and the streamflow, and a Biot number that relates the conductance at the river-aquifer interface to the aquifer conductance. The model, implemented in the Bill Williams River Basin, reproduces the observed streamflow patterns and the ground-water flow patterns. Fairly good results are obtained using multiple time steps in the simulation process.

Uncertainty surrounds the understanding of natural variability in hydrologic extremes such as droughts and floods and how these events are projected to change in the future. This thesis leverages Global Climate Model (GCM) data to analyse 738 year streamflow scenarios in the Nelson-Churchill River Basin. Streamflow scenarios include a 500 year stationary period and future projections forced by two forcing scenarios. Fifty three GCM simulations are evaluated for performance in reproducing observed runoff characteristics. Runoff from a subset of nine simulations is routed to generate naturalized streamflow scenarios. Quantile mapping is then applied to reduce volume bias while maintaining the GCM’s sequencing of events. Results show evidence of future increases in mean annual streamflow and evidence that mean monthly streamflow variability has decreased from stationary conditions and is projected to decrease further into the future. There is less evidence of systematic change in droughts and floods. / May 2016

Thesis (MScEng)--University of Stellenbosch, 2004. / ENGLISH ABSTRACT: This work investigates methods to theoretically determine the lateral velocity distribution across a river from which factors may be derived to translate a single point velocity into average velocity for the river as a whole. A wide range of field data from vanous nvers was analysed. This produced over a hundred velocity distributions with which to compare theoretical distribution results. Four theoretical approaches were considered: the one-dimensional method (Manning's equation), a two-dimensional flow formula solved as an initial-value-problem, a two dimensional flow formula solved as a boundary-value-problem and an empirical method developed from energy principles. The one-dimensional and initial-value-problem approaches were unsuccessful. The boundary-value and empirical approach did however produce promising results. Surprisingly the analysis of the field data revealed patterns of similarity which could produce accurate results without the need of a theoretical approach. / AFRIKAANSE OPSOMMING: Metodes word ondersoek om teoreties 'n laterale snelheidsverspreiding oor 'n rivier te bepaal en sodoende 'n faktor te vind waarmee 'n gemete enkelpuntsnelheid in die rivier omgeskakel kan word na 'n gemiddelde snelheid vir die rivier in geheel. Vloeimeetdata van verskeie nviere IS geanaliseer. Sodoende is meer as 100 snelheidsverspreidings gegenereer waarmee die teorie vergelyk kon word. Vier teoretiese benaderings is gevolg: Manning se een-dimensionele vloeivergelyking, 'n twee dimensionele vloei-vergelyking opgelos met behulp van 'n beginwaarde, 'n twee dimensionele vloei-vergelyking opgelos met behulp van randwaardes, en 'n empiriese metode ontwikkel vanuit energie beginsels. Die een-dimensionele- en beginwaarde-benaderings was me suksesvol me. Die randwaarde- en empiriese benaderings het wel belowende resultate gelewer. Selfs verwerking van die gemete stroommetings het waardevolle inligting gelewer: daar bestaan duidelike ooreenkomste in die snelheidsverspreidings wat gebruik kan word om die verspreidingsfaktor mee te bereken sonder om teoretiese oplossings te soek.

