Human and freshwater ecosystems are intrinsically interconnected. To better design effective policies, modeling tools and valuation methods are necessary to help understanding the complex reciprocal linkages between ecosystem processes and human actions, and coupled human and natural systems (CNHS) sets up a critical paradigm to do so. It is thus of both academic and empirical appeal to integrate reliable economic valuation methods with tools and models from multiple disciplines in order to quantify the feedbacks between human and natural systems and to inform better policy design. Using freshwater resources as an example, this dissertation contains three essays which integrate natural science and economics models to understand how changes in human behavior and societal policies lead to changes in ecosystem services, and how changes in ecosystem services, in return, affect human decisions. The first two essays focus on agricultural nonpoint source pollution problems in United States and examines the impacts of potential water polices on both water polluters and water demanders. Specifically, in the first essay, a novel coupling between an ecological model of within-lake hydrodynamics and an economic model of hedonic property prices has been developed to quantify the connections between nutrient loading, lake water quality, and economic outcomes. Linking ecological processes with human decision-making provides a basis for enhanced evidence-based decision making in the context of reducing nonpoint-source pollution. In the second essay, an economic mathematical programming model is coupled with an agro-ecosystem model to investigate the behavioral adjustments and environmental pollution outcomes of water quality policies. A complete quantification of costs from all regulating sources are necessary to help pinpoint the efficient water policy design and reflecting the connection between human decisions and ecosystem processes. The third essay focus on the water quantity problem in another developed country, Australia. A discrete choice experiment method has been explored and used to provide estimates of willingness to pay for purchasing irrigation rights to restore a Ramsar-convention wetland. Water policy scenario described in this essay could directly affect the feedback between human and ecosystem processes and serve as a baseline for future planning and policy designs. By offering both conceptual and methodological advancements, this dissertation aims to improve the understanding of coupled human and natural systems and the implementation of water policies. This dissertation also provides a framework to establish multi-disciplinary dialogues and cooperation between scientists and economists in the search of efficient water polices. / Doctor of Philosophy / Freshwater resources are one of the most important elements in our daily life. It provides important goods and services to our society, but at the same time, due to human behaviors, freshwater resources are under threat in both their quality and quantity. This dissertation contains three essays which integrate knowledge from multiple disciplines to help understand and quantify the linkages between human and freshwater resources, and provides information to come up with better water polices. In the first essay, I explore the connections between nutrient loading, lake water quality, and the economic outcomes. The essay illustrates how potential change in nutrient loadings affect lake water quality, and how that induces people’s housing purchase behavior, property sales price and local governments’ property tax revenue. In the second essay, I focus on the agricultural production problem, which is one of the largest source for water quality degradation. By exploring the impacts of water policy on farmers’ production decisions, the essay sheds light on how to better design water polices to maintain farmers’ profit while simultaneously alleviating the impact of agricultural production to water qualities. In the third essay, I utilize a survey method, choice experiments, to elicit people’s willingness to pay for wetland ecosystem health. This could better allocate water resources between agricultural production use and residential use and come up with better water quantity polices.

