Browsing by Subject "Environmental engineering"

Spatial variability of the Earth's surface has a considerable impact on the atmosphere at all scales and understanding the mechanisms involved in land-atmosphere interactions is hindered by the scarcity of appropriate observations. A measurement gap exists between traditional point sensors and large aircraft and satellite-based sensors in collecting measurements of atmospheric quantities. Point sensors are capable of making long time series of measurements, but cannot make measurements of spatial variability. Large aircraft and satellites make measurements over large spatial areas, but with poor spatial and temporal resolution. A helicopter-based platform can make measurements on scales relevant for towers, especially close to the Earth's surface, and can extend these measurements to account for spatial variability. Thus, the Duke University Helicopter Observation Platform (HOP) is designed to fill the existing measurement gap.

Because measurements must be made in such a way that they are as uncontaminated by the platform itself as much as is possible, it is necessary to quantify the aerodynamic envelope of the HOP. The results of an analytical analysis of the location of the main rotor wake at various airspeeds are shown. Similarly, the results of a numerical analysis using the commercial Computational Fluid Dynamics software Fluent are shown. The optimal flight speed for the sampling of turbulent fluxes is found to be around 30 m/s. At this airspeed, the sensors located in front of the nose of the HOP are in advance of the wake generated by the main rotor. This airspeed is also low enough that the region of high pressure due to the stagnation point on the nose of the HOP does not protrude far enough forward to affect the sensors. Measurements of differential pressures, variables and turbulent fluxes made while flying the HOP at different airspeeds support these results. No systematic effects of the platform are seen at airspeeds above about 10 m/s.

Processing of HOP data collected using the current set of sensors is discussed, including the novel use of the Empirical Mode Decomposition (EMD) to detrend and filter the data. The EMD separates the data into a finite number of Impirical Mode Functions (IMFs), each of which is unique and orthogonal. The basis is determined by the data itself, so that it need not be known a priori, and it is adaptive. The EMD is shown to be an ideal tool for the filtering and detrending of HOP data using data gathered during the Cloud and Land Surface Interaction Campaign (CLASIC).

The ability of the HOP to accurately measure atmospheric profiles of potential temperature is demonstrated. During experiments conducted in the marine boundary layer (MBL) and the convective boundary layer (CBL), HOP profiles are evaluated using profiles from an elastic backscatter lidar. The HOP and the lidar agree on the height of the boundary layer in both cases, and the HOP effectively locates other atmospheric structures.

Atmospheric sensible and latent heat fluxes, turbulence kinetic energy (TKE) and horizontal momentum fluxes are also measured, and the resulting information is used to provide context to tower-based data collected concurrently. A brief comparison made over homogeneous ocean conditions yields good results. A more exhaustive evaluation is made using short HOP flights made over an orchard during the Canopy Horizontal Turbulence Study (CHATS).

The objective of this research is to develop an open-source Python-based surface complexation modeling program to estimate the acidity of biological surface. Several computer software already exists with such function installed, such as ProtoFit and FITEQL. However, these programs lack capabilities in constraining fitting parameters, resulting in model fits that are not necessarily justified by the input data. Here, a new Python-based model algorithm was developed to estimate surface acidity and protonation constants for biological surfaces. The program was developed based on the algorithm of ProtoFit, and improved to allow for user-defined boundary conditions for model fitting parameters. This model was tested on potentiometric pH titration data for suspensions of Pseudomonas fluorescens and Bacillus subtilis bacterial cells and suspensions of colloidal particles (e.g., extracellular vesicles) that were isolated from cell cultures. Model testing was also performed for titration data collected for aqueous buffer solutions with known chemical species and concentration. The estimated surface acidities from Python script and ProtoFit are compared, and error analysis was conducted. Error analysis showed that the Python script modeled the titration data with lower curve-fitting error than models by ProtoFit, which suggests a better optimization performance in Python script. However, the model comparisons for the aqueous buffer titrations (for which acidity constants were known) did not showing such trend. We believe that is because experimental error is much larger than model error in our setups. Therefore, variance-based sensitivity analysis was further conducted on the Python script, and the result shows that the titrant concentration (N_tit) and adsorbent mass (M_ads) were two variables that contributed the most variance in our model output.

In recent years, NaTech disasters, which are defined as the technological accidents caused by natural events, have led to huge losses all over the world. To reduce these losses, assessments of the vulnerabilities of industrial facilities are necessary. In this study, an effort was made to locate aboveground storage tanks from remotely sensed imagery. A dataset that identifies different types of tanks was generated. The data were acquired from the National Agriculture Imagery Program (NAIP) and tanks were labeled as closed roof tank, external floating roof tank, spherical tank, water treatment tank, or water tower according to their shapes. After collecting these data, the Faster R-CNN algorithm, an object detection architecture, was applied to test the performance of this algorithm on the prelabeled dataset. Results of testing indicate that the algorithm could well achieve the goal that having a high recall rate for all the classes of tanks. The precision and recall rates were 82.92% and 90.03% for closed roof tanks, 85.85% and 91.68% for external floating roof tanks, 34.81% and 60.26% for spherical tanks, 49.63% and 89.33% water treatment tanks, 9.43% and 38.46% for water towers. For spherical tanks and water towers, although having low precision and recall, the percentage of missed tanks was extremely low, which is 2.08% and 0 respectively. These results suggest that this aboveground storage tank dataset and the pretrained model generated from Faster R-CNN could be further used in future work for tank detection and vulnerability assessment.

Powdered Activated Carbon (PAC) adsorption was studied in order to determine the influence of natural organic matter (NOM) on the adsorption of two acidic pharmaceutically active compounds (PhACs), clofibric acid and ketoprofen. Suwannee River humic acids (SRHAs) was used as substitute of NOM in natural water. Batch adsorption experiments were conducted to obtain the single compound adsorption kinetics and adsorption isotherm with and without SRHAs in the system. Three main findings resulted from this study. First, the adsorption isotherms showed that the adsorption of clofibric acid was not significantly affected in the presence of SRHAs (5 ppm); however, the adsorption of ketoprofen markedly decreased with SRHAs in the solutions. Higher initial concentrations of clofibric acid than ketoprofen together with the compressed double layer theory helped explain the different behaviors that were observed. Furthermore, the more hydrophobic ketoprofen molecules may increase the possibility that this compound would adsorb less on the surface area which was covered by the more hydrophilic humic acids. Second, the adsorption kinetics of both compounds were not affected by the SRHAs, although more research may be needed, as it is possible that slight differences exist during the initial adsorption phase. Lastly, possible intermolecular forces were discussed and a sequence of importance is proposed for their role in the adsorption process as A). electrostatic forces; B). electron donor-acceptor interaction; C & D). H-bond and London Dispersion forces.

Oil spills in marine environments often damage marine and coastal life if not remediated rapidly and efficiently. In spite of the strict enforcement of environmental legislations (i.e., Oil Pollution Act 1990) following the Exxon Valdez oil spill (June 1989; the second biggest oil spill in U.S. history), the Macondo well blowout disaster (April 2010) released 18 times more oil. Strikingly, the response methods used to contain and capture spilled oil after both accidents were nearly identical, note that more than two decades separate Exxon Valdez (1989) and Macondo well (2010) accidents.

The goal of this dissertation was to investigate new advanced materials (mechanically strong aerogel composite blankets-Cabot® Thermal Wrap™ (TW) and Aspen Aerogels® Spaceloft® (SL)), and their applications for oil capture and recovery to overcome the current material limitations in oil spill response methods. First, uptake of different solvents and oils were studied to answer the following question: do these blanket aerogel composites have competitive oil uptake compared to state-of-the-art oil sorbents (i.e., polyurethane foam-PUF)? In addition to their competitive mechanical strength (766, 380, 92 kPa for Spaceloft, Thermal Wrap, and PUF, respectively), our results showed that aerogel composites have three critical advantages over PUF: rapid (3-5 min.) and high (more than two times of PUF’s uptake) oil uptake, reusability (over 10 cycles), and oil recoverability (up to 60%) via mechanical extraction. Chemical-specific sorption experiments showed that the dominant uptake mechanism of aerogels is adsorption to the internal surface, with some contribution of absorption into the pore space.