Thesis (MScEng)--University of Stellenbosch, 2005. / ENGLISH ABSTRACT: This study sets out to describe the incipient motion process in cobble/boulder bed rivers in terms of the unit applied power approach. This objective has been met through the collection of data on stone movement from a total of thirteen flood events observed in two undisturbed rivers in the Western Cape, namely the Molenaars and Berg Rivers. The data were plotted on the original modified Liu diagram for incipient motion and it was found that the threshold of movement for the smaller stones did not conform with the portion of the Liu diagram which represents fully turbulent flow (i.e. where ~ gDs tv: = 0.12 for Re. >13). It was concluded that the only reason that could explain this deviation is the fact that the original modified Liu diagram had been derived for uniform particle size beds while the data of the Molenaars and Berg Rivers represent non-uniform particle size beds. This was proved through re-deriving the y-axis function of the original modified Liu diagram to include a factor that makes provision for the roughness of a non-uniform particle size bed. It was found that the average absolute roughness of the non-uniform particle size beds in the Molenaars and Berg Rivers is reasonably well approximated by a value of k=da. Design curves (in terms of the original modified Liu diagram parameters) for intensity of motion were also produced. Although it is not possible to read off accurate percentages of movement values directly from these curves, it should be possible to deduce reasonably accurate values in practical situations. / AFRIKAANSE OPSOMMING: Die studie poog om die begin van beweging proses in klip/rotsbed riviere te beskryf in terme van die eenheids aangewende drywing metode. Hierdie doel is bereik deur die insameling van klipbewegingdata in 'n totaal van dertien vloedgebeurtenisse in twee ongerepte riviere in die Wes-Kaap, naamlik die Molenaars en Bergriviere. Die data was geplot op die oorspronklike aangepaste Liu diagram vir begin van beweging en dit was gevind dat die grens van beweging vir die kleiner klippe nie ooreenstem met die gedeelte van die Liu diagram wat volle turbulente vloei verteenwoordig nie (waar ~ gDs tv; = 0.12 vir Re. >13). Daar was tot die gevolgtrekking gekom dat die enigste rede wat die afwyking kan beskryf is die feit dat die oorspronklike aangepaste Liu diagram afgelei was vir uniforme partikelgrootte beddens terwyl die data van die Molenaars and Bergriviere nie-uniforme partikelgrootte beddens verteenwoordig. Dit was bewys deur die heratleiding van die y-as funksie van die oorspronklike aangepaste Liu diagram om 'n faktor in te sluit wat voorsiening maak vir die ruheid van 'n nie-uniforme partikelgrootte bed. Dit was gevind dat die gemiddelde absolute ruheid van die nieuniforme partikelgrootte beddens in die Molenaars en Bergriviere word redelik goed benaderd met 'n waarde van k=ds4• Ontwerpkurwes (in terme van die oorspronklike aangepaste Liu diagram parameters) vir intensiteit van beweging was ook ontwikkel. Alhoewel dit nie moontlik is om baie akkurate persentasies van beweging af te lees van die kurwes nie, is dit moontlik om akkuraat genoeg waardes te verkry in praktiese situasies.

Thesis (MEng)--University of Stellenbosch, 2001. / ENGLISH ABSTRACT: This report deals with the use of Doppler meters to measure flow velocities and hence discharges in streams. The Doppler meter measures the shift in frequency of an acoustic wave, which it emits and then becomes reflected by a moving particle. The reading is converted into a velocity by dividing the shifted frequency by a calibration constant. The particles that reflect the signal need to follow the flow sufficiently closely so that their velocity may be assumed equal to the flow velocity. A previous study on the use of the Doppler meter at a Crump weir (Du Toit and Venter, 1999) indicated that velocities measured with a Doppler meter showed a distinct relationship with recorded water levels. However, the wide scatter of the observed frequencies in this study, necessitated further tests on the use of the Doppler meter at measuring structures as well as calibration tests on the instrument in the hydraulic laboratory of the University of Stellenbosch. The mam objective of this investigation was to establish the relationship between measured Doppler velocities at a Crump weir and the approach velocities in the stream. The instrument was to be tested in both modular and non-modular flow ranges. In addition, the instrument had to be calibrated in the hydraulic laboratory under varying flow conditions, such as very low flow velocities and different sediment concentrations. The placement of the probe at different depths of the flow was also investigated to comment on