Doctor of Philosophy / Department of Biological & Agricultural Engineering / Kyle R. Douglas-Mankin / Water quality trading (WQT) is a market-based approach to improve water quality. It is an innovative, voluntary program that connects point source (PS) dischargers who need to reduce their pollutant loads with land managers who could offset those loads with nonpoint source (NPS) reductions to economically achieve water quality improvements in a watershed. The potential issues impeding WQT are its inability to address trading risks and quantify the uncertainty of potential load reduction in trades between PS and NPS. Recent research has also shown that trading information level and transaction costs cause problems in implementing WQT. Therefore, the goals of this study were to quantify the uncertainties of pollutant load reduction and delivery effect for potential trades, to estimate their spatiotemporal variations, and to provide information for stakeholders to reduce intangible costs of WQT. This study simulated agricultural cropland with more than 225 alternative land management practices to identify trends among these scenarios. Both total nitrogen and total phosphorus loads were modeled with SWAT and EUTROMOD for 36 years to analyze the potential load reduction, in-field uncertainty ratio, in-stream delivery ratio, and overall trading ratio (TR) in Lower Kansas watershed, Kansas. The analyses of site-specific effects in both geospatial and temporal aspects were also applied on subbasin level. The variant loading patterns and time distributions of each subbasin showed strong site-specific phenomena. The ANOVA of in-field nutrient load showed significant differences among the design criteria of scenarios. The results also showed a significant delivery and lake effects within the subbasins. The overall TR ranged from 1 to 2.2 or more in different scenarios. The advanced cluster analysis presented a potential method to eliminate the problems involved in fixed TRs while keeping the method simpler than finer-resolution floating TR system. Based on WQT geospatial data model, a three-tier GIS-based web interface Water Quality Trading Information Platform System (WQTIPS) was then developed for WQT information and assessment. A case study demonstrated WQTIPS can provide systematic, spatially information for stakeholders to assess the potential environmental benefit changes from the land management shifts using a simple interface. This study demonstrated that it is possible to automate water-quality trades, use watershed models to minimize trading risk and maximize water-quality benefits, and prioritize among possible trades both spatially and by BMP.

The State of Kuwait has three large refineries, namely, Mina Al Ahmadi Refinery, Mina Abdullah Refinery and Shuaiba Refinery. These refineries process and refine Kuwait Crude Oil through different process units. There are many heaters, boilers and flares that form a part of the complicated equipment which enhance the different petroleum processes. Fuel gas is used as a firing fuel for those heaters and boilers. As a result, stack emissions like SO2, NOx, CO are predominantly present in the flue gases and this study focuses on those emissions and their impact on the surrounding residential area. The area of interest will be Umm Al-Hyman residential area. This study accumulates emission inventories from the three refineries and the respective ground level concentration of the pollutants in the neighboring residential area. It also focuses on the impact of emissions from the refinery operations on the ground level concentrations in the surrounding areas by using the inventory model and latest emission factors to provide accurate emission estimates. The models were developed and the results were verified with the actual data from the area of impact. As a result of the findings of the major pollutants, namely SO2, NOx & CO, it is found that SO2 and CO are not exceeding Kuwait EPA Ambient Air Quality Standards for Residential Areas normally. However, NOx is observed to exceed occasionally. Even though, NOx emissions from refineries sources represented by plume models were much less, there is a consistent increase in the measured NOx. Furthermore, in 2007, the measured hourly, daily and annual NOx concentration exceeded the international standard many times. The increasing trend in NOx is attributed to continuous increase in population and the number of motor vehicles. The study will go further step in recommending engineering solutions and best practices to reduce the pollutants concentrations which will help in the reduction of human health risks and protect the environment.

Tidal freshwater reaches of estuaries may play an important role in mitigating nutrient fluxes from watersheds to the coastal zone due to their location at the interface between riverine and estuarine systems. We developed annual N and P budgets for the tidal, freshwater James River over 4 calendar years (2007-2010) taking into account riverine inputs at the Fall Line, local points sources (including CSO events), ungagued inputs, riverine outputs, and tidal exchange. The tidal freshwater James River experiences high areal loading rates of TN (383 mg/m2/d) and TP (70 mg/m2/d) due to the combined effects of large watershed area and local point source discharges. On an annual basis, riverine sources dominated TN and TP inputs (59% and 84%, respectively), whereas during low discharge summer months (May-Oct) point sources were more important. Proportional retention of TP inputs (59±7%) was greater than TN retention (27±4%) with annual absolute retention being 1,800±350 kg TP/d, and 5,900±2,700 kg TN/d. Proportional retention of TN and dissolved inorganic fractions of N and P was highest during the low discharge summer months due to reduced loading rates and increased residence times and biotic activity. TP retention was greatest during high discharge winter months (Nov-Apr) when loading rates were highest. High retention during this period of low biotic activity suggests that trapping of riverine derived particulate-bound P via sedimentation was an important mechanism of P retention. Understanding this seasonal variation in nutrient inputs and retention can help to inform management decisions regarding reducing nutrient inputs to the Chesapeake Bay and improving local water quality.