Second, we investigated the potential environmental impacts (energy and chemical burdens) associated with manufacturing, use, and disposal of SL aerogel and PUF to remove the oil (i.e., 1 m3 oil) from a location (i.e., Macondo well). Different use (single and multiple use) and end of life (landfill, incinerator, and waste-to-energy) scenarios were assessed, and our results demonstrated that multiple use, and waste-to-energy choices minimize the energy and material use of SL aerogel. Nevertheless, using SL once and disposing via landfill still offers environmental and cost savings benefits relative to PUF, and so these benefits are preserved irrespective of the oil-spill-response operator choices.

To inform future aerogel manufacture, we investigated the different laboratory-scale aerogel fabrication technologies (rapid supercritical extraction (RSCE), CO2 supercritical extraction (CSCE), alcohol supercritical extraction (ASCE)). Our results from anticipatory LCA for laboratory-scaled aerogel fabrication demonstrated that RSCE method offers lower cumulative energy and ecotoxicity impacts compared to conventional aerogel fabrication methods (CSCE and ASCE).

The final objective of this study was to investigate different surface coating techniques to enhance oil recovery by modifying the existing aerogel surface chemistries to develop chemically responsive materials (switchable hydrophobicity in response to a CO2 stimulus). Our results showed that studied surface coating methods (drop casting, dip coating, and physical vapor deposition) were partially successful to modify surface with CO2 switchable chemical (tributylpentanamidine), likely because of the heterogeneous fiber structure of the aerogel blankets. A possible solution to these non-uniform coatings would be to include switchable chemical as a precursor during the gel preparation to chemically attach the switchable chemical to the pores of the aerogel.

Taken as a whole, the implications of this work are that mechanical deployment and recovery of aerogel composite blankets is a viable oil spill response strategy that can be deployed today. This will ultimately enable better oil uptake without the uptake of water, potential reuse of the collected oil, reduced material and energy burdens compared to competitive sorbents (e.g., PUF), and reduced occupational exposure to oiled sorbents. In addition, sorbent blankets and booms could be deployed in coastal and open-ocean settings, respectively, which was previously impossible.

In the spirit of American individualism, the majority of the United States’ forested landscape is controlled by private landowners, who make autonomous decisions that impact a shared wealth of biodiversity and ecosystem services. It is important to understand not only the forest management decisions made by private landowners, but also the motivations that incentivize these consequential actions. Furthermore, it is useful to have the capacity to infer such insights using publicly available data, and by employing transparent, flexible, and scalable statistical frameworks. This dissertation seeks to elucidate the motivations and actions of private landowners in the United States using a variety of data sources, including Zillow home estimates, the American Community Survey, satellite remote sensing imagery, and the Forest Inventory and Analysis database, and by implementing interpretable modeling frameworks, such as the hedonic pricing method and structural equation modeling. I uncover nuanced insights about human-environmental systems, including (1) a positive feedback loop between affluence and tree-shading in metropolitan areas; (2) the dominance of normative pressures on forest owners’ harvest intentions; and (3) a causal link between invasive insects and the quantity and sizes of harvested trees. Understanding such relationships benefits policymakers, forest managers, and urban planners tasked with optimizing human-natural systems.

Ammonia is an odorous gaseous compound emitted by a variety of industrial facilities. This study aimed to address the feasibility of ammonia gas removal using a biotrickling filter (BTF) coupled with an anammox bioreactor. In the BTF, the influent ammonia gas partitioned into the trickling water and was converted to nitrite via partial nitrification. The effluent liquid from the BTF, containing nitrite and ammonium concentrations, was fed into the anammox reactor where autotrophic denitrifying bacteria converted the ammonium and nitrite to dinitrogen gas. For the anammox reactor to operate efficiently, the influent ammonium and nitrite concentrations must be in a 1 to 1 molar ratio. To evaluate the feasibility of this system, a lab scale BTF and anammox reactor were constructed and operated and a conceptual model for this system was developed. To obtain a nitrite to ammonium ratio close to 1, it was found that the effluent pH from the BTF must be maintained below 7, and the loading rate could not exceed 8.7 g N/m3h. At this loading rate, complete ammonia gas removal occurred. A recycle rate of 1.4 times that of the influent was implemented in the BTF to increase performance and improve the nitrite to ammonium ratio. The addition of the recycle line achieved a nitrite of ammonium ratio of 0.97 at a pH value of 7.67. The anammox reactor achieved 88% removal of ammonium and nitrite at a loading rate of 10.5 g N /m3h. The fact that the BTF was able to achieve a 1 to 1 nitrite to ammonium ratio indicated that coupling of a BTF with the anammox reactor should be feasible. The mathematical model underpredicted effluent ammonium and nitrite concentrations in the BTF and greatly overpredicted the effluent concentrations from the anammox reactor. To improve the BTF model inhibition factors and oxygen supply need to be accounted for. Further development of the growth kinetics in the annamox model are necessary as well.

The settling of inertial particles in turbulence boundary layers plays an essential role in many meteorological, industrial and environmental processes, and is governed by multifarious mechanisms. First, turbulence alters the settling velocity of inertial particles through different effects, like preferential sweeping mechanism, loitering effect and vortex trapping. Second, the existence of a wall introduces extra effects that can influence particle settling, such as turbophoresis. The Rouse model was the most famous model in predicting particle settling in vertical wall-bounded settling. Nevertheless, it is only valid for inertia-less particles in the logarithmic region. A theory by Bragg et al., based on phase-space probability density theory, incorporates particle inertia into the Rouse model, and quantifies the contributions from the aforementioned mechanisms to the particle vertical velocity. The theory is valid for all particle Stokes numbers, yet it still lacks a closed form.In this work, one way to close the equations presented by Bragg et al. (the extended Rouse model) was examined. Using a central differencing scheme combined with an iterative method, the nonlinear second-order differential equation of the variance of vertical particle velocity was solved. The predictions of the variance of vertical particle velocity S and the particle concentration PDF ρ by the model were studied and compared to DNS. The comparison indicates that the extended Rouse model is able to predict many features of S and ρ, like the accumulation of particles close to the wall and turbophoretic drift. However, the quantitative agreement between the predictions by the model and DNS is poor. There are two probable reasons for the discrepancies between the predictions and DNS. First, the closure of the term in the equation may be a source of errors. Second, the lower boundary condition, whose validity is suspicious for particles with weak inertia, can be a reason for the discrepancies. In order to investigate the cause for the disagreement, three different boundary conditions (zero-gradient condition, asymptotic matching, iterative condition) were examined. The results indicate that the boundary conditions have a very limited influence on the predictions. As a result, the closure of the terms is more likely to be responsible for the discrepancies.

A cost-effective pathway towards net-zero electric power systems requires an extraordinary deployment of new solar and wind generation assets. This aggressive expansion driving unprecedented investment entails a fundamental understanding of the challenge ahead of us. This dissertation seeks to provide a multidisciplinary perspective of research questions that shine the light on rapid and cost-efficient strategies for the energy transition. Integrating operations research and geospatial analysis methods, the dissertation utilizes a multidisciplinary approach when addressing three questions.

First, the dissertation examines the role of battery energy storage technologies (i.e., utility-scale lithium-ion batteries) on reducing the greenhouse gas emissions of an electric power system while simultaneously achieving a reduction in carbon abatement costs. The study uses a cost-based production model (day-ahead unit commitment and a real-time economic dispatch) to simulate the optimal operation of all the generation resources in the largest vertically-integrated electric service region in the U.S. The study explores a multitude of configurations to identify optimal sizing of battery energy storage systems when paired with utility-scale photovoltaics.