the accuracy of the Doppler readings at these depths. The results of these tests should serve as guidelines for any additional tests required for use of this instrument in open channel discharge measurements. The Doppler meter used for this study was supplied and manufactured in Stellenbosch by Flotron, and is being marketed as DFM-P-067. It was calibrated in the laboratory in a channel with limited width and hence non-two-dimensional flow conditions. Conclusions were drawn on the calibration constant that was established. The calibration of the instrument requires the division of the cross-sectional flow area into a number of sub-divisions over which the flow was integrated. The calibration constant of 1460 established in this study differs by approximately 6 percent from the theoretical constant value of 1375. The sensitivity of the Doppler meter to different sediment concentrations was also investigated. For the instrument to read a shifted frequency, it is essential that suspended particles that follow the water movement sufficiently closely are present in the stream. It was observed that readings of the instrument in "sediment-free" water differed only by 3.6% from the readings taken in water containing sediments. The instrument was thus not very sensitive to different sediment concentrations. It was also found that the angle at which the probe was placed in the water had no effect on the accuracy of the observed Doppler velocity. It was furthermore found that the Doppler meter worked reliably at all depths, including levels very close to the channel floor and levels just below the water surface. One drawback of the apparatus was the minimum velocity that it can measure accurately. This minimum velocity of 0.046 mis does not compare well with that for other commercially available Doppler meters. The Argonaut-Acoustic Doppler meter for example can measure velocities as low as O.OOOlm/s, meaning that the DFM-P-067 measures a minimum velocity 460 times swifter than the minimum velocity of the Argonaut-Acoustic Doppler meter. After the Doppler meter had been calibrated, it was tested at a Crump weir in the laboratory to determine the relationship between the Doppler velocities, measured at the weir's crest, and the velocities in the approach channel. These tests were performed for both modular and non-modular flow conditions. The report concludes that, within the flow range in which the instrument was tested, there is a linear relationship between the two velocities mentioned. It is likely that the results obtained in the modular flow range can be used to extrapolate for high flows, especially for submergence ratios less than 0.93. The wide scatter of results obtained in the previous study was due to the readings not being averaged. The Doppler meter does not measure a point velocity but an average velocity within the acoustic field that it emits. This acoustic field is very small and depends on the geometry of the probe. Finally it is recommended that the linear relationship in the non-modular flow range be investigated further in a larger model, where the submergence ratio can be better controlled. The Doppler meter should in future also be calibrated in a wide channel in which two-dimensional flow conditions are approached and these results should be compared to the results obtained in this study. Every instrument is expected to have its own calibration constant, and depending on its application, it can either be calibrated at a weir or in the laboratory. The calibration of the instrument at a Crump weir should allow for a wider range of flows, and also very low flow velocities. At the end of this report guidelines were drawn up that are based on the results and conclusions obtained in this investigation. They may serve as an aid for measurements that could be carried out with this instrument in open channels. / AFRIKAANSE OPSOMMING: Hierdie verslag handeloor die gebruik van die Doppler-meter om vloeisnelhede en derhalwe die vloeitempos in riviere te meet. Die Doppler meter word gebruik om die verandering in die frekwensie van 'n akoustiese golf wat deur bewegende deeltjies in die water gereflekteer word te meet. Die lesing word dan omgeskakel in 'n snelheid deur die gewysigde frekwensie deur 'n kalibrasie konstante te deel. Die bewegende deeltjies wat die sein reflekteer, volg die vloei genoegsaam sodat aanvaar kan word dat hulle snelhede gelyk aan die vloeisnelheid is. 