AMANDA-II ist ein Neutrino-Teleskop, das sich im Eis des Suedpols befindet. Es wurde optimiert um Spuren von hochenergetischen Myonen, die in Neutrino-Wechselwirkungen entstanden sind, anhand ihrer Emission von Cerenkov-Licht zu detektieren. In dieser Arbeit analysieren wir die Daten, die in 1001 Tagen effektiver Detektorlaufzeit in den Jahren 2000 bis 2004 gesammelt worden sind, um ein Signal von einer Neutrino-Punktquelle zu finden. Ein derartiges Signal wird von kosmischen Objekten erwartet, die Hadronen zu sehr hohen Energien beschleunigen, welche daraufhin mit Photonen und Protonen in der Umgebung des Objekts wechselwirken. Die wichtigste Signatur um ein Myon aus einer Neutrino-Wechselwirkung zu identifizieren ist eine nach oben laufende Spur. Ein Datensatz mit 4282 aufwaertslaufenden Ereignissen wurde aus den ca. 10 Milliarden im Zeitraum dieser Analyse registrierten Ereignissen extrahiert. Diese Zahl ist konsistent mit der erwarteten Anzahl atmosphaerischer Neutrinos. In der Suche nach Punktquellen wird nach einem lokalen Ereignissueberschuss in diesem Datensatz gesucht. Zuerst wird ein Katalog von Quellkandidaten untersucht. Danach wird eine Rastersuche nach unbekannten Quellen auf dem ganzen noerdlichen Himmel durchgefuehrt. Darueberhinaus werden Methoden entwickelt um die Chancen einer Detektion von Quellen zu erhoehen von denen vermutet wird, dass ihre Neutrinoemission hochvariabel ist. Kein signifikanter lokaler Ereignissueberschuss wurde im analysierten Datensatz gefunden. Deswegen berechnen wir obere Grenzen fuer Neutrinofluesse, die mit dieser Beobachtung vertraeglich sind. Die mittlere obere Flussgrenze fuer ein Neutrinospektrum dPhi/dE proportional zu E^-2, die mit dieser Analyse erreicht wird, ist E^2 dPhi/dE = 1.0^-7 GeV cm^-2 s^-1 fuer den aufaddierten Fluss von nu_mu + nu_tau unter der Annahme eines Flavor-Verhaeltnisses von 1:1. Dies entspricht der momentan niedrigsten Flussgrenze fuer Neutrinofluesse von Punktquellen. / AMANDA-II is a neutrino telescope located in the glacial ice at the South Pole. It is optimized to detect neutrino induced muon tracks with energies larger than 100 GeV by their Cerenkov light emission. We analyzed the data collected in 1001 effective days of detector operation between the years 2000 and 2004 for a signal from point-like sources of neutrinos. Such a signal is expected from cosmic objects that accelerate hadrons to very high energies, which subsequently interact with ambient protons or photons. The dominant event class recorded in AMANDA-II are muons produced in the interactions of cosmic rays in the atmosphere. Due to their energy loss, the muons cannot penetrate the Earth and have down-going directions. The main signature to identify a neutrino induced event is therefore its up-going direction. A sample of 4282 up-going events is extracted from the 10 billion events triggered in the period selected for this analysis. The search for point sources is accomplished on this data sample by looking for a localized excess over the isotropic background of atmospheric neutrinos. The procedure is applied for the directions of candidate sources, like Active Galactic Nuclei, Supernova remnants and X-ray binaries. In a second step, the full northern sky is scanned for unknown sources. Further, we investigate methods to enhance the detection chance for sources which are suspected to be highly variable neutrino emitters. No localized excess has been found in the analyzed dataset. Therefore, we calculate upper limits on the neutrino fluxes which are compatible with this observation. The average upper limit achieved for a combined nu_mu + nu_tau flux with a spectrum of dPhi/dE proportional to E^-2 is E^2 dPhi/dE = 1.0 10^-7 GeV cm^-2 s^-1 assuming a flavor ratio of 1:1. It represents the most stringent upper limit on neutrino fluxes from point-like sources reported so far.