Next, the dissertation studies the effect of incorporating high-resolution data when identifying suitable land for renewable energy projects over a geographically defined region. Using a python-based user-friendly siting tool implemented in ArcGIS Pro to perform suitability and cost analysis of utility-scale photovoltaic projects in North Carolina under three scenarios (representing conditions ranging from favorable to restrictive). The study finds that the land suitable for utility-scale photovoltaics reduces substantially when parcel-level data reflecting local land-use restrictions are incorporated. The study's findings highlight the necessity of integrating detailed land-use data that reflects local regulation (zoning ordinances) into siting models while simultaneously increasing their spatial granularity.

Lastly, the dissertation analyzes the benefits of weatherizing wind power farms enabling their operation under extreme climates (winter storms). The study uses global reanalysis data with operational information from the Electric Reliability Council of Texas (ERCOT) during the 2021 winter storm Uri to simulate a continued operation of wind power farms under low-temperature environments. The study finds that the financial benefits that winterized wind turbines would have received during winter storm Uri would have outweighed the capital costs required to implement ice-accretion mitigation actions (before winter storm URI).

Despite significant advances in public health and engineering over the last 100 years, diarrheal disease remains one of the highest global burdens of disease, particularly for children under 5 years of age. Access to clean water, sanitation, and hygiene greatly reduces this risk, but access to improved sanitation remains a challenge for a large percentage of the world. Due to the lack of access to safely managed sanitation and rapid urbanization, sustainable onsite fecal sludge treatment systems need to be developed and deployed to reduce the burden of diarrheal disease.

The Anaerobic Digestion Pasteurization Latrine (ADPL) is a concept that was developed by Professor Marc Deshusses to meet this need. The goal of the ADPL was to produce a pathogen-free effluent through pasteurization powered by biogas produced from anaerobic digestion of fecal sludge. The concept was supported through laboratory studies on the anaerobic digestion of a simulant fecal sludge and inactivation of E. coli in a pasteurization system using a heater maintained at 65-75 °C and a tube-in-shell counter-flow heat exchanger for heat recovery.

The goal of this research was to build upon initial laboratory-based research on the ADPL and demonstrate the feasibility of the ADPL concept at full-scale in field conditions, simulate improved digester designs to increase digestion efficiency, evaluate digester effluent post-treatment for residual organic and nutrient removal, and develop a remote data acquisition and controls system to improve system understanding and operation of the pasteurization system. The desired outcome of this work is a complete, self-sustaining system that efficiently digests fecal sludge for maximum biogas production, produces a polished effluent that can be reused, and pasteurizes the effluent efficiently and reliably, all while being low-cost with minimal operation and maintenance requirements.

Two ADPL systems were installed on residential plots with 15-35 residents in a peri-urban area outside of Eldoret, Kenya. Each system was comprised of 3 toilets built above a floating dome digester and heat pasteurization system. The ADPLs are simple systems, with no moving parts and relying on gravity-induced flows. Adoption at two sites was successful, and residents reported that the systems had little to no odor or flies, and the residents were interested in the possibility of excess biogas and effluent reuse. The ADPLs were monitored daily for biogas production and temperatures in the pasteurization system. The ADPL serving 35 residents produced on average 350 Lbiogas d-1, and the temperature in the heating tank was greater than 65 °C on 87% of sampling days. The treated effluent was analyzed periodically for chemical oxygen demand (COD), biochemical oxygen demand (BOD), total ammonia nitrogen (TAN), and pH. On average, the effluent contained 4,500-5,600 mg COD L-1 (an 87-89% reduction of the estimated input), 2,000-3,900 mg BOD L-1, 2,400-4,800 mg NH3-N, and had a pH of 7.4-7.7. Results from this field study show that anaerobic digestion of minimally diluted fecal sludge can provide enough energy to pasteurize the effluent, and that the ADPL can be a suitable option for onsite fecal sludge treatment.

Three variations of a 2 m3 anaerobic digester were simulated with a flow of 120 Lwater d-1 – a reactor with no internal baffle walls (CSTR), a reactor with baffle walls that forced flow to wind in the xy-direction (HABR), and a reactor with baffle walls that forced flow to wind in the xz-direction (ABR). Results showed that increasing the number of baffle walls significantly improved the hydraulic performance of the reactor in terms of residence time, dead space, and Morrill Index. Adding angled portions to the end of baffle walls and adjusting the D:U ratio in the ABR had minimal impact while a variable inflow had a moderate impact on performance. Overall, these results suggest that adding 3-5 baffle walls inside of an anaerobic digester would greatly improve the digester’s hydraulic efficiency and better utilize the reactor volume. These adjustments would thus cause enhanced solids removal and digestion efficiency, resulting in higher biogas production and a cleaner effluent. However, simulation work including solids and biological reactions would be beneficial to future reactor design considerations.

The biological filter study analyzed the treatment of high-strength anaerobic digester effluent using trickling filters for nitrification and then submerged attached growth filters for denitrification. Five media types were tested in the trickling filters (8 L volume): biochar, granular activated carbon (GAC), zeolite (clinoptilolite), Pall rings, and gravel. Five columns were tested for denitrifying filters (4 L volume) using sand, bamboo wood chips, eucalyptus wood chips, bamboo with sand, and eucalyptus with sand. Wood chips were used in denitrifying filters as a supplemental carbon source for denitrification. From six months of operation, biochar, GAC, zeolite, Pall rings, and gravel media had turbidity removal efficiencies of 90, 91, 77, 74, and 74%, respectively, and NH3-N removal efficiencies of 83, 87, 85, 30, and 80%, respectively. The primary mechanism for ammonia removal was nitrification to nitrate, but some adsorption was seen in biochar, GAC, and zeolite filters. From four months of operation, sand, bamboo, bamboo with sand, eucalyptus, and eucalyptus with sand filters had NO3-N removal efficiencies of 30, 59, 51, 31, and 30%, respectively, and turbidity removal efficiencies of 88, 89, 84, 89, and 88%, respectively. Bamboo had the greatest NO3-N removal rate at 0.054 kg N m-3 d-1 and released more COD than eucalyptus (0.076-0.120 gCOD gbamboo-1 compared to 0.012-0.043 gCOD geucalyptus-1). Biochar and bamboo were selected as the best media types from this study for the nitrification and denitrification filters, respectively, due to their low-cost and sustainable supply. Based on an average initial influent of 600 mg NH3-N L-1 and 980 NTU, the biochar filter’s expected effluent would be 97 mg NH3-N L-1, 450 mg NO3-N L-1, and 120 NTU. The bamboo filter would then produce an effluent of 82 mg NH3-N L-1, 180 mg NO3-N L-1, and 13 NTU. This theoretical combined performance would thus result in 56% removal of total N and 98.7% removal of turbidity. Based on nitrate removal rate, full denitrification could be achieved by doubling reactor volume. Total nitrogen removal efficiency of 80-90% could thus be achievable. These filter media were successful in treating high-strength digester effluent and present an alternative for sustainable, low-cost, and low-maintenance post-treatment options for nitrogen management.

A low-cost data acquisition and controls system with remote, real-time data access was developed using the Particle Electron. This device records temperature and liquid flow data while controlling a gas valve and igniter as part of pasteurization system. The device was tested in lab and field conditions. The power consumption is low, 34 Wh per day, and data acquisition matched the results of standard laboratory devices. The field deployment (Eldoret, Kenya) successfully operated the pasteurization system in its target range while reporting real-time data. This low-cost and low-power device has improved the operation of the onsite pasteurization system, and adaptations of the device would be valuable in many other onsite fecal sludge treatment systems.

