Skip to main content
Open Access Publications from the University of California

UCLA Electronic Theses and Dissertations

Cover page of Climatic Controls on Streamflow and Snowpack over the Colorado River Basin

Climatic Controls on Streamflow and Snowpack over the Colorado River Basin


The Colorado River is the main source of surface water for the Southwestern U.S. and Mexico. It is heavily regulated by two large reservoirs, and many smaller ones. Although summer precipitation is about the same amount as winter precipitation averaged over the basin, runoff is mainly generated from the melting of snowpack accumulated during the cold season. Therefore, fresh water availability is challenged by warming temperatures that have occurred over the last few decades. A noticeable downward trend in the basin’s naturalized streamflow appears to be related to the increasing temperatures, as precipitation changes have been small. In this dissertation, I evaluate the effects of climatic controlling factors on two major hydrologic components (runoff and snow water equivalent) of the region’s water cycle. The primary tool I use is offline land surface simulations from macro-scale hydrological models. In particular, this dissertation comprises three studies that have or will be published as journal articles. First, I employed the Variable Infiltration Capacity (VIC) model to investigate the causes of the century-long decreasing trend in Colorado River streamflow. I separated the influences of warming temperatures and unevenly (spatially) distributed (and mostly small) precipitation changes by conducting a parallel set of experiments that isolated the various effects. My experiments suggest that more than half of the downward trend in streamflow is attributable to warming temperatures. Compared with an earlier drought period (1953-1968) caused mainly by insufficient precipitation over the UCRB, about half of runoff losses during the Millennium Drought (2000-2014) is attributed to warm temperature. Second, I explored the factors that control snow ablation processes across the West and differences in their representation in a multi-model suite. I selected ten USDA Snow Telemetry (SNOTEL) stations distributed across the mountainous ranges of the Western U.S. Consistent with earlier studies, I find that during the ablation period net radiation generally has stronger effects on melt rates than does air temperature. However, estimates of melt rates vary greatly across the models, in part because of differences in the way they represent the effects of vegetation on the surface energy balance. The canopy effect of each model on snow melting is also evaluated with parallel experiments. Finally, I reconstructed snowpack in the Upper Colorado River Basin (UCRB) over the last 67 years (1949-2015) using the macroscale VIC model implemented at 1/16� latitude-longitude spatial resolution. I then investigated the storms that were associated with accumulation of snow water equivalent (SWE) using 86 SNOTEL stations distributed across the UCRB. In particular, I classified storms associated with SWE accumulation into Atmospheric River (AR) related and non-AR events. The storms I identified (both AR and non-AR) during the study period account for an average of 78.2% of annual peak SWE. On average, 69% of the storms are AR-related; they contribute 56.3% of the annual snowpack maxima. I find no statistically significant basin-wide trends in the number of storms (of either type) or their contributions to SWE. However, in the middle of the basin, there are a number of grid cells with significant upward trends in the storm contributions to snow, which suggests some movement of snow accumulation towards the UCRB mid-zone over the last few decades.

  • 3 supplemental PDFs
Cover page of Query-Efficient Black-box Adversarial Attacks

Query-Efficient Black-box Adversarial Attacks


Machine learning systems have been shown to be vulnerable to adversarial examples. We study the most practical problem setup for evaluating adversarial robustness of a machine learning system with limited access: the hard-label black-box attack setting for generating adversarial examples, where limited model queries are allowed and only the decision is provided to a queried data input. Several algorithms have been proposed for this problem but they typically require huge amount (>20,000) of queries for attacking one example. Among them, one of the state-of-the-art approaches (Cheng et al., 2019) showed that hard-label attack can be modeled as an optimization problem where the objective function can be evaluated by binary search with additional model queries, thereby a zeroth order optimization algorithm can be applied. In this thesis, we adopt the same optimization formulation but propose to directly estimate the sign of gradient at any direction instead of the gradient itself, which enjoys the benefit of single query. Using this single query oracle for retrieving sign of directional derivative, we develop a novel query-efficient Sign-OPT approach for hard-label black-box attack. We provide a convergence analysis of the new algorithm and conduct experiments on several models on MNIST, CIFAR-10 and ImageNet. We find that Sign-OPT attack consistently requires 5X to 10X fewer queries when compared to the current state-of-the-art approaches and usually converges to an adversarial example with smaller perturbation.

Cover page of Information Economics and Financial Economics

Information Economics and Financial Economics


My thesis consists of three chapters on information economics and financial economics.