Deutsche Zusammenfassung Die vorliegende Dissertation beschäftigt sich einerseits mit der Suche nach atmosphärischen Myonneutrinos und andererseits mit der Suche nach extraterrestrischen Neutrinopunktquellen in dem Datensatz, welcher im Jahre 1997 durch den AMANDA-Detektor aufgenommenen wurde. In dieser Arbeit wird zunächst die kosmische Strahlung eingeführt. Die Suche nach den Quellen dieser Strahlung wurde bisher insbesondere mit Hilfe der geladenen kosmischen Strahlung selber, sowie mit Hilfe von Cherenkovteleskopen für Gammastrahlen durchgeführt. Die mit diesen Techniken gewonnenen Erkenntnisse lassen bisher jedoch noch keine eindeutigen Schlüsse über die Quellen der geladenen kosmischen Strahlung zu. Dies motiviert den Versuch, mit Hilfe der Neutrinoastronomie ein neues Fenster zu den Quellen zu eröffnen. Es gibt theoretische Modelle für verschiedene potentielle Neutrinoquellen. Szenarien, in denen massive Teilchen zerfallen und die Zerfallsprodukte ihre kinetische Energie aus der freigewordenen Ruhemasse gewinnen, spielen in dieser Arbeit eine untergeordnete Rolle. Ausführlicher dargestellt werden die konventionellen Modelle, in denen geladene Teilchen mit Hilfe der sogenannten Fermibeschleunigung in astrophysikalischen Schockwellen und/oder Magnetfeldern beschleunigt werden. Hochenergetische Neutrinos entstehen nur bei Quellen, welche Hadronen beschleunigen. Insbesondere die Klasse der aktiven galaktischen Kerne (AGNs) sind hierbei interessant. Die vor einigen Jahren entwickelten vereinheitlichten AGN-Schemata sind ein wichtiger Schritt auf dem Weg, diese Objekte zu verstehen. Andere potentielle Quellen sind Supernovae und ihre überreste, Mikroquasare, sowie die Quellen hochintensiver Gammastrahlenausbrüche (GRBs). Hoch spekulativ sind Quellen, welche im elektromagnetischen Spektrum unsichtbar sind, oder auch bisher vollkommen unerwartete Quellen. Unabhängig von der genauen Natur möglicher Neutrinoquellen müssen für die Beschreibung der von ihnen ausgesendeten Neutrinoflüsse Oszillationseffekte zwischen den verschiedenen Neutrinofamilien berücksichtigt werden. Der Nachweis der hochenergetischen Neutrinos soll mit dem AMANDA-Detektor oder ähnlichen Teleskopen erfolgen. Das derzeitige AMANDA-Teleskop AMANDA-II wurde in den Jahren 1995 bis 2000 aufgebaut. Es basiert auf dem Nachweis von neutrinoinduzierten Myonen mit Hilfe des Cherenkoveffektes. Das Cherenkovlicht wird hierbei von, in einem Gitter angeordneten, großen Sekundärelektronenvervielfachern registriert. Die gewonnene Zeitinformation erlaubt eine Richtungsrekonstruktion. Das Charakteristikum für ein Neutrinoereignis ist ein aufwärts laufendes Myon, da das Neutrino das einzige bekannte Teilchen ist, welches die Erde durchqueren und ein aufwärts laufendes Myon erzeugen kann. Die vorliegende Arbeit untersucht Daten, die mit Hilfe des AMANDA-B10-Detektors im Jahre 1997 genommen wurden. Die Daten bestehen aus etwa $10^9$ atmosphärischen Myon- und etwa 5000 atmosphärischen Neutrinoereignissen. Um die experimentell gewonnenen Daten analysieren zu können, wird der Vergleich zu simulierten Daten benötigt. Im Rahmen dieser Arbeit wurden Flüsse atmosphärischer Myonen mit den Programmen basiev und Corsika, Flüsse neutrinoinduzierter Myonen mit nusim und nu2mu generiert. Während eine hohe Zahl von neutrinoinduzierten Myonen simuliert werden konnte, blieb die Zahl der simulierten atmosphärischen Myonen weit hinter der Zahl der experimentell gemessenen zurück. Die Propagation der simulierten Myonen sowie die Detektorsimulation erfolgte mit den Programmen mudedx bzw. amasim. Die Zeit-, Orts- und Amplitudeninformationen der gemessenen -- wie auch der simulierten -- Daten wurden anschließend kalibriert. Das erste Ziel der