Together, these objectives have demonstrated the ADPL concept works in field conditions, digester performance can be improved with simple modifications, digester effluent can be further treated to encourage reuse or for safe disposal with biological filters using sustainable media that have low operational requirements, and low-cost controls can improve the pasteurization system efficiency and reliability while generating more data to expand understanding of the system.

Exposure to higher PM2.5 can lead to increased risks of mortality; however, the spatial concentrations of PM2.5 are not well characterized, even in megacities, due to the sparseness of regulatory air quality monitoring (AQM) stations. This motivates novel low-cost methods to estimate ground-level PM2.5 at a fine spatial resolution so that PM2.5 exposure in epidemiological research can be better quantified and local PM2.5 hotspots at a community-level can be automatically identified. Wireless low-cost particulate matter sensor network (WLPMSN) is among these novel low-cost methods that transform air quality monitoring by providing PM information at finer spatial and temporal resolutions; however, large-scale WLPMSN calibration and maintenance remain a challenge because the manual labor involved in initial calibration by collocation and routine recalibration is intensive, the transferability of the calibration models determined from initial collocation to new deployment sites is questionable, as calibration factors typically vary with urban heterogeneity of operating conditions and aerosol optical properties, and the stability of low-cost sensors can drift or degrade over time. This work presents a simultaneous Gaussian Process regression (GPR) and simple linear regression pipeline to calibrate and monitor dense WLPMSNs on the fly by leveraging all available reference monitors across an area without resorting to pre-deployment collocation calibration. We evaluated our method for Delhi, where the PM2.5 measurements of all 22 regulatory reference and 10 low-cost nodes were available for 59 days from January 1, 2018 to March 31, 2018 (PM2.5 averaged 138 ± 31 μg m-3 among 22 reference stations), using a leave-one-out cross-validation (CV) over the 22 reference nodes. We showed that our approach can achieve an overall 30 % prediction error (RMSE: 33 μg m-3) at a 24 h scale and is robust as underscored by the small variability in the GPR model parameters and in the model-produced calibration factors for the low-cost nodes among the 22-fold CV. Of the 22 reference stations, high-quality predictions were observed for those stations whose PM2.5 means were close to the Delhi-wide mean (i.e., 138 ± 31 μg m-3) and relatively poor predictions for those nodes whose means differed substantially from the Delhi-wide mean (particularly on the lower end). We also observed washed-out local variability in PM2.5 across the 10 low-cost sites after calibration using our approach, which stands in marked contrast to the true wide variability across the reference sites. These observations revealed that our proposed technique (and more generally the geostatistical technique) requires high spatial homogeneity in the pollutant concentrations to be fully effective. We further demonstrated that our algorithm performance is insensitive to training window size as the mean prediction error rate and the standard error of the mean (SEM) for the 22 reference stations remained consistent at ~30 % and ~3–4 % when an increment of 2 days’ data were included in the model training. The markedly low requirement of our algorithm for training data enables the models to always be nearly most updated in the field, thus realizing the algorithm’s full potential for dynamically surveilling large-scale WLPMSNs by detecting malfunctioning low-cost nodes and tracking the drift with little latency. Our algorithm presented similarly stable 26–34 % mean prediction errors and ~3–7 % SEMs over the sampling period when pre-trained on the current week’s data and predicting 1 week ahead, therefore suitable for online calibration. Simulations conducted using our algorithm suggest that in addition to dynamic calibration, the algorithm can also be adapted for automated monitoring of large-scale WLPMSNs. In these simulations, the algorithm was able to differentiate malfunctioning low-cost nodes (due to either hardware failure or under heavy influence of local sources) within a network by identifying aberrant model-generated calibration factors (i.e., slopes close to zero and intercepts close to the Delhi-wide mean of true PM2.5). The algorithm was also able to track the drift of low-cost nodes accurately within 4 % error for all the simulation scenarios. The simulation results showed that ~20 reference stations are optimum for our solution in Delhi and confirmed that low-cost nodes can extend the spatial precision of a network by decreasing the extent of pure interpolation among only reference stations. Our solution has substantial implications in reducing the amount of manual labor for the calibration and surveillance of extensive WLPMSNs, improving the spatial comprehensiveness of PM evaluation, and enhancing the accuracy of WLPMSNs. Satellite-based ground-level PM2.5 modeling is another such low-cost method. Satellite-retrieved aerosol products are in particular widely used to estimate the spatial distribution of ground-level PM2.5. However, these aerosol products can be subject to large uncertainties due to many approximations and assumptions made in multiple stages of their retrieval algorithms. Therefore, estimating ground-level PM2.5 directly from satellites (e.g., satellite images) by skipping the intermediate step of aerosol retrieval can potentially yield lower errors because it avoids retrieval error propagating into PM2.5 estimation and is desirable compared to current ground-level PM2.5 retrieval methods. Additionally, the spatial resolutions of estimated PM2.5 are usually constrained by those of the aerosol products and are currently largely at a comparatively coarse 1 km or greater resolution. Such coarse spatial resolutions are unable to support scientific studies that thrive on highly spatially-resolved PM2.5. These limitations have motivated us to devise a computer vision algorithm for estimating ground-level PM2.5 at a high spatiotemporal resolution by directly processing the global-coverage, daily, near real-time updated, 3 m/pixel resolution, three-band micro-satellite imagery of spatial coverages significantly smaller than 1 × 1 km (e.g., 200 × 200 m) available from Planet Labs. In this study, we employ a deep convolutional neural network (CNN) to process the imagery by extracting image features that characterize the day-to-day dynamic changes in the built environment and more importantly the image colors related to aerosol loading, and a random forest (RF) regressor to estimate PM2.5 based on the extracted image features along with meteorological conditions. We conducted the experiment on 35 AQM stations in Beijing over a period of ~3 years from 2017 to 2019. We trained our CNN-RF model on 10,400 available daily images of the AQM stations labeled with the corresponding ground-truth PM2.5 and evaluated the model performance on 2622 holdout images. Our model estimates ground-level PM2.5 accurately at a 200 m spatial resolution with a mean absolute error (MAE) as low as 10.1 μg m-3 (equivalent to 23.7% error) and Pearson and Spearman r scores up to 0.91 and 0.90, respectively. Our trained CNN from Beijing is then applied to Shanghai, a similar urban area. By quickly retraining only RF but not CNN on the new Shanghai imagery dataset, our model estimates Shanghai 10 AQM stations’ PM2.5 accurately with a MAE and both Pearson and Spearman r scores of 7.7 μg m-3 (18.6% error) and 0.85, respectively. The finest 200 m spatial resolution of ground-level PM2.5 estimates from our model in this study is higher than the vast majority of existing state-of-the-art satellite-based PM2.5 retrieval methods. And our 200 m model’s estimation performance is also at the high end of these state-of-the-art methods. Our results highlight the potential of augmenting existing spatial predictors of PM2.5 with high-resolution satellite imagery to enhance the spatial resolution of PM2.5 estimates for a wide range of applications, including pollutant emission hotspot determination, PM2.5 exposure assessment, and fusion of satellite remote sensing and low-cost air quality sensor network information. We later, however, found out that this CNN-RF sequential model, despite effectively capturing spatial variations, yields higher average PM2.5 prediction errors than its RF part alone using only meteorological conditions, most likely the result of CNN-RF sequential model being unable to fully use the information in satellite images in the presence of meteorological conditions. To break this bottleneck in PM2.5 prediction performance, we reformulated the previous CNN-RF sequential model into a RF-CNN joint model that adopts a residual learning ideology that forces the CNN part to most effectively exploit the information in satellite images that is only “orthogonal” to meteorology. The RF-CNN joint model achieved low normalized root mean square error for PM2.5 of within ~31% and normalized mean absolute error of within ~19% on the holdout samples in both Delhi and Beijing, better than the performances of both the CNN-RF sequential model and the RF part alone using only meteorological conditions. To date, few studies have used their simulated ambient PM2.5 to detect hotspots. Furthermore, even the hotspots studied in these very limited works are all “global” hotspots that have the absolute highest PM2.5 levels in the whole study region. Little is known about “local” hotspots that have the highest PM2.5 only relative to their neighbors at fine-scale community levels, even though the disparities in outdoor PM2.5 exposures and their associated risks of mortality between populations in local hotspots and coolspots within the same communities can be rather large. These limitations motivated us to concatenate a local contrast normalization (LCN) algorithm at the end of the RF-CNN joint model to automatically reveal local PM2.5 hotspots from the estimated PM2.5 maps. The RF-CNN-LCN pipeline reasonably predicts urban PM2.5 local hotspots and coolspots by capturing both the main intra-urban spatial trends in PM2.5 and the local variations in PM2.5 with urban landscape, with local hotspots relating to compact urban spatial structures while coolspots being open areas and green spaces. Based on 20 sampled representative neighborhoods in Delhi, our pipeline revealed that on average a significant 9.2 ± 4.0 μg m-3 long-term PM2.5 exposure difference existed between the local hotspots and coolspots within the same community, with Indian Gandhi International Airport area having the steepest increase of 20.3 μg m-3 from the coolest spot (the residential area immediately outside the airport) to the hottest spot (airport runway). This work provides a possible means of automatically identifying local PM2.5 hotspots at 300 m in heavily polluted megacities. It highlights the potential existence of substantial health inequalities in long-term outdoor PM2.5 exposures within even the same local neighborhoods between local hotspots and coolspots. Apart from posing serious health risks, deposition of dust and anthropogenic particulate matter (PM) on solar photovoltaics (PVs), known as soiling, can diminish solar energy production appreciably. As of 2018, the global cumulative PV capacity crossed 500 GW, of which at least 3–4% was estimated to be lost due to soiling, equivalent to ~4–6 billion USD revenue losses. In the context of a projected ~16-fold increase of global solar capacity to 8.5 TW by 2050, soiling will play an increasingly more important part in estimating and forecasting the performance and economics of solar PV installations. However, reliable soiling information is currently lacking because the existing soiling monitoring systems are expensive. This work presents a low-cost remote sensing algorithm that estimates utility-scale solar farms’ daily solar energy loss due to PV soiling by directly processing the daily (near real-time updated), 3 m/pixel resolution, and global coverage micro-satellite surface reflectance (SR) analytic product from the commercial satellite company Planet. We demonstrate that our approach can estimate daily soiling loss for a solar farm in Pune, India over three years that on average caused ~5.4% reduction in solar energy production. We further estimated that around 437 MWh solar energy was lost in total over the 3 years, equivalent to ~11799 USD, at this solar farm. Our approach’s average soiling estimation matches perfectly with the ~5.3% soiling loss reported by a previous published model for this solar farm site. Compared to other state-of-the-art PV soiling modeling approaches, the proposed unsupervised approach has the benefit of estimating PV soiling at a precisely solar farm level (as in contrast to coarse regional modeling for only large spatial grids in which a solar farm resides) and at an unprecedently high temporal resolution (i.e., 1 day) without resorting to solar farms’ proprietary solar energy generation data or knowledge about the specific components of deposited PM or these species’ dry deposition flux and other physical properties. Our approach allows solar farm owners to keep close track of the intensity of soiling at their sites and perform panel cleaning operations more strategically rather than based on a fixed schedule.

