Doctor of Philosophy / Department of Biological & Agricultural Engineering / Stacy L. Hutchinson / The Soil Conservation Service-Curve Number (SCS-CN) method is widely used to estimate direct runoff from rainfall events; however, the method does not account for the dynamic rainfall-runoff relationship. This study used back-calculated curve numbers (CNs) and Normalized Difference Vegetation Index (NDVI) to develop NDVI-based CNs (CN[subscript]NDV) using four small northeastern Kansas grassland watersheds with average areas of 1 km² and twelve years (2001–2012) of daily precipitation and runoff data. Analysis indicated that the CN[subscript]NDVI model improved runoff predictions compared to the SCS-CN method. The CN[subscript]NDVI also showed greater variability in CNs, especially during growing season, thereby increasing the model’s ability to estimate relatively accurate runoff from rainfall events since most rainfall occurs during the growing season. The CN[subscript]NDVI model was applied to small, disturbed grassland watersheds to assess the model’s ability to detect land cover change impact for military maneuver damage and large, diverse land use/cover watersheds to assess the impact of scaling up the model. CN[subscript]NDVI application was assessed using a paired watershed study at Fort Riley, Kansas. Paired watersheds were identified through k-means and hierarchical-agglomerative clustering techniques. At the large watershed scale, Daymet precipitation was used to estimate runoff, which was compared to direct runoff extracted from stream flow at gauging points for Chapman (grassland dominated) and Upper Delaware (agriculture dominated) watersheds. In large, diverse watersheds, CN[subscript]NDVI performed better in moderate and overall flow years. Overall, CN[subscript]NDVI more accurately simulated runoff compared to SCS-CN results: The calibrated model increased by 0.91 for every unit increase in observed flow (r = 0.83), while standard CN-based flow increased by 0.506 for every unit increase in observed flow (r = 0.404). Therefore, CN[subscript]NDVI could help identify land use/cover changes and disturbances and spatiotemporal changes in runoff at various scales. CN[subscript]NDVI could also be used to accurately estimate runoff from precipitation events in order to instigate more timely land management decisions.

<p>Streamflow prediction in ungauged basins (PUB) is a process generating streamflow time series at ungauged reaches in a river network. PUB is essential for facilitating various engineering tasks such as managing stormwater, water resources, and water-related environmental impacts. Machine Learning (ML) has emerged as a powerful tool for PUB using its generalization process to capture the streamflow generation processes from hydrological datasets (observations). ML’s generalization process is impacted by two major components: data splitting process of observations and the architecture design. To unveil the potential limitations of ML’s generalization process, this dissertation explores its robustness and associated uncertainty. More precisely, this dissertation has three objectives: (1) analyzing the potential uncertainty caused by the data splitting process for ML modeling, (2) investigating the improvement of ML models’ performance by incorporating hydrological processes within their architectures, and (3) identifying the potential biases in ML’s generalization process regarding the trend and periodicity of streamflow simulations.</p><p>The first objective of this dissertation is to assess the sensitivity and uncertainty caused by the regular data splitting process for ML modeling. The regular data splitting process in ML was initially designed for homogeneous and stationary datasets, but it may not be suitable for hydrological datasets in the context of PUB studies. Hydrological datasets usually consist of data collected from diverse watersheds with distinct streamflow generation regimes influenced by varying meteorological forcing and watershed characteristics. To address the potential inconsistency in the data splitting process, multiple data splitting scenarios are generated using the Monte Carlo method. The scenario with random data splitting results accounts for frequent covariate shift and tends to add uncertainty and biases to ML’s generalization process. The findings in this objective suggest the importance of avoiding the covariate shift during the data splitting process when developing ML models for PUB to enhance the robustness and reliability of ML’s performance.</p><p>The second objective of this dissertation is to investigate the improvement of ML models’ performance brought by Physics-Guided Architecture (PGA), which incorporates ML with the rainfall abstraction process. PGA is a theory-guided machine learning framework integrating conceptual tutors (CTs) with ML models. In this study, CTs correspond to rainfall abstractions estimated by Green-Ampt (GA) and SCS-CN models. Integrating the GA model’s CTs, which involves information on dynamic soil properties, into PGA models leads to better performance than a regular ML model. On the contrary, PGA models integrating the SCS-CN model's CTs yield no significant improvement of ML model’s performance. The results of this objective demonstrate that the ML’s generalization process can be improved by incorporating CTs involving dynamic soil properties.</p><p>The third objective of this dissertation is to explore the limitations of ML’s generalization process in capturing trend and periodicity for streamflow simulations. Trend and periodicity are essential components of streamflow time series, representing the long-term correlations and periodic patterns, respectively. When the ML models generate streamflow simulations, they tend to have relatively strong long-term periodic components, such as yearly and multiyear periodic patterns. In addition, compared to the observed streamflow data, the ML models display relatively weak short-term periodic components, such as daily and weekly periodic patterns. As a result, the ML’s generalization process may struggle to capture the short-term periodic patterns in the streamflow simulations. The biases in ML’s generalization process emphasize the demands for external knowledge to improve the representation of the short-term periodic components in simulating streamflow.</p>