Analyse war die Extrahierung eines Satzes atmosphärischer Neutrinos und ein Verständnis der absoluten Sensitivität des Detektors. Die hierfür notwendige Prozessierung der Daten erfolgte in mehreren Schritten. Dabei wechselten sich immer exaktere (aber auch immer langsamere) Richtungsrekonstruktionen mit immer strengeren Qualitäts- und Winkelschnitten zur Datenreduktion ab. Die Rekonstruktionen bestanden sowohl aus schnellen analytischen Richtungsapproximationen, wie auch aus solchen, die auf langsamen Minimierung von Likelihoodfunktionen basierten. Die Schnitte wurden auf topologische Grössen, wie auch auf Parameter, welche aus der Rekonstruktion gewonnen wurden, angewandt. Die Schnitte waren notwendig, um Neutrinoereignisse aus dem weit zahlreicheren Untergrund atmosphärischer Myonenereignisse herauszufiltern. Es wurden zwei stark reduzierte Datensätze ("BG-10" mit 223 Ereignissen und "BG-100" mit 369 Ereignissen) aus den gemessenen Daten extrahiert. Es konnte gezeigt werden, dass hiervon ca. 15 bzw. 100 Ereignisse durch atmosphärische Myonen bedingt waren. Die Ergebnisse stimmen sehr gut mit der Erwartung für atmosphärische Neutrinos überein, wobei die Erwartung eine Unsicherheit von bis zu 63% aufweist. Mit dem BG-10 Datensatz war somit das erste Ziel der Analyse erfüllt. Der BG-100 Datensatz sollte dem zweiten Ziel dieser Arbeit dienen: der Suche nach extraterrestrischen Neutrinoquellen. Für dieses zweite Ziel musste zunächst mit der effektiven Fläche} ein Maß für die Sensitivität des Detektors bezüglich extraterrestrischer Neutrinos eingeführt werden. Anschließend wurde die Güte der Richtungsrekonstruktion bestimmt. Hiermit konnte die optimale Grösse der zu benutzenden Suchfenster festgelegt werden. Für diese Suchfenster wurde dann die Effizienz der Rekonstruktion bestimmt. Die Effizienz ist ein Maß für den Anteil der Neutrinoereignisse, für den die Rekonstruktion korrekt bestimmt, aus welchem Suchfenster sie stammen. Nach diesen vorbereitenden Untersuchungen konnte die Quellsuche beginnen. Für diese Suche waren nun sowohl atmosphärische Myonereignisse, als auch Ereignisse, die durch atmosphärische Neutrinos hervorgerufen wurden, Untergrund. Für die Suche wurden drei verschiedene Strategien angewendet. Zunächst wurde ein Netz aus 374 aneinander angrenzenden Suchfenstern konstruiert. Basierend auf der erwarteten Anzahl von Untergrundereignissen für jedes Suchfenster wurden Wahrscheinlichkeiten berechnet, dass die Ereignisse ausschließlich Untergrundereignisse sind. Durch die große Zahl an Suchfenstern gab es einige Fenster, bei denen diese Wahrscheinlichkeit recht gering war. Insgesamt jedoch gab es keinen signifikanten Hinweis darauf, dass die Messung nicht auschließlich durch Untergrund erklärt werden kann. Die zweite Strategie bestand in dem Versuch, mit Hilfe einer Clusteranalyse Punktquellen zu finden. Auch hier wurden keine Hinweise auf Punktquellen gefunden. Schließlich wurde in Richtung von 62 vorselektierten potentiellen Quellen nach Ereignisüberschüssen gesucht -- ebenfalls ohne ein positives Ergebnis. Daraufhin wurden obere Flussgrenzen abgeleitet. Diese Grenzen wurden sowohl richtungsunabhängig als auch für die zuvor selektierten potentiellen Quellen berechnet. In beiden Fällen wurde dabei für das Quellspektrum ein spektraler Index gamma = -2 angenommen. Für Neutrinoenergien E > 10 GeV und Deklinationen > 33 Grad wurden integral folgende globale, obere Flussgrenzen berechnet: Myonfluss: 1.41 * 10^-14 cm^-2 s-1 und Neutrinofluss: 1.65 * 10^-7 cm^-2 s^-1 . Nach der "Eichung" am Fluss atmosphärischer Neutrinos konnte die systematische Unsicherheit auf diese Grenzen zu 46% (systematisch) plus 7% (statistisch) abgeschätzt werden. Für die 62 ausgewählten Quellen wurden