Deep representation learning has shown its effectiveness in many tasks such as text classification and image processing. Many researches have been done to directly improve the representation quality. However, how to improve the representation quality by cooperating ancillary data source or by interacting with other representations is still not fully explored. Also, using representation learning to help other tasks is worth further exploration.

In this work, we explore these directions by solving various problems in natural language processing and image processing. In the natural language processing part, we first discuss how to introduce alternative representations to improve the original representation quality and hence boost the model performance. We then discuss a text representation matching algorithm. By introducing such matching algorithm, we can better align different text representations in text generation models and hence improve the generation qualities.

For the image processing part, we consider a real-world air condition prediction problem: ground-level $PM_{2.5}$ estimation. To solve this problem, we introduce a joint model to improve image representation learning by incorporating image encoder with ancillary data source and random forest model. We the further extend this model with ranking information for semi-supervised learning setup. The semi-supervised model can then utilize low-cost sensors for $PM_{2.5}$ estimation.

Finally, we introduce a recurrent kernel machine concept to explain the representation interaction mechanism within time-dependent neural network models and hence unified a variety of algorithms into a generalized framework.

As biological challenges to environmental health, such as the proliferation of antibiotic resistance genes, continue to emerge there is a greater need for the generation of models capable of predicting the fate and transport of biological particles. For over 100 years, Smoluchowski’s watershed 1917 work has provided a foundation for the construction of such models. Classically, this approach has been used for inert nano-scale particles. However, given that several of the most pressing challenges are biological in nature, it is imperative that predictive models of particle transport be adapted to include particles with a biological signature.This work uses modified Smoluchowskian aggregation modeling parameters to investigate the transport of three primary biological particles: bacteriophages, extracellular vesicles, and extracellular DNA, each of them often existing on the nanoscale. This was done by 1) using modified Smoluchowskian aggregation parameters to predict phage-induced host lysis, 2) characterize phage-kaolinite heteroaggregation, and 3) construct a multi-particle predictive model incorporating the heteroaggregation of all three particle types. This work found that modified Smoluchowskian aggregation parameters used in concert with appropriate population balances were largely successful in predicting such particles’ transport and provided unique insight into possible design features for engineered environmental systems.

The dipole-flow test is a novel aquifer characterization technique that utilizes a single-borehole measurement system to yield the vertical hydraulic conductivity, horizontal hydraulic conductivity, and storativity within confined aquifers. The test implements a packer and a pump system that creates a hydraulic dipole flow pattern by pumping water at a constant rate thought a suction screen, transferring it within the well to a second chamber, and injecting it back into the aquifer. Various mathematical models have been developed to derive the drawdown in each chamber and estimating water flow parameters. This thesis derives and proposes a new mathematical model that deals with packers containing asymmetrical chamber lengths. It further tests this formula by implementing in on a particular aquifer of interest and contrasting the numerical findings with those obtained in field testing and simulations as described in a Johnson and Simmon 2007 publication.

In order to derive this equation we utilize the principles of superposition, the Taylor series, the Newton Raphson model, and the implementation of an error function. We also draw elements of the Hantush leaky well function and the infinity aquifer simplifications suggested by Zlotnik. The results obtain from this computation demonstrated that this developed hydrologic model yields accurate and rational measurements for drawdown and conductivity. We conclude that our modeled formulas surpass those proposed in the Johnson article, and provides experimenters with a valuable and efficient mathematical tool for aquifer characterization.

Silver nanoparticles (AgNPs) are increasingly being integrated into a wide range of consumer products, such as air filters, washing machines, and textiles, due to their antimicrobial properties [1]. However, despite the beneficial applications of AgNPs into consumer products, it is likely that their use will facilitate the release of AgNPs into wastewater treatment plants, thereby possibly negatively impacting key microorganisms involved in nutrient removal. For this reason, it is important to characterize the effects of AgNPs in natural and engineered systems and to measure the antimicrobial effect of AgNPs on wastewater microorganisms. Polyvinyl alcohol coated AgNPs have already been linked to decreased nitrifying activity [2] and it is important to determine if AgNPs coated with other materials follow similar trends. Furthermore, it is likely that, with repeated exposure to AgNPs microbial communities could evolve and develop resistance to silver. Thus, a long-term effect of silver nanoparticle exposure could be a reduction of the efficacy of such products in a similar fashion to the development of microbial antibiotic resistance [3]. Therefore, it is critical that the impacts of these materials be ascertained in wastewater treatment systems to prevent long-term negative effects.