No description available.

Indiana University-Purdue University Indianapolis (IUPUI) / The transport of excess nutrients, sediment, and other contaminants to surface waters has been shown to cause a number of environmental and human health concerns. An understanding of the export pathways that these contaminants follow to surrounding water bodies is crucial to the anticipation and management of peak concentration events. Several studies have demonstrated that the majority of annual contaminant loading in the Midwest occurs during periods of elevated discharge. However, many studies use a limited number of sampling points to determine concentration patterns, loadings, and fluxes which decreases accuracy. Through high-resolution storm sampling conducted in a 2945 km2 (1137 mi2) area of central Indiana’s Upper White River Watershed, this research has documented the complex concentration signals and fluxes associated with a suite of cations, nutrients, and contaminants and isolated their primary transport pathways. Additionally, by comparing the results of similar studies conducted on smaller areas within this watershed, differences in concentration patterns and fluxes, as they relate to drainage area, have also been documented. Similar to the results of previous studies, NO3- concentrations lacked a well-defined relationship relative to discharge and was attributed to primarily subsurface contribution. DOC was exported along a shallow, lateral subsurface pathway, TP and TSS via overland flow, and TKN through a combination of both. Near or in-channel scouring of sediment increased DOC, TKN, TP, and TSS concentrations during Storm 2. Atrazine export was attributed to a combination of overland and subsurface pathways. 2-MIB and geosmin derived from different sources and pathways despite being produced by similar organisms. 2-MIB concentration patterns were characterized by dilution of an in-stream source during Storm 1 and potential sediment export during Storm 2 while in-stream concentrations or a sediment source of geosmin was rapidly exhausted during Storm 1. Many of the concentration patterns were subject to an exaggerated averaging effect due to the mixing of several larger watersheds, especially during Storm 1. This research illustrates the need for high-frequency sampling to accurately quantify contaminant loads for total maximum daily load (TMDL) values, developing best management practices (BMPs), and confronting the challenges associated with modeling increasingly larger-scale watersheds.

<p dir="ltr">This study analyzes one subset of twelve lakes within Kosciusko County, Indiana between 2015 and 2021 to provide a quantitative understanding of the mechanisms which influence onset and occurrence of HABs. Analysis of water samples, balanced by imagery from satellite remote sensing platforms, are used to quantify the biogeochemical state of these water systems and better understand the mechanisms involved in formation of HABs. Parameters studied include in-situ measurements (e.g., water temperature), laboratory measurements (e.g., microcystin, nitrogen, and phosphorous concentrations), and satellite derived responses (chlorophyll-a). Results indicate no single parameter is correlated with cyanotoxin concentrations, but instead multiple parameters have a synergistic effect on algal bloom growth and toxicity.</p>

<p>  </p> <p>The management of water resources in East Africa is inherently challenged by rainfall variability and the uneven spatial distribution of freshwater resources. In addition to these issues, meteorological and water data collection has been inconsistent over the past decades, and unclearly defined purposes or end goals for collected data have left many datasets ineffectively curated. In light of the data intensiveness of current modelling and planning methods, data scarcity and inaccessibility have become substantial impediments to informed decision-making. Among the outputs of this research are 1) a revised technique for evaluating bias correction performance on reanalysis data for use in regions where precipitation data is temporally discontinuous which can potentially be applied to other types of climate data as well, 2) a new methodology for quantifying qualitative information contained in legislation and official documents and websites for the assessment of relationships between documented meteorological and water data policies and resulting outcomes in terms of data availability and accessibility, and 3) a fresh look at data needs and the value data holds with respect to water resources decision-making and management in the region.</p>

<p> The hydrological cycle is affected by land use variability. Land use spatial and temporal variability has the power to alter watershed runoff, water resource quantity and quality, ecosystems, and environmental sustainability. In recent decades, agriculture lands, pastures, plantations, and urban areas have increased, resulting in significant increases in energy, water, and fertilizer usage, as well as significant biodiversity losses. </p>

No description available.