'n Vorige studie in die gebruik van die Doppler meter by 'n Crump meetwal het baie belowende resultate getoon deurdat daar gevind is dat die gemete Doppler snelheid 'n duidelike verwantskap toon met veranderings in gemete water vlakke. As gevolg van die wye band in die waargenome frekwensies in die studie is aanbeveel dat verdere toetse op die gebruik van die Doppler meter by meetstasies gedoen moet word. Die instrument moet ook in die laboratorium gekalibreer word. Die hoofdoel van hierdie ondersoek was om die verwantskap tussen die gemete Doppler snelhede by 'n Crump meetwal en die aankomssnelhede in die stroom te bepaal. Dit moes gedoen word in beide die modulêre en niemodulêre vloeibestekke. Behalwe vir die kalibrasie van die instrument in die laboratorium moes die betroubaarheid daarvan onder verskillende vloei toestande ook getoets word, soos byvoorbeeld by lae vloei snelhede en by verskillende sediment konsentrasies. Die instrument is ook op verskillende vlakke binne die vloei getoets om te bepaal of daar op hierdie vlakke betroubare lesings verwag kon word. Resultate verkry, kan dan dien as riglyne vir enige verdere toetse wat nog op die instrument in oop kanale uitgevoer moet word. Die Doppler meter wat vir die ondersoek gebruik is, word in Stellenbosch vervaardig deur Flotron en word onder die naam DFM-P-067 bemark. Dit is in die laboratorium in 'n kanaal met 'n beperkte breedte getoets en IS daarom in nie-twee dimensionele vloei gekalibreer. Gevolgtrekkings IS gebaseer op die kalibrasie konstante verkry uit die toetse. Die kalibrasie van die instrument vereis dat die deursnee area van die vloei in verskeie segmente onderverdeel moes word. Die kalibrasie konstante van 1460 bepaal in hierdie studie verskilongeveer 6% van die teoretiese waarde van 1375 vir die konstante. Die Doppler meter se sensitiwiteit vir verskillende sediment konsentrasies is ook ondersoek. Dit is noodsaaklik dat daar gesuspendeerde deeltjies teenwoordig in die water is en dat die deeltjies saam met die water beweeg om te verseker dat die instrument die gewysigde frekwensie kan registreer. Daar is egter gevind dat die lesings van die instrument in sediment-vrye water slegs met 3,6% verskil van lesings wat in water met sediment geneem is. Dit lei tot die gevolgtrekking dat die instrument nie baie sensitief vir veranderlike sediment konsentrasies in die water is nie. Daar is ook gevind dat die hoek waarteen die sender in die water geplaas word nie die akkuraatheid van die Doppler snelhede beinvloed nie. Verder is gevind dat die Doppler meter bevredigende resultate lewer, ongeag op watter diepte lesings geneem word. Tydens toetse is waarnemings baie nabyaan die kanaal bodem asook nabyaan die water se oppervlak gedoen. 'n Tekortkoming van hierdie instrument is die minimum snelheid wat dit akkuraat kan meet. Daar is gevind dat die Doppler meter se muurnum snelheid lesing van 0.046 mis nie goed vergelyk met dié van ander meters wat kommersieël beskikbaar is nie. Die Argonaut-Acoustic Doppler meter kan byvoorbeeld vloeisnelhede so laag as 0.0001 mis meet wat beteken dat die DFM-P-067 se minimum betroubare vloeisnelheid 460 keer vinniger is as die Argonaut-Acoustic Doppler meter se minimum betroubare vloeisnelheid. Nadat die Doppler meter gekalibreer is, is dit by 'n Crump meetwal in die laboratorium getoets om die verhouding tussen die Doppler snelhede gemeet by die oorloopkruin en die snelhede wat in die aanloopkanaal gemeet is, te bepaal. Hierdie toetse is uitgevoer op beide modulêre en nie-modulêre vloei toestande. Daar is gevind dat daar binne die vloeibestek waarin die toetse plaasgevind het 'n liniêere verband tussen die twee bogenoemde snelhede bestaan. Dit is hoogs waarskynlik dat die resultate wat in die modulêre vloeibestek gevind is gebruik kan word om vir hoë vloeie te ekstrapoleer, veral vir grade van versuiping laer as 0.93. Die vorige studie se uiteenlopende resultate kan toegeskryf word aan lesings waarvan die gemiddelde lesing vir 'n spesifieke vloeitoestand nie bepaal is nie. Die Doppler meter meet nie 'n bepaalde punt-snelheid nie, maar 'n gemiddelde snelheid binne die akoestiese veld wat dit uitstraal. Hierdie akoestiese veld is baie klein en afhanklik van die geometrie van die sender. Ten slotte word aanbeveel dat die lineêre verband in die nie-modulêre vloeibestek in 'n groter model, waar die graad van versuiping makliker beheerbaar is, verder ondersoek moet word. Die Doppler meter moet ook in 'n breë kanaal waarin twee dimensionale vloei voorkom, gekalibreer word. Resultate so verkry moet vergelyk word met die wat in hierdie studie behaal is. Elke instrument behoort sy eie kalibrasie konstante te hê en afhangende van waar dit gebruik word, kan dit of by 'n meetwal of in die laboratorium gekalibreer word. Die kalibrasie van die instrument by 'n Crump meetwal behoort 'n wyer reeks vloeie toe te laat met ook baie lae snelhede. Die verslag word afgesluit met riglyne gebaseer op die resultate en gevolgtrekkings wat uit die ondersoek voortgespruit het. Hierdie riglyne en gevolgtrekkings kan dan dien as 'n hulpmiddel vir metings wat met hierdie instrument in oop kanale uitgevoer word.