The objectives of this dissertation were to: 1) characterize the effect of several different AgNPs on the ammonia oxidizing bacterium (AOB) Nitrosomonas europaea and investigate possible mechanisms for toxicity, 2) test the effects of consumer product AgNPs on a wide range of heterotrophic bacteria, 3) evaluate the effects of AgNPs on bench scale wastewater sequencing batch reactors, and lastly 4) assess the impacts on microbial communities that are applied with AgNP spiked wastewater biosolids.

First, Nitrosomonas europaea was was selected because wastewater nitrifying microorganisms carry out the first step in nitrification and are known to be sensitive to a wide range of toxicants [4].The antimicrobial effects of AgNPs on the AOB N. europaea were measured by comparing nitrite production rates in a dose response assay and analyzing cell viability using the LIVE/DEAD® fluorescent staining assay. AgNP toxicity to N. europaea appeared to be largely nanoparticle coating dependent. While PVP coated AgNPs have shown reductions up to 15% in nitrite production at 20 ppm, other AgNPs such as gum arabic (GA) coated showed the same level of inhibition at concentrations of 2 ppm. The first mechanism of inhibition appears to be a post-transcriptional interference of AMO/HAO by either dissolved Ag or ROS, in treatments where membranes are not completely disrupted but nitrite production decreased (2 ppm GA AgNP and 2 ppm PVP AgNP treatments). The disruption of nitrification is dependent on AgNP characteristics, such as zeta potential and coating, which will dictate how fast the AgNP will release Ag+ and ROS production Finally, total membrane loss and release of internal cellular matter occur.

In order to test the effects of AgNP products available to consumers, simple bacterial toxicity tests were carried out on well-studied heterotrophic bacteria. A model gram-positive and gram-negative bacterium (B. subtilis and E. coli, respectively) was selected to assess any differences in sensitivity that may occur with the exposure to AgNPs. A third model gram-negative bacterium (P. aeruginosa) was chosen for its biofilm forming capabilities. In addition to testing pure nanoparticles, three silver supplements meant for ingestion, were randomly chosen to test with these three bacteria. Growth curve assays and LIVE/DEAD staining indicate that the consumer product AgNPs had the most significant inhibition on growth rates, but not membrane integrity. Overall, P. aeruginosa was most negatively affected by all AgNPs with nearly 100% growth inhibition for all 2 ppm AgNP treatments. TEM imaging also confirmed cell wall separation in P. aeruginosa and internal density differences for E. coli. The effects on B. subtilis, a gram-positive bacterium, were not as severe but toxicity was observed for several AgNPs at concentrations greater than 2 ppm. Citrate AgNPs appeared to have the most impact on membrane integrity, while other mechanisms such as internal thiol binding might have been at work for other AgNPs.

The effects of varying concentrations of pure AgNPs on complex microbial wastewater reactors are currently being tested. Eight bench-scale sequencing batch reactors were set up to follow the typical "fill, react, settle, decant, idle" method with an 8 hour hydraulic retention time and constant aeration. Reactors were fed synthetic wastewater and treatment efficiency is measured by monitoring effluent concentrations of COD, NH4+, and NO3-. The reactors were seeded with 500 mL of activated sludge from a local wastewater treatment plant. After reaching steady state, the reactors were spiked with 0.2 ppm gum arabic and citrate coated AgNPs. Treatment efficiency was monitored and results showed significant spikes and ammonia and COD immediately following the first spike, but the microbial community appeared to adapt for future AgNP spikes. Microbial community analysis (terminal restriction fragment length polymorphism) showed confirmed this hypothesis.

Overall, this dissertation asserts that by examining AgNP coating type, Ag+ dissolution rates and Stern layer surface charge, it may be possible to predict which AgNPs may be more detrimental wastewater treatment, but not all AgNPs will have the same effect. The results obtained herein must be expanded to other types of AgNPs and microorganisms of ecological importance.

The antimicrobial properties of silver have been taken advantage of by societies for thousands of years. Its use has come back in favor in the form of silver nanoparticles, which are highly efficacious antimicrobial agents. Silver nanoparticles are incorporated into a myriad of products specifically designed for clinical use, but also for general use by consumers. Silver nanoparticles can be found in textiles such as clothing and stuffed toys, and in home appliances including washing machines and curling irons. A large number of products specifically marketed for use by children are also available to consumers, including pacifiers, sippy cups, and even breast milk storage bags. The hazards and toxicities associated with silver nanoparticles are not well understood, however modes of toxicity have been reported for ionic silver. It is assumed that the main mechanism of toxicity of silver nanoparticles relates to the release of ionic silver, however studies have indicated an additional nano-effect, likely due to nanoparticle size, differential coatings, and means of sustained dosing of ionic silver. However we are sure that these silver nanoparticles will accumulate in the waste stream, likely arriving during different stages of a product’s lifespan. A main sink of these nanoparticles travelling through both natural and engineered environments is wastewater treatment plants. As a society we rely on the biological removal of nutrients, which takes place primarily in the activated sludge of secondary treatment. Studies have already indicated possible, temporary decreases in removal efficiencies as well as changes in microbial communities, including losses of diversity, following exposure to silver nanoparticles. Therefore, it is of paramount importance to examine the effects of both silver nanoparticles and ionic silver on the community and function of wastewater bacteria.

Sequencing batch reactors were operated to mimic wastewater treatment. They were fed synthetic wastewater and after reaching acclimation, were dosed over time with varying concentrations of both ionic and nanosilver. Cell samples were collected periodically to assess the presence and identity of cultivable silver resistant bacteria and to map the microbial community changes taking place under different treatments using Next Generation Sequencing. Isolates were tested for the presence of known silver resistance (sil) genes as were activated sludge samples from a collection of domestic wastewater treatment plants, by designing TaqMan probe assays and performing quantitative PCR. The silver resistant isolates were also used to test the growth implications, as well as sil gene expression changes, following treatment with ionic silver and a variety of silver nanoparticles with various coatings, all at multiple concentrations. This was accomplished by performing multiple batch experiments and then using the TaqMan assays and reverse transcription-quantitative PCR.

Overall, microbial community changes were observed in the sequencing batch reactors, and there were differences noted based on treatment, including ionic silver versus nanosilver and between the two silver nanoparticle coatings. Most notably, the possibility of nitrification in wastewater treatment being particularly susceptible was strongly indicated. Individual wastewater bacteria isolates all contained sil genes, as did the majority of the wastewater treatment plant activated sludge, however the levels of actual sil gene expression were inconsistent. This particular finding supports a current body of work indicating that there are alternate modes of bacterial silver resistance in play that we are just becoming aware of.

Coal fired power plants generate approximately 45% of the electricity produced in the United States every year, and each year, over 100 million tons of coal ash are produced as a by-product of electricity generation. Coal ash is a solid waste made up principally of bottom ash, fly ash, and flue gas desulfurization materials. The chemical composition of coal ash varies depending on the feed coal source, combustion parameters, and the presence and type of air pollution control devices that remove contaminants from the flue gas into the solid waste stream. Although a significant portion of coal ash waste is recycled, the majority of coal ash is disposed in landfills and holding ponds. Coal ash impoundments have a long history of environmental degradation, which includes: contaminant leaching into groundwater, the discharge of contaminant-laden effluent into surface waters, and catastrophic impoundment failures and ash spills. Despite these known problems, coal ash is not considered a hazardous waste, and thus is not subject to stringent disposal requirements. The current coal ash management system is based on risk assessments of coal ash that do not include environmental parameters that have a profound impact on coal ash contaminant mobility, particularly for the toxic elements such as mercury, arsenic, and selenium. This dissertation research focused on the biogeochemical transformations of mercury, arsenic, and selenium associated with coal ash materials in an effort to: (1) define the key environmental parameters controlling mercury, arsenic, and selenium fate during disposal and ash spills; and (2) delineate the relationship between coal ash characteristics, environmental parameters, and leaching potential.