<p dir="ltr">Alpine glacier meltwater is an important source of recharge supporting groundwater flow processes in the high mountains. In the face of rapid ice loss, knowledge of response times of mountain aquifers to loss of glacial ice is critical in evaluating the sustainability of alpine water resources for human communities and alpine ecosystems. Glaciers are very sensitive to changes in climate, they advance during periods of global or regional cooling, and they retreat in response to global or regional warming conditions. When the glaciers grow, the equilibrium-line altitude separating the zone of accumulation and zone of ablation on the glacier moves downslope; it moves upslope when they retreat. The latter is not a sustainable condition for the glacier. Previous studies have shown that glacial meltwater is an important source of groundwater recharge. However, we lack fundamental information on the importance of glacial meltwater in mountain groundwater processes such as supporting baseflow generation to alpine streams, perennial flow to alpine springs, and the geochemical evolution of groundwater in mountain aquifers. Thus, continued glacial ice loss may have severe consequences for alpine hydrological and hydrogeological systems.</p><p dir="ltr">Glacier National Park (GNP) and Mount Hood National Forest (MH), both have alpine glaciers. These two study sites show different responses to climate change since their glaciers are in different states of retreat. GNP glaciers are in advanced stages of retreat compared to MH glaciers. Groundwater samples were collected from springs, seasonal snow, glacial ice, and glacial melt (subglacial flow) in GNP and MH. The samples were analyzed for a suite of environmental isotopes and geochemical tracers to address the following questions: 1) How are isotopic fingerprints of glacial meltwater preserved in mountain-block aquifers? What does the isotopic fingerprint of subglacial flow tell us about melting, meltwater processes, and mixing processes? 2) Is the preservation of the isotopic fingerprint of glacial meltwater affected by aspect controls on ice preservation? Aspect is defined as the compass direction of the slope where the glacier is found. 3) What controls groundwater flow and flowpath connectivity from high elevations (near glacier) to lower elevations? What geologic units support groundwater flow to local- and regional-scale springs and flowpath connectivity across spatial scales in each study site?</p><p dir="ltr">The flow of groundwater in mountainous terrain is heavily dependent on the hydraulic properties of the bedrock including presence/absence of dipping layers and structural features, primary and secondary porosity, and presence/absence of ongoing tectonic activity. Strontium isotopes (⁸⁷Sr/⁸⁶Sr) were used to identify the rock units that host groundwater flowpaths and to quantify flowpath connectivity across spatial scales in both study sites. The ⁸⁷Sr/⁸⁶Sr data show that flowpaths in GNP are primarily hosted in the Helena Formation and permeable facies in the Snowslip Formation. Groundwater also flows through alluvium and younger bedrock units, and there is some flow along or through the volcanic sill in the Helena Formation. Hydrostratigraphy also affects groundwater flow and the spatial distribution of alpine springs in GNP. At MH, the rock units hosting flowpaths are young reworked volcanic rock units that are Quaternary in age. Flowpaths in MH appear to be connected across spatial scales since warm springs emerging along the lower southern slopes of Mount Hood preserve stable isotopic signatures of glacial meltwater. In comparison, nearly all the sampled springs in GNP emerge on south-facing slopes. This is not an indication of ice preservation, instead it’s controlled by hydrostratigraphy. In fact, it’s unlikely that high-elevation groundwater is strongly connected to low-elevation sites due to hydrostratigraphy. There are more springs on south-facing slopes at MH as well; however, they do not preserve an isotopic signature of recharge from glacial meltwater except for the warm springs. Springs on north-facing slopes in MH, however, do preserve the signature.</p><p dir="ltr">Tritium (³H) and chlorine-36 (³⁶Cl/Cl) were measured to assess how the isotopic fingerprint of glacial meltwater is preserved in mountain aquifers. The ³H activities in spring water are elevated in GNP and it’s difficult to differentiate between modern precipitation and glacial meltwater. Tritium activities are lower in MH, but it’s also difficult to differentiate between potential endmembers. This discrepancy could imply that glacial meltwater doesn’t contribute to groundwater recharge, but this doesn’t support the Bayesian stable isotope mixing model results of an earlier study. Instead, I infer that englacial mixing processes are affecting the isotopic fingerprint of subglacial melt. An englacial mixing model (EMM) was developed to explain how the isotopic fingerprint of subglacial flow (glacial meltwater) changes in relation to the stage of retreat. The stage of retreat is important because it controls the proportion of glacial meltwater to runoff from snowmelt and rain that enters the englacial network from the surface of the glacier. Mixing occurs in the englacial network, and the mixed water is transported to the base of the glacier. Englacial mixing in conduits, fractures, and moulins affects the ³H and ³⁶Cl/Cl fingerprint of subglacial flow and will, in turn, affect the isotopic fingerprint of recharge from glacial meltwater. For this study, the ³H is not robust by itself; however, ³⁶Cl/Cl shows some additional benefits over ³H. The EMM suggests that the impact of englacial mixing and the influence of modern precipitation on the isotopic composition of subglacial flow increases as the glacier retreats in both GNP and MH. This model is novel to the best of our knowledge. Additional testing of the EMM should be prioritized in the near future.</p>

