<p> This study investigates the use of large scale ensemble weather predictions provided by the National Centers for Environmental Prediction (NCEP) medium range forecast (MRF) modeling system, for short-term hydrologic forecasting. The weather predictors are used to downscale daily precipitation and temperature series at two meteorological stations in the Saguenay watershed in northeastern Canada. Three data-driven methods, namely, statistical downscaling model (SDSM), time lagged feedforward neural network (TLFN), and evolutionary polynomial regression (EPR), are used as downscaling models and their downscaling results are compared. The downscaled results of the best models are used as additional inputs in two hydrological models, Hydrologiska Byrans Vattenbalansavdelning (HBV) and Bayesian neural networks (BNN), for up to 14 day ahead reservoir inflow and river flow forecasting. The performance of the two hydrological forecasting models is compared, the ultimate objective being to improve 7 to 14 day ahead forecasts. </p> <p> The downscaling results show that all the three models have good performance in downscaling temperature time series, the correlation between the observed and downscaled data is more than 0.90, however the downscaling results are less accurate for precipitation, the correlation coefficient is no more than 0.62. TLFN and EPR models have quite close performance in most cases, and they both perform better than SDSM. </p> <p> Therefore the TLFN downscaled meteorological data are used as predictors in the HBV and BNN hydrological models for up to 14 day ahead reservoir inflow and river flow forecasting, and the forecasting results are compared with the case where no downscaled data is included. The results show that for both reservoir inflow and river flow, HBV models have better performance when including downscaled meteorological data, while there is no significant improvement for the BNN models. When comparing the performance of HBV and BNN models through scatter plots, it can be found that BNN models perform better in low flow forecasting than HBV models, while less good in peak flow forecasting. </p> / Thesis / Master of Science (MSc)

In cold and snowy countries, water resources management and planning require accurate and reliable spring peak flow forecasts which call for adequate snowmelt estimation techniques. Thus, exploring the potential of snowmelt models to improve the spring peak flow prediction has been an active research area. Snow models vary in degree of complexity from simple empirical models to complex physically based models. Whereas majority of studies on snowmelt modeling have focused on comparing the performance of empirical snowmelt estimation techniques with physically based methods, very few studies have investigated empirical methods and conceptual models for hydrological applications. This study investigates the potential of a simple Degree-Day Method (DDM) to effectively and accurately predict peak flows compared to sophisticated SNOW-17 model at La-Grande River Basin (LGRB), Quebec and Upper Assiniboine river at Shellmouth Reservoir (UASR), Manitoba. Moreover, since hydrologic models highly rely on estimated parameter vectors to produce accurate streamflow simulations, accurate and efficient parameter optimization techniques are essential. The study also investigates the benefits of seasonal model calibration versus annual model calibration approach. The study is performed using two hydrological models, namely MAC-HBV (McMaster University Hydrologiska Byrans Vattenbalansavdelning) and SAC-SMA (Sacramento Soil Moisture Accounting) and their model combinations thereof. Results indicate that the simple DDM performed consistently better at both study sites and showed significant improvement in prediction accuracy at UASR. Moreover, seasonal model calibration appears to be an effective and efficient alternative to annually calibrated model especially when extreme events are of particular interest. Furthermore, results suggest that SAC-SMA model outperformed MAC-HBV model, no matter what snowmelt computation method, calibration approach or study basin is used. Conclusively, DDM and seasonal model optimization approach coupled with SAC-SMA hydrologic model appears to be a robust model combination for enhanced spring peak flow prediction. A significant advantage of aforementioned modeling approach for operational hydrology is that it demonstrates computational efficiency, ease of implementation and is less time-consuming. / Thesis / Master of Science (MSc)

