Streamflow data is used throughout the world in applications such as flooding, agriculture, and urban planning. Understanding daily and seasonal patterns in streamflow is important for decision makers, so that they can accurately predict and react to seasonal changes in streamflow for the region. This understanding of daily and seasonal patterns has historically been achieved through interpretation of observed historical data at stream reaches throughout the individual regions. Developing countries have limited and sporadic observed stream and rain gage data, making it difficult for stakeholders to manage their water resources to their fullest potential. In areas where observed historical data is not readily available, the European Reanalysis Interim (ERA-Interim) data provided by the European Center for Medium-Range Weather Forecasts (ECMWF) can be used as a surrogate. The ERA-Interim data can be compared to historic observed flow to determine the accuracy of the ERA-Interim data using statistical measures such as the correlation coefficient, the mean difference, the root mean square error, R2 coefficients and spectral angle metrics. These different statistical measures determine different aspects of the predicted data's accuracy. These metrics measure correlation, errors in magnitude, errors in timing, and errors in shape. This thesis presents a suite of tests that can be used to determine the accuracy and correlation of the ERA-Interim data compared to the observed data, the accuracy of the ERA-Interim data in capturing the overall events, and the accuracy of the data in capturing the magnitude of events. From these tests, and the cases presented in this thesis, we can conclude that the ERA-Interim is a sufficient model for simulating historic data on a global scale. It is able to capture the seasonality of the historical data, the magnitude of the events, and the overall timing of the events sufficiently to be used as a surrogate dataset. The suite of tests can also be applied to other applications, to make comparing two datasets of flow data a quicker and easier process.

A major challenge in freshwater ecosystem management is to predict future changes in streamflow regime. This thesis focused on identifying and modeling specific characteristics of streamflow that are important to stream ecosystems. The need to evaluate the potential impacts of climate change on stream ecosystems makes it important to study how streamflow regime may change. In this thesis we sought to advance understanding of the effect of climate change on streamflow regime by (1) examining the spatial variation in streamflow attributes across the continental US, (2) modeling how these streamflow attributes vary with current climate and watershed features, and (3) using this model with future climate projections of changes in precipitation and temperature to predict how streamflow attributes change with climate change. We used long-term daily flow measurements for 601 gauged streams whose watersheds were in relatively unimpaired condition to characterize streamflow regimes. Sixteen streamflow variables were identified which in our judgment sufficiently characterized aspects of the streamflow regime most relevant to stream ecosystem structure and function. These are computed for each stream. Principal component analysis with Varimax rotation reduced the dimensionality to five uncorrelated streamflow factors that quantify lowflow, magnitude, flashiness, timing and constancy. These independent factors were used to hereafter classify the streams based on distances in factor space into three broad classes which were further divided into eight classes. We used Random Forests to develop a model to predict these stream classes using watershed and climate attributes. The model had an accuracy of about 75%. Downscaled climate projections of precipitation and temperature were used to predict the changes in these stream classes by 2100 using the RF model. Thirty-three percent of selected sites were predicted to change into a different stream class by 2100. The least changes were predicted in snow-fed streams in the west while most of changes were predicted for rain-fed small perennial streams and intermittent streams in the central and eastern US. Class changes predicted, due to projected climate change provide a basis for (i) considering the extent of projected changes and (ii) formulating approaches to protect ecosystems that may be subject to change.

To meet the surging water demand due to rapid population growth and changing climatic conditions around the world, and to reduce the impact of floods and droughts, comprehensive water management and planning is necessary. Climatic variability, hydrologic uncertainty and variability of hydrologic quantities in time and space are inherent to hydrological modeling. Hydrologic modeling using a physically-based model can be very complex and typically requires detailed knowledge of physical processes. The availability of data is an important issue to justify the use of these models. Data-driven models are an alternative choice. This is a relatively new and efficient approach to modeling. Data-drive models bridge the gap between the classical regression and physically-based models. By using a data-driven model that relies on the machine learning approach, it is possible to produce reasonable predictions from a limited data set and limited knowledge of underlying physical processes of the system by just relating input and output. This dissertation uses the Multivariate Relevance Vector Machine (MVRVM) and Support Vector Machine (SVM) for predicting a variety of hydrological quantities. These models are used in this dissertation for identifying influential climate indicators, and are used for long-term streamflow prediction for multiple lead times at different locations in Utah. They are also used for prediction of Great Salt Lake (GSL) elevation series. They provide reasonable predictions of hydrological quantities from the available data. The predictions from these models are robust and parsimonious. This research presents the first attempt to identify influential climate indicators and predict long lead-time streamflow in Utah, and to predict lake elevation using machine learning models. The approach presented herein has potential value for water resources planning and management especially for irrigation and flood management.