The impact of coal ash on mercury transformations in anaerobic systems was assessed using anaerobic sediment-ash microcosms to mimic an ash spill into a benthic aquatic system. Anaerobic sediments are the primary zones for the microbial conversion of inorganic mercury to methyl mercury (MeHg), a process that is mediated by anaerobic bacteria, particularly sulfate reducing bacteria (SRB). MeHg is a potent neurotoxin that biomagnifies up the aquatic food chain, presenting a human health risk-- especially to children and pregnant women. The results of the sediment-ash microcosm experiments indicated negligible net production of MeHg in microcosms with no ash and in microcosms amended with the low-sulfate/low-Hg ash. In contrast, microcosms amended with sulfate and mercury-rich ash showed increases in MeHg concentrations that were two to three times greater than control microcosms without ash. The enhancement MeHg production in the microcosms was likely due to large quantities of leachable sulfate that stimulated the activity of methylating bacteria. Overall, these results highlight the importance of considering both the geochemical conditions of the receiving environment and the chemical composition of the coal ash in assessing the MeHg potential of coal ash.

The hypothesis that sulfate-rich coal ash can change sediment microbial communities, enhancing MeHg production, was tested by analyzing coal ash impacts on the SRB community in the sediment-ash microcosms using Terminal Restriction Fragment Length Polymorphism (T-RFLP), Quantitative Polymerase Chain Reaction (q-PCR), and Reverse Transcription-qPCR (RT-qPCR). Coal ash did not appear to cause significant changes to the structure of the overall bacterial community, though results showed that it may have caused a decrease in the evenness for species distribution for both SRB and the overall microbial community. During the five-day incubation experiment, the coal ash had a temporary significant effect on SRB abundance during the first one to two days of the experiment and a more sustained effect on SRB activity. This stimulation of SRB population growth and activity also corresponded with increasing net MeHg production. Overall, results indicate that coal ash amendments do not cause large shifts in the overall microbial community or the SRB community, but results indicate that there are connections between SRB abundance/activity and MeHg production. More research is needed to determine how coal ash directly impacts Hg methylating microorganisms, which include diverse array of microorganisms outside of SRB.

The effect of aerobic and anaerobic conditions on arsenic and selenium leaching from coal ash in an ash spill scenario was also assessed using sediment-ash microcosms. The fate of arsenic and selenium associated with coal ash is of particular concern due to the leachability of these elements at neutral pH and their tendency to bioaccumulate in aquatic organisms. Both the redox speciation of arsenic and selenium, and the pH of the aquatic system, are known to influence leaching into the environment, yet current environmental risk assessments of coal ash focus on pH alone as the primary driving force for arsenic and selenium leaching from coal ash and do not take into account the effects of anaerobic conditions and microbial activity. In this research, total dissolved concentrations of arsenic and selenium, dissolved speciation of arsenic, and solid phase speciation of selenium were monitored to determine the biogeochemical transformations and leaching of arsenic and selenium under differing redox conditions. The results from the sediment-ash microcosm studies showed that redox potential was the major determinant of arsenic and selenium mobility in the microcosm systems with greater arsenic leaching occurring in anaerobic microcosms and greater selenium leaching occurring in aerobic microcosms. Furthermore, the experiments provided clues to how coal ash influences the geochemistry of the benthic environment and how these influences affect the speciation and longer term solubility of arsenic and selenium.

Finally, experiments were conducted to determine how differing CaO, SO3, and Fe2O3 concentrations in coal ash affect the release of arsenic and selenium from sediment-ash mixtures in a simulated ash spill environment. Aerobic and anaerobic sediment-ash microcosms were constructed to mimic an ash spill into a benthic aquatic system, and a variety of coal ash materials were tested as amendments, including seven fly ashes, one lime-treated fly ash sample, and two FGD samples. Results showed that, in most cases, the sediment in the microcosm buffered the system at neutral, which counteracted leaching impacts of differing CaO and SO3 concentrations in the microcosms. Regardless of ash material, leaching of selenium was greater under aerobic conditions and was correlated with the total selenium content of the microcosm. Maximum leaching of arsenic occurred in anaerobic microcosms for some ash materials and in aerobic microcosms for other materials, suggesting that ash material chemistry played a significant role in controlling arsenic mobility. In both aerobic and anaerobic microcosms, dissolved arsenic concentration was correlated with total arsenic content of the ash material and in anaerobic microcosms, dissolved arsenic concentrations also correlated with the total iron content of the ash material. Overall, the results of these experiments showed that arsenic and selenium release under environmentally relevant conditions cannot be predicted by the CaO and SO3 content of the ash material. Rather, the total arsenic, total selenium content, and total iron content of the ash material are good predictors of the worst case environmental leaching scenario.

These investigations illuminated two major conclusions: (1) microbial activity and differing redox conditions are key in determining the impact of coal ash on the environment and in determining the mobility of coal ash contaminants, and (2) coal ash characteristics, such as sulfate and iron content, can change the redox chemistry and microbial activity of the surrounding environment, further influencing the fate of ash contaminants. This work will be useful in designing a framework that accurately predicts the leaching potential of ash contaminants under environmentally relevant conditions. The results will also be helpful in developing treatment technologies for ash impoundment effluent, guiding decisions on ash pond closure and remediation, and in designing long-term monitoring plans and remediation strategies for ash-impacted sites.

This dissertation presents a framework to improve bioremediation of soils polluted with polycyclic aromatic hydrocarbons (PAHs). PAHs are of great concern because they are recalcitrant and toxic. PAHs enter the environment from a variety of sources such as incomplete combustion and coal tar distillation. The PAHs focused on in this dissertation have polluted soils as a result of creosote-based wood treatment operations that took place at Holcomb Creosote and Atlantic Wood Industries, Inc. (AWI) both of which are now classified as Superfund sites. There are numerous sites analogous to these two Superfund sites throughout the world which have been polluted through similar wood-treatment operations, as creosote was once industry’s foremost wood preservative.

There is room for existing PAH treatment options, which are mainly physical and chemical in nature, to be expanded to include more sustainable options. Commonly used technologies include excavation, in situ stabilization, and soil washing. Historically, bioremediation strategies relying on bacteria to transform pollutants have been challenged by the tight sorption of heavy- and middle-weight PAHs to soils, as this restricts aqueous phase transport required for bacterial degradation. Multiple studies have demonstrated fungi to be capable of degrading these inaccessible pollutants and other mixtures of hydrophobic pollutants (mycoremediation). Yet, when fungi have been introduced to polluted soils (mycoaugmentation), they have not been able to outcompete the native microbiota long enough to degrade the contaminants of concern over the long term. It is possible that a thorough characterization of the indigenous fungi at a given site may provide some insights into the development of targeted in situ mycoremediation strategies.

Although incorporating site microbes has been generally acknowledged as important for some time, the techniques enabling thorough assessment of microbial ecosystems are relatively new. Consequently, little is known about PAH-associated microbiomes in general, and even less is known about PAH-associated fungal communities. The work presented in this dissertation aims to address this knowledge gap by leveraging recent advances in high-throughput sequencing technology to design and validate targeted mycoremediation strategies. To this end, the overarching goal of this dissertation was to develop and test a framework for incorporating native fungi into a bioremediation strategy to expand such sustainable remediation options to sites where they have not been relevant in the past.