<p>Land use/cover has shown significant changes during the past three decades in Ethiopia especially in the highlands of the country. That resulted in changes in streamflows and other hydrological processes. The existing land and water resources system of the area is adversely affected due the rapid growth of population, deforestation, surface erosion and sediment transport. The main objective of this study is to evaluate the impact of land use/cover changes in the hydrology of <em>Beles</em> Basin, Ethiopia. The physically based hydrologic model, SWAT, was developed for the <em>Beles</em> basin, Ethiopia by combining geospatial and climatic data. ArcGIS has been used to process geospatial data which includes the Digital Elevation Model (DEM) which has a resolution of 90 m, land use/cover and soil maps. A simple Interpolation technique has been used to fill in the missing precipitation data. The GIS interface version of SWAT (ArcSWAT) has the capability to utilize ArcGIS to facilitate input data preparation and output data generation. Idrisi Andes in cooperation with ArcGIS 9.2 used to generate landuse/cover maps from Landsat data of three different years. Three SWAT models were set up using the three generated land use/cover maps and used to evaluate the land use/cover change and its impacts on the streamflow of study basin. The primary hydrological model was evaluated through sensitivity analysis, model calibration, and model validation for realistic prediction of the different hydrological components in the basin. Out of twenty six flow parameters sixteen parameters were found to be sensitive. But the most sensitive ten parameters were selected and used for model calibration. The model calibration was carried out using observed streamflow data from 01 January 2001 to 31 December 2002 and a validation period from 01 January 2003 to 31 December 2004. The coefficient of determinations (<em>R2</em>) was 0.74 and the Nash-Sutcliffe simulation efficiency (<em>NSE</em>) was 0.62which indicated that the model was able to predict streamflow with reasonable accuracy. However, the hydrograph of the cumulative hydrographs of the calibration and validation periods showed significant discrepancies between the observed and the simulated data of each period.  The average yearly flow volume of the observed streamflow on the cumulative hydrograph of the calibration period has exceeded the simulated streamflow. On the other hand on the cumulative hydrograph of the validation period the average yearly flow volume of the simulated streamflow was higher than the observed streamflow. The simulated result of the streamflow data from different land use/cover maps revealed that the change in the land use/cover classes of the basin throughout the study periods.</p> / QC 20100707

The growing demand for water in the arid regions of the West increases the need for optimal allocation of water among competing uses. An efficient allocation of water between instream and out-of-stream uses has been impeded by institutional constraints and the scarcity of information regarding instream flow benefits. The objectives of this thesis were to provide preliminary economic data on the value of instream water in "producing" recreational fishing and to examine the effect of forestry, agriculture, and livestock practices on temporal streamflow patterns and anadromous fish production. The steelhead trout (Salmo gairdneri) sport fishery within the John Day River basin in north-central Oregon provided the setting for this research. The interdisciplinary methodology employed in estimating the marginal value of water with respect to steelhead production consisted of two tasks. The first task involved valuing a marginal change in the quality of the steelhead recreational fishery. The contingent valuation method (CVM) was selected for this purpose. Both open- and closed-ended willingness-to-pay (WTP) questions were included in a questionnaire administered to John Day River steelhead anglers during the 1986/87 steelhead fishing season. Survey data were analyzed to arrive at individual and aggregate bid functions relating WTP to expected angling success rates. Results indicate that, under current conditions, the average angler is willing to pay approximately $7.20 to catch an additional steelhead. The second task of the instream water valuation methodology was directed at deriving a streamflow/steelhead production relationship. By including variables influencing steelhead production in a Ricker stock-recruitment model, it was possible to develop a model which could be estimated using linear regression techniques. Some difficulty arose, however, with interpretation of the model due to the unavailability of cohort escapement data and the subsequent use of standing crop data. While possibly masking the true magnitude of streamflow's effect on fish production, this drawback was not deemed limiting within the general context of the interdisciplinary methodology. Results of the biological model conformed to a priori expectations. Increases in summer and winter streamflows led to increased steelhead survival, whereas higher spring flows increased mortality levels. Other results indicate that the John Day Dam was responsible for a 31.5 percent decline in the population index for the 1969-1983 period. Combining the economic and biological results into one equation yielded an estimate of the marginal value of summer instream water in "producing" recreational steelhead angling. Similar equations were developed for winter and spring flows. The marginal value of water in producing recreational steelhead fishing within the John Day basin was estimated at $0.56 per acre-foot for summer flows, $0.046 for winter flows, and -$0.075 for spring flows. By including out-of-basin benefits, these values increased to $2.26, $0.19, and -$0.30, respectively. In comparison, water's value in irrigation within the John Day basin has been estimated at between $10 to $24 per acre-foot. However, nonuse values of steelhead, as well as the increased production of other fish species (such as spring chinook salmon) were not included in the instream water values. In addition, no attempt was made at valuing instream water's contribution to boating, camping, or other benefit-producing activities. A secondary objective of this thesis was to briefly examine the possible benefits accruing to other instream and out-of-stream users due to an alteration in streamflow patterns. In addition, the impact of activities by other resource users -- namely forestry, agriculture, and livestock production --on anadromous fish production was reviewed. Improper management practices by these activities can negatively impact the aquatic and riparian ecosystems. While no firm conclusions were drawn, it appears the quality of these ecosystems, as opposed to the amount of streamflow, has the largest marginal impact on anadromous fish populations. / Graduation date: 1988