Bryan Tikalsky Department of Geography Master of Science Mountain water resources are essential to those living along the Salt Lake City urban corridor. Water resource planners base their policy on twentieth century climate conditions and streamflow records. Often these records only account for a small amount of the natural variability in streamflow and climate. By utilizing dendrochronology this study seeks to better understand variability of streamflow in the Jordan River Drainage Basin over the last 828 years. A GIS model was used to identify potential sampling sites where tree growth would be sensitive to climate and factors affecting stream run-off. Over eighty samples from ancient limber pine (Pinus flexilis) and Douglas-fir (Pseudotsuga menziesii) were obtained to perform the reconstruction. Results indicate significant correlation between tree growth and streamflow. A multiple linear regression model created with tree-ring width as the predictor of October - March American Fork River streamflow explained 51.7% of streamflow variance. Analysis of the reconstruction indicates that present records do not adequately represent potential streamflow variability, and several droughts of greater severity and length occurred before the instrumental period.

This study developed an integrated biogeochemical and hydrological modelling system by incorporating the latest versions of the nitrogen coupled Canadian Land Surface Scheme-Canadian Terrestrial Ecosystem Model (CLASS-CTEM) into the Modelisation Environmentale Communautaire (MEC) Surface and Hydrology system (MESH), hereafter referred to as MESH-CTEM. The newly developed MESH-CTEM modelling system allows simulations of energy, water, carbon and nitrogen fluxes and their feedbacks on vegetation growth and exploration of impacts of future climatic changes on catchment-scale processes. Performance of the MESH-CTEM system was tested at the Big Creek watershed within Norfolk county, Ontario, Canada, which is a 573 km2 crop-dominated catchment with areas of broadleaf and needleleaf forests, using observed eddy covariance flux, meteorological and hydrological datasets from October 2004 to December 2017 at a grid resolution of 0.02o latitude × 0.02o longitude. MESH-CTEM showed a significant increase in the simulated streamflow as compared to MESH running with only CLASS, excluding dynamic vegetation growth and carbon fluxes, resulting in an overall increase in the accuracy of streamflow with Nash-Sutcliffe Efficiency (NSE) indices of 0.38 and 0.12 respectively. Significant improvements were also seen for each Plant Functional Type (PFT) within the catchment with respect to energy fluxes, evaporation and soil water regimes. Many of these improvements in simulated fluxes were due in part by changes in the canopy conductance formulation, more realistic soil heat and water processes due to the introduction of fine soil layers, inter-grid transfers of water and other spatial components and vegetation cover feedbacks on energy, water and carbon exchanges by using dynamic vegetation growth processes. Simulated averaged gross ecosystem productivity, ecosystem respiration, latent heat flux and sensible heat flux for the entire catchment were respectively 660 g C m−2 yr−1, 640 g C m−2 yr−1, 32.5 W m-2 and 27.1 W m-2. Application and use of MESH-CTEM will help to study the impact of climate change and extreme events on energy, water and carbon fluxes and associated feedbacks at the catchment scale. Additionally, this will help bridge a major gap in hydrologic modelling studies through integration of biogeochemical processes. / Thesis / Master of Science (MSc)