<p dir="ltr">Tropical Cyclones (TCs) are intense low-pressure weather systems that acts as a meteorological monster causing severe rainfall and widespread freshwater flooding, leading to extensive damage and disruption. Quantitative precipitation estimates (QPEs) are crucial for accurately understanding and evaluating the impacts of TCs. However, QPEs derived from various modalities, such as rain gauges, ground-based merged radars, and satellites, can differ significantly and require thorough comparison. Understanding the limitations/advantages of using each QPE is essential to simulate a hydrological model especially to estimate extreme events like TCs. The objective of the dissertation is to 1) characterize the tropical cyclone precipitation (TCP) using three gridded products, 2) characterize the impact of using different QPEs in estimation of hydrological variables using a hydrology model, and 3) understand the usability of satellite-derived QPEs for eight cases of TC and its impact on the estimate of hydrological variables. The QPEs include near real-time and post-processed satellite data from NASA’s Global Precipitation Mission-Integrated Multi-sensor Retrievals for GPM Rainfall Product (IMERG), merged ground radar observations (Stage IV) from the National Centers for Environmental Prediction (NCEP), and interpolated gauge observations from the National Weather Service Cooperative Observer Program (GCOOP). The study quantifies how differences in rainfall intensity and location, as derived from these gridded precipitation datasets, impact surface hydrology. The Variable Infiltration Capacity (VIC) model and the geographic information system (GIS) routing assess the propagation of bias in the daily rainfall rate to total runoff, evapotranspiration, and flooding. The analysis covers eight tropical cyclones, including Hurricane Charley (2004), Hurricane Frances (2004), Hurricane Jeanne (2004), Tropical Storm Fay (2008), Tropical Storm Beryl (2012), Tropical Storm Debby (2012), Hurricane Irma (2017) and Hurricane Michael (2018) focusing on different regions in South-Atlantic Gulf region and land uses. The findings indicate that IMERG underpredicts precipitation at higher quantiles but aligns closely with ground-based and radar-based products at lower quantiles. IMERG reliably estimates total runoff and evapotranspiration in 90% of TC scenarios along the track and in agricultural and forested regions. There is substantial overlap ~ 70% between IMERG and GCOOP/Stage IV for the 90th percentile rainfall spatially for the case of TC Beryl 2012. Despite previous perceptions of underestimation, the study suggests that satellite-derived rainfall products can be valuable in simulating streamflow, particularly in data-scarce regions where ground estimates are lacking. The relative error in estimation is 12% and 22% when using IMERG instead of Stage IV and GCOOP rainfall data. The findings contribute to a broader perspective on usability of IMERG in estimating near real-time hydrological characteristics, paving the way for further research in this area. This analysis demonstrates that IMERG can be a reliable data product for hydrological studies even in the extreme events like landfalling TCs. This will be helpful in improving the preparedness of vulnerable communities and infrastructure against TC-induced flooding in data scare regions.</p>