Surface mining activities are known to affect the quantity and quality of stormwater runoff. This can create flooding and water quality degradation of receiving streams. The Surface Mining Control and Reclamation Act (SMCRA) of 1977 provides regulations intended to produce environmentally acceptable results from mining operations. The SMCRA requires that extensive pre-mining monitoring be carried out to assist in determining the probable hydrologic consequences (PHC) of mining. The Finite Element Storm Hydrograph Model (FESHM) was used to demonstrate the utility of hydrologic modeling concepts in simulating runoff volumes and peak flows. Guidelines were proposed for using this methodology to simulate selected pre-mining hydrologic conditions. The use of geographic information system (GIS) technology as a tool for improving data management and modeling efficiency was demonstrated. The required input watershed characteristics were digitized, stored, and manipulated using a computerized GIS. Appropriate software was developed to integrate the GIS with FESHM. The ability of FESHM to simulate runoff events in an ungaged context was evaluated using an experimental watershed. First, simulations were conducted using two separate data bases, "lumped" and "detailed", in order to evaluate the effect of limited data availability, as expected in mining regions, on FESHM's predictive ability. The "lumped" data base produced better simulation results, however, more thorough and detailed research is needed to determine the level of data resolution necessary for a given level of simulation accuracy. Next., significant runoff events from 17-years of the historical record were simulated using data from the "lumped" data base. Statistical analyses were used to make judgments on parameter estimation and model usage. Regression methodology was used to assess expected error and model bias. Simulation bias was found to be related to the input rainfall intensity levels. The results suggest that either spatial variability or parameter values were not adequately defined and that some form of calibration is needed. Two additional drainage basins were used to evaluate FESHM's predictive capabilities in situations considered representative of mining regions. The results indicated that more thorough investigations of watershed characteristics must be made, that calibration procedures should be performed for each watershed, and that FESHM does not adequately model the physical processes involved in forest hydrology. / Master of Science

Globally, peatlands are vulnerable to degradation via drainage, with consequences for ecosystem structure and function such as increased fire vulnerability, soil oxidation, and altered vegetation composition. Peatland function is largely dependent on hydrologic regimes and their influences on the accumulation and properties of peat soil. Therefore, an understanding of soil-hydrology interactions is needed to inform management in drained peatlands, including expansive systems such as the Great Dismal Swamp (GDS; Virginia and North Carolina, USA) where hydrologic restoration is underway. Two physically distinct soil layers have been observed at GDS, the upper layer thought to be a result of past drainage and the lower layer more representative of an undisturbed state. To understand the occurrence and consequences of these distinct layers, we integrated continuous water level data, peat profile characterization, and analyzed soil physical and hydraulic properties. The transition from upper to lower peat soil layers typically occurred at depths below contemporary water level observations, suggesting that the upper layer may be a result of historical drainage with limited recovery following hydrologic restoration. We also found distinct differences between the properties of the two layers, where upper layers had lower fiber and organic matter contents and higher bulk densities. Further, upper layers had higher proportions of macropores, resulting in an overall lower water retention capacity. These differences in layer properties suggest the upper layer is more susceptible to drying, increasing fire vulnerability, oxidation, and shifts in vegetation composition that do not support current management objectives. / Master of Science / Peatlands provide many valuable ecosystem services, including carbon storage, water quality maintenance, and habitat provision. However, peatlands have been subjected to centuries of drainage (i.e., lowered water levels) to support timber harvesting, land conversion, and other land use actions. Drainage and the resulting drier conditions can lead to soil carbon loss, increased fire vulnerability, and changes in vegetation communities. Additionally, peatland drainage has consequences for peat soil properties and their role in ecosystem services. In an effort to restore peatland ecosystem services, hydrologic restoration, usually in the form of water control structures, is often implemented to reduce drainage and reestablish historical water levels. To guide restoration practices, research is needed to understand how drained peat soils respond to such hydrologic management. In this study, we investigated peat soil profiles, current water level regimes, and soil properties at the Great Dismal Swamp (Virginia and North Carolina, USA), a drained peatland currently undergoing hydrologic restoration. We found a visibly distinct upper soil layer, which we suggest developed as a result of past drainage and with little recovery under restored, wetter conditions. We also found that this upper layer has altered soil properties and thus is more vulnerable to drying, with implications for ecosystem function such as fire vulnerability, carbon sequestration and vegetation composition. Together, our findings will help inform restoration and water level management at GDS and our understanding of drained peatlands more broadly.

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.

Low Impact Development (LID) is gaining popularity as a solution to erosion, flooding, and water quality problems that stormwater ponds partially address. LID analysis takes a spatially lumped approach, based on maintaining the predevelopment Curve Number and time of concentration, precluding consideration of the spatial distribution of impervious areas and Integrated Management Practices (IMP's), runoff-runon processes, and the effects of land grading. Success is thus dependent on the accuracy of the assumption of watershed uniformity, applied to both land cover distribution and flow path length. Considering the cost of long-term paired watershed monitoring, continuous, spatially-distributed hydrologic modeling was judged a better method to compare the response of LID, forest, and conventional development. Review of available models revealed EPA-SWMM 4.4H as the most applicable to the task. A 4.3-acre subwatershed of a local subdivision was adapted to LID using impervious surface disconnection, forest retention, and IMPs. SWMM was applied to the LID development at a fine spatial scale, yielding an 80-element SWMM model. The LID model was modified to reflect conventional development, with gutters, storm sewer, and detention. A predevelopment forest model was also developed. Two parameter sets were used, representing a range of assumptions characterized as favorable or unfavorable toward a particular development form. Modeled scenarios included favorable and unfavorable versions of Forest, LID, uncontrolled Conventional Development, and Conventional Development with Stormwater Management. SWMM was run in continuous mode using local rainfall data, and event mode using NRCS design storms. Runoff volumes, peak flows, and flow duration curves were compared. / Master of Science