individuelle Flussgrenzen berechnet. Diese waren im Durchschnitt etwa einen Faktor drei besser als die integralen Grenzen für den entsprechenden Deklinationsbereich. Bei 48 potentiellen Quellen waren dies sowohl die ersten Grenzen auf ihren Neutrino- als auch die ersten Grenzen auf ihren neutrinoinduzierten Myonenfluss. Für eine weitere Quelle konnte erstmals eine Grenzen auf den Neutrinofluss abgeleitet werden. Bei den 14 restlichen Quellen konnten in fünf Fällen beide bisher publizierten Grenzen verbessert werden, in zwei weiteren zumindest die Grenze auf den Neutrinofluss. Im Anhang werden Daten bereitgestellt, mit denen die errechneten Grenzen auch in Grenzen für andere spektrale Indizes umgerechnet oder auch Grenzen für weitere Quellen abgeleitet werden können. / Abstract The young field of high energy neutrino astronomy can be motivated by the search for the origin of the charged cosmic rays. Large astrophysical objects like AGNs or supernova remnants are candidates to accelerate hadrons which then can interact to eventually produce high energy neutrinos. Neutrino-induced muons can be detected via their emission of Cherenkov light in large neutrino telescopes like AMANDA. More than 10^9 atmospheric muon events and approximately 5000 atmospheric neutrino events were registered by AMANDA-B10 in 1997. Out of these, 223 atmospheric neutrino candidate events have been extracted. This data set contains approximately 15 background events. It allows to confirm the expected sensitivity of the detector towards neutrino events. A second set containing 369 events (approximately 270 atmospheric neutrino events and 100 atmospheric muon events) was used to search for extraterrestrial neutrino point sources. Neither a binned search, nor a cluster search, nor a search for preselected sources gave indications for the existence of a strong neutrino point source. Based on this result, flux limits were derived. Assuming E^-2 neutrino spectra, typical flux limits for selected sources of the order of 10^-14 cm^-2 s^-1 for muon fluxes and 10^-7 cm^-2 s^-1 for neutrino fluxes have been obtained.

Sampling data of many surface water bodies in Broward County tend to reveal abnormal levels of bacterial concentrations for total coliform, Escherichia coli and Enterococcus. Given the presence of septic systems in geohydrologically sensitive areas throughout the County, several studies (Morin 2005 and Bocca 2007) suggested a correlation in the high bacterial densities and the septic sites. / Includes bibliography. / Thesis (M.S.)--Florida Atlantic University, 2014. / FAU Electronic Theses and Dissertations Collection

The Finite Difference Time Domain (FDTD) Method has been a powerful tool in numerical simulation of electromagnetic (EM) problems for decades. In recent years, it has also been applied to biomedical research to investigate the interaction between EM waves and biological tissues. In Wireless Body Area Networks (WBANs) studies, to better understand the localization problem within the body, an accurate source/receiver model must be investigated. However, the traditional source models in FDTD involve effective volume and may cause error in near field arbitrary direction. This thesis reviews the basic mathematical and numerical foundation of the Finite Difference Time Domain method and the material properties needed when modeling a human body in FDTD. Then Coincident Phase Centers (CPCs) point sources models have been introduced which provide nearly the same accuracy at the distances as small as 3 unit cells from the phase center. Simultaneously, this model outperforms the usual sources in the near field when an arbitrary direction of the electric or magnetic dipole moment is required.