Accurate hydrologic forecasting is vital for proper water resource management. Practices that are impacted by these forecasts include power generation, reservoir management, agricultural water use, and flood early warning systems. Despite these needs, the models largely used are simplifications of the real world and are therefore imperfect. The forecasters face other challenges in addition to the model uncertainty, which includes imperfect observations used for model calibration and validation, imperfect meteorological forecasts, and the ability to effectively communicate forecast results to decision-makers. Bayesian methods are commonly used to address some of these issues, and this thesis will be focused on improving methods related to recursive Bayesian estimation, more commonly known as data assimilation. Data assimilation is a means to optimally account for the uncertainties in observations, models, and forcing data. In the literature, data assimilation for urban hydrologic and flood forecasting is rare; therefore the main areas of study in this thesis are urban and semi-urban watersheds. By providing improvements to data assimilation methods, both hydrologic and flood forecasting can be enhanced in these areas. This work explored the use of alternative data products as a type of observation that can be assimilated to improve hydrologic forecasting in an urban watershed. The impact of impervious surfaces in urban and semi-urban watersheds was also evaluated in regards to its impact on remotely sensed soil moisture assimilation. Lack of observations is another issue when it comes to data assimilation, particularly in semi- or fully-distributed models; because of this, an improved method for updating locations which do not have observations was developed which utilizes information theory’s mutual information. Finally, we explored extending data assimilation into the short-term forecast by using prior knowledge of how a model will respond to forecasted forcing data. Results from this work found that using alternative data products such as those from the Snow Data Assimilation System or the Soil Moisture and Ocean Salinity mission, can be effective at improving hydrologic forecasting in urban watersheds. They also were effective at identifying a limiting imperviousness threshold for soil moisture assimilation into urban and semi-urban watersheds. Additionally, the inclusion of mutual information between gauged and ungauged locations in a semi-distributed hydrologic model was able to provide better state updates in models. Finally, by extending data assimilation into the short-term forecast, the reliability of the forecasts could be improved substantially. / Dissertation / Doctor of Philosophy (PhD) / The ability to accurately model hydrological systems is essential, as that allows for better planning and decision making in water resources management. The better we can forecast the hydrologic response to rain and snowmelt events, the better we can plan and manage our water resources. This includes better planning and usage of water for agricultural purposes, better planning and management of reservoirs for power generation, and better preparing for flood events. Unfortunately, hydrologic models primarily used are simplifications of the real world and are therefore imperfect. Additionally, our measurements of the physical system responses to atmospheric forcing can be prone to both systematic and random errors that need to be accounted for. To address these limitations, data assimilation can be used to improve hydrologic forecasts by optimally accounting for both model and observation uncertainties. The work in this thesis helps to further advance and improve data assimilation, with a focus on enhancing hydrologic forecasting in urban and semi-urban watersheds. The research presented herein can be used to provide better forecasts, which allow for better planning and decision making.

Anthropogenic climate change will impact Virginia's hydrologic processes in unforeseen ways in the coming decades. This research describes variability in meteorology (temperature and precipitation) and associated hydrologic processes (evapotranspiration) throughout an ensemble of 31 general circulation models (GCMs) used by the Chesapeake Bay Program (CBP). Trends are compared with surface runoff generation patterns for a variety of land uses to investigate climate's effect on runoff generation. Scenarios representing pairings of the tenth, fiftieth, and ninetieth percentiles of precipitation and temperature in the CBP 31-model ensemble were run through VADEQ's VA Hydro hydrologic model to investigate streamflow's response to climate. Temperature changes across the study area were minimized in the tenth percentile scenario (+1.02 to +1.24◦C) and maximized in the ninetieth (+2.20 to +3.02◦C), with evapotranspiration change following this trend (tenth: +2.84 to +3.81%; ninetieth: +6.53 to +10.2%). Precipitation change ranged from -10.9 to -7.30% in the tenth to +22.1 to +28.0% in the ninetieth. Runoff per unit area was largely dependent on land use, with the most extreme changes in runoff often seen in forested and natural land uses (-24% in tenth; +53% in ninetieth) and the least extreme seen in impervious and feeding space land(tenth: -11%; ninetieth: +30%). Both overall runoff per unit area and streamflow changed drastically from the base in the tenth (-20.4% to -25.9% change in median runoff; -19.8% to -27.1% change in median streamflow) and ninetieth (+30.4% to +53.7% change in median runoff; +33.0% to +77.8% change in median streamflow) percentile scenarios. / Master of Science / Human-caused climate change will impact Virginia's hydrologic processes in unforeseen ways in the coming decades. This research describes variability in meteorology (temperature and precipitation) and associated hydrologic processes (evapotranspiration) throughout an ensemble of 31 general circulation models (GCMs) used by the Chesapeake Bay Program (CBP). Trends are compared with surface runoff generation patterns for a variety of land uses to investigate climate's effect on runoff generation. Scenarios representing pairings of the tenth, fiftieth, and ninetieth percentiles of precipitation and temperature in the CBP 31-model ensemble were run through VADEQ's VA Hydro hydrologic model to investigate streamflow's response to climate. Temperature changes across the study area were minimized in the tenth percentile scenario (+1.02 to +1.24◦C) and maximized in the ninetieth (+2.20 to +3.02◦C), with evapotranspiration change following this trend (tenth: +2.84 to +3.81%; ninetieth: +6.53 to +10.2%). Precipitation change ranged from -10.9 to -7.30% in the tenth to +22.1 to +28.0% in the ninetieth. Runoff per unit area was largely dependent on land use, with the most extreme changes in runoff often seen in forested and natural land uses (-24% in tenth; +53% in ninetieth) and the least extreme seen in impervious and feeding space land(tenth: -11%; ninetieth: +30%). Both overall runoff per unit area and streamflow changed drastically from the base in the tenth (-20.4% to -25.9% change in median runoff; -19.8% to -27.1% change in median streamflow) and ninetieth (+30.4% to +53.7% change in median runoff; +33.0% to +77.8% change in median streamflow) percentile scenarios.