In 2016 the Iowa Watershed Approach (IWA) was created to increase community resiliency against flooding, to develop hydrologic assessments that would identify strategies to reduce flooding, and to implement those strategies within nine identified watersheds that experienced flooding between 2011 and 2013. One of the nine watersheds was the Upper Iowa watershed located in northeast Iowa. This thesis focuses on the work that has been done to create a hydrologic assessment of the Upper Iowa watershed. The hydrologic assessment identifies potential conservation practices, creates a hydrologic model to assess the hydrologic cycle over the past ten years, and identifies strategies to reduce flooding within the watershed. Many potential agricultural conservation practices within the Upper Iowa watershed were identified and trends relating to the soil, land use, and topography were determined. In addition, a methodology to compare potential conservation practices with existing conservation practices actually in place was developed including a tool to estimate the size of grassed waterways to NRCS design guidelines. The comparison validated the methodologies used to identify potential practice placements, identified locations where potential practices could be implemented, and showed how stakeholder preferences influence conservation implementation. Additionally, a hydrologic model of the Upper Iowa watershed was developed, using the new Generic Hydrologic Overland-Subsurface Toolset model and calibrated to simulate the time period of 2007 through 2016. The model was evaluated against water balance ratios and performance statistics calculated from measured data. The model achieved Nash Sutcliffe Efficiency scores for streamflow above 0.7 and percent bias scores between ±12% for the three wettest years of 2008, 2013, and 2016. With the calibrated model, the benefits of continuous cover crop implementation were investigated under current conditions and under increased extreme precipitation intensity expected from climate change over the next half century. The results of this investigation determined that continuous cover crops increased evapotranspiration within the early half of the year creating more storage within the soil. Thus the flood risk from convective storms during the summer was lowered. In addition, the benefits from cover crops in terms of peak flow and volume reductions were cumulative increasing each consecutive year and were proportional to the percentage of cover cropped area. Lastly, a scenario using cover crops in a future extreme precipitation environment resulted in a reduction of peak discharge to current conditions. The results of this thesis will guide both future work within the Upper Iowa watershed and contribute to the knowledge of hydrologic planning and modeling within agricultural watersheds.

The Hydrologic Information System Project of the Consortium of Universities for the Advancement of Hydrologic Science, Inc. (CUAHSI) has successfully created a large-scale prototype Hydrologic Information System (HIS). This system catalogs and provides access to over 23 million time series of hydrologic data, which are distributed across the United States at various academic, research, and governmental data providers. The service-oriented architecture that enables the HIS comprises distributed hydrologic data servers, a centralized series catalog, and various client software applications, and is supported by WaterML, a standardized language for transmission of hydrologic data. The current architectural model, termed the Network-Observations Model, of the HIS relies on a searchable central catalog of series metadata. Harvesting series metadata from large federal data providers, such as the USGS, EPA, and NCDC, has proven a laborious undertaking and involves custom database migration tools. This time-consuming harvesting task, coupled with a multitude of custom-coded solutions at the central series catalog has led to concerns with the long-term sustainability of the current architectural model. A new architectural model, termed the Services Stack Model, is proposed in this thesis. In the proposed model, a catalog of services metadata, rather than of series metadata is used to connect hydrologic data consumers with data providers. Internationally-recognized web service and data encoding standards, including the upcoming WaterML2.0 specification, from the Open Geospatial Consortium are used as the backbone of the new model. The proposed model will hopefully lead to greater acceptance of the CUAHSI-HIS, and result in increased sustainability and reduced maintenance of the system in the long-term. / text