In the first aim of this dissertation research, advances in high-throughput sequencing were used to identify potential biostimulation targets in soils moderately polluted with PAHs. The next generation sequencing (NGS) platform, Illumina, was utilized to sequence the large sub-unit (LSU) gene commonly used as a marker gene in fungal community studies. Relationships were examined between concentrations of over 31 different polycyclic aromatic hydrocarbons and the pollutant-associated communities to test whether there were any fungi capable of tolerating high levels of these toxic contaminants. In this aim, fungal genera were identified that contained species closely related to known PAHs and petroleum hydrocarbon degraders. In all, this work identified 32 targets for biostimulation, based on Spearman rank correlations between prevalence and mid- and high-molecular weight PAHs. Ascomycetes were found to have higher levels of diversity than any other phylum in this subset of biostimulation targets. These data suggest that ascomycete fungi are more likely to be present in heavily polluted soils than basidiomycete fungi (which had previously been subjects of much interest). Overall, this work illustrates that polluted soils harbor fungal biostimulation targets, specifically within Ascomycota.

The second aim of this thesis research was to use the precision bioremediation assessment in highly polluted soils and then to evaluate a range of amendments with the goal of identifying strategies to stimulate the fungal communities that dominate these PAH-associated fungal communities. Here we applied the approach we fine-tuned in the first aim to the AWI soils, as these soils have some of the highest documented PAH-concentrations. Again, Ascomycota were found to be more prevalent in these soils, so an isolate obtained from AWI was used to compare alternative stimulation techniques between three substrates they are known to grow on: chitin, cellulose, and wood. We used anthracene degradation as a proxy for PAH degradation, which we monitored in sacrificial simplified bioreactors responding to the three amendments. T. harzianum is also known to have enzymes which degrade PAHs, but it is unknown which ecological role uses those enzymes, and thus which ecological role we should promote. T. harzianum was grown in the presence of chitin, cellulose, and wood as substrates in liquid culture with anthracene. Chitin was found to stimulate the highest anthracene removal, with a 0.1% (w/v) amendment resulting in ~93% degradation. While ~13% less than chitin, 1% (w/v) cellulose was also found to stimulate ~46% more anthracene degradation than wood, which had no improvement over the abiotic losses (~33% on average). This is notable because the “go to” method for stimulating fungi in the past has been wood supplements. This work provided insight into alternative stimulation strategies to target specific ecological roles that may better degrade PAHs in situ.

For the third and final aim of this dissertation research, the two most promising amendments were added with and without Trichoderma harzianum spores to test several mycoremediation treatment strategies in soil bioreactors and compare them with a (no carbon added) nutrient stimulation treatment. Pollutants were added as aged Atlantic Wood Industries soil delivering aged pollutants. Triplicate reactors from each treatment were sequenced at time zero, after two weeks, and after one month. At each sampling time, RNA was extracted, converted to cDNA, and submitted to Illumina MiSeq library preparation targeting the LSU region for fungal community analysis in addition to the V4 region of the 16S rDNA for bacterial community analysis. Statistical analyses using DESeq2 identified responders among the groups of reactors subjected to the different biostimulation treatments. Taxa from both the fungal and the bacterial communities responded differentially to the amendments. Fungi were found to comprise the majority of the significant responders. This work also found that mycoaugmented strains were not successful in establishing themselves as prominent members of the active community. This represents one of the earliest studies to directly measure mycoaugmentation failure. These data propose a hypothesis about functional redundancy inhibiting establishment of augmented fungi as already established fungi outcompete them for freshly added nutrients. Over 90% degradation was observed over the course of one month regardless of treatment-interestingly, the highest degradation was found in the nutrient amendment (no carbon added) treatment. These results show similar degradation across the soil bioreactors, yet different microbial growth, which supports the hypothesis that there is community-level functional redundancy and multiple metabolic food webs that result in the observed pollutant degradation.

Overall, this dissertation work demonstrates how significant advances in sequencing technology can be implemented in design and monitoring stages of bioremediation. This work also suggests that significant advances could be possible through the application of targeted metatranscriptomic analysis. Through incorporating such insights as described in this dissertation, this research brings the field of bioremediation one step closer to successfully engineering microbiomes to degrade contaminants of concern.

Hydrogen sulfide (H2S) is a toxic gas and common odor nuisance produced in a variety of chemical and environmental processes. The biological oxidation of H2S to sulfate/sulfuric acid is a well-documented treatment method that is efficient both in removal and cost. Sulfate ions produced in a BTF can interact with various cations, specifically calcium, and form insoluble salts. Gypsum (CaSO4*2H2O) formed within a BTF treating H2S can affect system performance by causing pressure buildup and reducing pollutant mass transfer. An experimental approach was developed to quantify gypsum precipitation in BTFs as a function of critical system parameters. Effluent liquid from one laboratory and four industrial BTFs was used to induce gypsum precipitation at various levels of pH, total sulfate concentration, calcium content, and ionic strength. A computer model was developed to predict gypsum precipitation based on the ionic composition of the reactor trickling liquid. The results support the hypothesis that gypsum precipitation in a BTF treating H2S is a realistic concern for industrial systems. The computer model demonstrates the ability to successfully predict gypsum precipitation within a correction factor of 2. The presence of gypsum and elemental sulfur in solid samples collected from industrial BTFs illustrates the feasibility of mineral deposition in full-scale treatment systems. Ethylene diamine tetraacetic acid (EDTA) shows the potential of being an effective additive for the prevention of gypsum formation within a BTF treating hydrogen sulfide.

Surface chemistry plays an essential role in determining the reactivity, bioaccessibility, bioavailability and toxicity of nanoparticles (NPs) in the environment. Processes such as aggregation, deposition and biouptake are controlled in part by the attachment efficiency, α, between particles and the surfaces they encounter. One premise of this research is that surface hydrophobicity is a pivotal property of NP surfaces that can affect the behavior of NPs in aquatic environment and potentially decide the fate and transport of NPs. However, there are multiple challenges in the characterization of hydrophobicity for NPs. Methods developed for macroscopic surfaces or organic compounds may not be readily applied or interpreted for the case of nano-scale surfaces. This dissertation addresses theoretical basis for applying methods to determining hydrophobicity of NPs. The use of an octanol-water partitioning method analogous to that used for organic compounds was evaluated on the basis of trends anticipated by thermodynamics, and by experimental observations. This work shows that partitioning of NPs in two phases systems, such as water and octanol, is not uniquely determined by hydrophobicity, but also influenced by surface charge and particle size. The water-oil interface rather than the bulk phases becomes the thermodynamically favored location for NP accumulation once NPs are larger than 1-10 nm and/or the surface is amphiphilic.

Nonetheless, the relative hydrophobicity of selected NPs, as characterized by adsorption of molecular probes (i.e. organic dye and naphthalene), was consistent with the macroscopic contact angle measurements and octanol-water distribution coefficients. The in-situ adsorption of these molecular probes offers the most solid grounds for measurement of hydrophobicity. Other measure of hydrophobicity or hydrophilicity such as water-affinity based methods that measure water vapor adsorption to nanomaterial powders, or immersion microcalorimetry and thermogravimetric analysis, yielded similar results to the molecular probes. However, possible physical or chemical transformations to NP surfaces during characterization by these other methods limited the use of results to infer hydrophobicity based on a rigorous thermodynamic model.

Column experiments suggested that the attachment efficiency of NPs to biofilm was generally greater for more hydrophobic NPs, though polymeric coatings might stabilize NPs against the attachment. The affinity of NPs for a variety of bacterial surfaces (i.e. different species, planktonic or biofilm, with or without extracellular polymeric substances (EPS)) of different hydrophobicities, which correlated with the quantity of proteins in EPS, was also investigated. It was found that the attachment of hydrophobic NPs increased with the hydrophobicity of bacterial surfaces, but not for hydrophilic NPs. Environmental conditions such as divalent ions and pH influenced the affinity of nanoparticle for bacterial surface by changing the bacterial surface hydrophobicity and electric double layer interaction, respectively.