A Disaggregation model was tested for arid land stream flow generating. The test was performed on data from Black River, near Fort Apache, Arizona. The model was tested in terms of preserving the relevant historical statistics on both monthly and daily levels, the monthly time series were disaggregated to a random observation of their daily components and the daily components were then reaggregated to yield monthly values. A computer model (DSGN) was developed to perform the model implementation. The model was written and executed on the Macintosh plus personal computer Data from two months were studied; the October data represented the low flow season, while the April data represented the high flow season. Twenty five years of data for each month was used. The generated data for the two months was compared with the historical data.

This thesis explores the effect of Acacia mearnsii on streamflow in the Eastern Cape. There is a need for data on the localised effects of removing alien trees from the riparian zones within the Fynbos Biome. Fynbos catchments throughout the Western and Eastern Cape yield large quantities of good quality water which is an essential resource in the region. To convince local land owners to manage their riparian zones, small scale experimental results will prove invaluable to assure them of the immediate advantages for themselves and for downstream water users. Three permanent weirs were built 500 m apart to monitor the effect of removing A. mearnsii on streamflow in the Sand River, Eastern Cape. Consecutive weirs allowed for the comparison of streamflow between a cleared and uncleared section of the river without significant differences in riparian conditions, channel morphology and vegetation densities. A site survey confirmed comparable densities of A. mearnsii in both sections. A sample of trees was weighed and a relationship was found between diameter at breast height and above ground wet biomass. Between the first two weirs, 2.5 ha of riparian zone was cleared amounting to approximately 160tlha. Streamflow was monitored from the 10th of January 1996 to the 9th of September 1996. The average streamflow reduction for the duration of the experiment was 15.1 m³/ha/day or 551 mm per annum. Initially, after a period of above average rainfall, streamflow was augmented by discharge from the riparian zone but as conditions dried out, there was a net uptake of water with the highest average uptake of 23. 7m³/ha/day in June. A comparison between weather conditions and streamflow reduction indicated there is a complex relationship, with evidence of A. mearnsii exhibiting control of water loss during dry conditions. Acacia mearnsii trees in the riparian zone have been shown to cause significant streamflow reduction. Permanent weirs were found to be appropriate for this type of study. There is a need for further research on the effect of alien trees in riparian zones around South Africa as there is potential for significant increases in streamflow.

Many hydrologic basins in the southwest have seen their perennial streamflows turn to ephemeral, their riparian communities disappear or be jeopardized, and their aquifers suffer from severe overdrafts. Under -management of ground -water exploitation and of conjunctive use of surface and ground waters are the main reasons for these events.