The segment of the Pecos River that extends from Red Bluff Reservoir until it discharges to the Rio Grande/Bravo near Langtry was studied in this project. Hydrologic behavior of the basin was analyzed between 1981 and 2000, the first ten year period for calibration and the second ten year period for validation by using Water Evaluation and Planning Software (WEAP, SEI, 2006). Simulated streamflows were compared with naturalized streamflows (RJBCO, 2003) at two control points, one in the middle of the basin near Girvin and the other one is at the end of the basin near Langtry. The purpose of the project is to create a valid model for water availability simulations in the Pecos River Basin to be used for future water availability simulations considering climate change effects. The basin was divided into two parts in order to evaluate the results, the upper basin and the entire basin (below Red Bluff reservoir) according to the location of control gages. Simulated streamflows closely match the naturalized flows at the Girvin station in the upper basin. Although the results at the Langtry station for the entire basin are not as good as Girvin, the model still reproduces streamflows well enough to represent the hydrologic behavior of the basin, especially for the base flow. Considering the complex geological structure of the Pecos River Basin below Red Bluff Reservoir, the results can be considered satisfactory. The model can be used for future water availability predictions in the basin considering climate change effects. / text

abstract: Ponderosa pine forests are a dominant land cover type in semiarid montane areas. Water supplies in major rivers of the southwestern United States depend on ponderosa pine forests since these ecosystems: (1) receive a significant amount of rainfall and snowfall, (2) intercept precipitation and transpire water, and (3) indirectly influence runoff by impacting the infiltration rate. However, the hydrologic patterns in these ecosystems with strong seasonality are poorly understood. In this study, we used a distributed hydrologic model evaluated against field observations to improve our understandings on spatial controls of hydrologic patterns, appropriate model resolution to simulate ponderosa pine ecosystems and hydrologic responses in the context of contrasting winter to summer transitions. Our modeling effort is focused on the hydrologic responses during the North American Monsoon (NAM), winter and spring periods. In Chapter 2, we utilized a distributed model explore the spatial controls on simulated soil moisture and temporal evolution of these spatial controls as a function of seasonal wetness. Our findings indicate that vegetation and topographic curvature are spatial controls. Vegetation controlled patterns during dry summer period switch to fine-scale terrain curvature controlled patterns during persistently wet NAM period. Thus, a climatic threshold involving rainfall and weather conditions during the NAM is identified when high rainfall amount (such as 146 mm rain in August, 1997) activates lateral flux of soil moisture and frequent cloudy cover (such as 42% cloud cover during daytime of August, 1997) lowers evapotranspiration. In Chapter 3, we investigate the impacts of model coarsening on simulated soil moisture patterns during the NAM. Results indicate that model aggregation quickly eradicates curvature features and its spatial control on hydrologic patterns. A threshold resolution of ~10% of the original terrain is identified through analyses of homogeneity indices, correlation coefficients and spatial errors beyond which the fidelity of simulated soil moisture is no longer reliable. Based on spatial error analyses, we detected that the concave areas (~28% of hillslope) are very sensitive to model coarsening and root mean square error (RMSE) is higher than residual soil moisture content (~0.07 m3/m3 soil moisture) for concave areas. Thus, concave areas need to be sampled for capturing appropriate hillslope response for this hillslope. In Chapter 4, we investigate the impacts of contrasting winter to summer transitions on hillslope hydrologic responses. We use a distributed hydrologic model to generate a consistent set of high-resolution hydrologic estimates. Our model is evaluated against the snow depth, soil moisture and runoff observations over two water years yielding reliable spatial distributions during the winter to summer transitions. We find that a wet winter followed by a dry summer promotes evapotranspiration losses (spatial averaged ~193 mm spring ET and ~ 600 mm summer ET) that dry the soil and disconnect lateral fluxes in the forested hillslope, leading to soil moisture patterns resembling vegetation patches. Conversely, a dry winter prior to a wet summer results in soil moisture increases due to high rainfall and low ET during the spring (spatially averaged 78 mm ET and 232 mm rainfall) and summer period (spatially averaged 147 mm ET and 247 mm rainfall) which promote lateral connectivity and soil moisture patterns with the signature of terrain curvature. An opposing temporal switch between infiltration and saturation excess runoff is also identified. These contrasting responses indicate that the inverse relation has significant consequences on hillslope water availability and its spatial distribution with implications on other ecohydrological processes including vegetation phenology, groundwater recharge and geomorphic development. Results from this work have implications on the design of hillslope experiments, the resolution of hillslope scale models, and the prediction of hydrologic conditions in ponderosa pine ecosystems. In addition, our findings can be used to select future hillslope sites for detailed ecohydrological investigations. Further, the proposed methodology can be useful for predicting responses to climate and land cover changes that are anticipated for the southwestern United States. / Dissertation/Thesis / Ph.D. Geological Sciences 2012