In Taiwan, non-point source (NPS) pollution is one of the major causes of the impairment of surface waters. NPS pollutants, which are associated with stormwater runoff from agricultural land uses can be quite diffuse and difficult to treat. The I-Liao Creek Basin, located in southern Taiwan, flows through approximately 90-km and drains towards the Kaoping River. It is one of the major sub-basin in the Kaoping River watershed, which is the largest and the most intensively used watershed in Taiwan. Field investigation results indicate that the main water pollution sources in the I-Liao Creek Basin are domestic wastewater and NPS pollutants from agricultural activities. In this study, an Integrated Watershed Management Model (IWMM) was applied for simulating the water quality in the I-Liao Creek watershed. The model includes a global atmosphere module, a land module, a human impact module, a canopy module, and a global ocean module. Those modules can be linked and managed by a graphic user-interface. The model was calibrated and verified with field data, and was used to investigate potential NPS pollution management plans. Moreover, the Storm Water Management Model (SWMM) was used to verify the accuracy of the simulated results of flow and water qualities. Results from this study show that geographical information system (GIS) is an important mean for land-use identification and waste load estimation in the catchment. Linking the information of land utilization with the NPS pollution simulation model may further provide essential information of pollution potential of NPS pollution for all sub-regions in the river basin. Results and experience obtained from this study will be helpful in designing the watershed management and NPS pollution control strategies for other similar river basins.

Polybrominated diphenyl ethers (PBDEs), flame retardants, have been applied to consumer goods, such as furniture, electrical devices, textiles, and appliances for decades. Due to their physico-chemical properties, PBDEs are semi-volatile and easily leach off the consumer good during aging, stress, or normal wear and tear of the good. Once airborne, they pose an environmental health threat because they can adsorb onto dust particles, soil, or other particulates that can be inhaled, ingested, or come into contact with the dermal layer. Additionally, PBDEs have a molecular structure similar to other persistent organic pollutants, such as polychlorinated biphenyls and polychlorinated dibenzo-p-dioxins and furans. They are a health threat due to their endocrine-disrupting nature by affecting thyroid functioning, fertility, and child development. The purpose of the study is to measure selected PBDEs in a wastewater treatment plant (WWTP) that produces reclaimed water, such that a mass balance can be completed, and to compare this mass balance with theoretically expected concentrations. The mass balance includes the collection of samples from wastewater, sewage sludge, and air at points within the WWTP. The PBDEs examined are BDE-28, 47, 99, 100, 153, 154, and 183. The second part of the study will compare effluent concentrations to reclaimed water concentrations in order to examine the potential exposure (if any) of using reclaimed water. Influent concentration of mean (sum7)PBDE was found to be 49,117 pg/L and effluent concentration was 4,603 pg/L, illustrating a 91% removal rate of PBDEs during the wastewater treatment plant. Sludge samples contained the highest total concentrations of PBDEs with mean (sum7)PBDE ranging from 14.0 to 41.3 ug/kg dry weight. Air samples were highest at the post-aeration (248 pg/m3 mean (sum7)PBDE) step due to the use of highly oxygenated air assisting in the release and volatilization of the PBDEs. Sludge was found to carry the largest mass loading at 14.2 lb/day (sum7)PBDE. Of the total mass loading of PBDEs from the WWTP, sludge is responsible for 86.7%, followed by reclaimed water and effluent (11.7% and 1.6%, respectively). The mass loading from air was negligible with less than 0.01% contribution to the total mass loading. Whereas reclaimed water overall had higher PBDE congener mean concentrations than the effluent, the independent samples t-test found no statistically significant differences between the two groups. The results of this study can be used to improve the wastewater treatment process to reduce the impact of PBDEs being released into the environment by WWTPs, and to educate the public on utilizing reclaimed water in a safe and healthy manner.