Chapter 1:

Many decentralized over-the-counter (OTC) markets have recently become subject to new regulations requiring transparency. I build up an information model that features bilateral trade in a double auction, endogenous public signal, and inter-dealer network formation to study the effect of TRACE on the inter-dealer markets. In the trading stage, I study the private information diffusion process and endogenize the public information contained in the disseminated trading price. I show that in markets with a relatively low degree of information asymmetry, post-trade transparency makes the adverse selection more severe and reduces the surplus from asset reallocation between dealers, and thus hurts the inter-dealer network formation. Investors are more likely to be symmetrically uninformed about thinly traded bonds. The empirical results provide evidence that TRACE has a significant negative effect on the inter-dealer trading frequency for thinly traded bonds.

Chapter 2 (with Kim-Sau Chung):

During currency crises, large traders once simultaneously short the asset markets and currency market. We study the large trader's information manipulation in crises by introducing a large trader in an asset market and a currency-attack coordination game with imperfect information. The asset price realized in the asset market aggregates dispersed private information acting as a public signal in the currency attack game. We show that the incentive of the large trader to manipulate the asset price in favor of its currency attack leads to financial contagion. In equilibrium, the large trader's manipulating the asset price to be lower and attacking the currency regime are concurrent; the large trader's manipulation in the asset market is most significant when the public signal is in the intermediate range. To draw policy implication regarding the market transparency, we show that when the asset market is transparent, a natural equilibrium refinement that incorporates forward induction reasoning would select the equilibrium where every trader behaves most aggressively in the currency-attack game and the currency regime is most fragile.

Chapter 3 (with Yinqiu Lu):

The way central banks manage their foreign reserve assets has evolved over the past decades. One major trend is managing reserves in two or more tranches -- liquidity tranche and investment tranche -- especially for those with adequate reserves. Incorporating reserve tranching, we have developed in this paper a central bank’s reserve portfolio choice model to analyze the determinants of the currency composition of reserves. In particular, we adopt the classical mean-variance framework for the investment tranche and the asset-liability framework for the liquidity tranche. Building on these frameworks, the roles of currency compositions in imports invoicing and short-term external debt, and risk and returns of reserve currencies can be quantified by our structural model -- a key contribution of our paper given the absence of structural models in the literature. Finally, we estimate the potential paths of the share of RMB in reserves under different scenarios to shed light on its status as an international currency.

Field-Programmable Acoustic Platform for Deep, Subwavelength-Resolution Patterning of Micro-Objects into Complex and Non-Periodic Shapes


Manipulation of biological micro-objects in vitro is essential for many biomedical applications, such as the study of cells’ or molecular interaction, single-cell analysis, drug delivery, and tissue engineering. There are various physical mechanisms employed to achieve the manipulation; electrokinetic, optical, magnetic, hydrodynamic, and acoustic mechanisms are all conventional examples that have been deeply researched. In comparison, acoustics has been demonstrated to have superior biocompatibility and wide range of operable sizes of target, making it an attractive option to be widely utilized in many applications. On the other hand, acoustics suffers from limited Degree-of-Freedom (DOF) in objects’ manipulation due to the fundamental constraint stemmed from the principle that conventional acoustic devices built upon. Furthermore, acoustic patterning of micro-objects that allows not only real-time, versatile adjustment but also re-configuration to the patterning profile has not yet been realized. This re-configurable feature is particularly crucial in taking the acoustic manipulation technology from the current research phase to the next stage for broader applications.

For this dissertation, two important concepts are investigated and realized to overcome the acoustic limitations described. The first concept explores a new field of acoustofluidics where a novel manipulation platform is developed, making a breakthrough to the conventional approach in objects’ manipulation in achieving highly complex and non-periodic patterning shapes. The second concept involves integration of a photothermal mechanism with the developed acoustic platform which allows the acoustic potential field to be adjusted and re-configured, leading to a programmable platform for highly complex patterning of micro-objects.

In the first concept, a new acoustofluidic field using deep, sub-wavelength approach is exploited. Different from the traditional techniques that rely on using standing waves to generate the acoustic potential wells only spaced periodically at half the wavelength, the new approach can generate the acoustic wells spaced arbitrarily within half the wavelength. To achieve such approach, we developed a “Complaint Membrane Acoustic Patterning” (CMAP) platform utilizing an air-embedded, viscoelastic Polydimethylsiloxane (PDMS) structure to precisely control the wave fronts by providing barriers to incoming waves using acoustic impedance mismatch between air and PDMS, creating a non-uniform energy field and thus the acoustic wells. The embedded air cavities dictate the shape of the wells. Since the cavities can be fabricated into any geometry, complicated profiles of the wells can be realized. As experiments have demonstrated, we succeeded in the patterning of micro-polystyrene beads and HeLa cells into various numeric-letter shapes with resolution one tenth of the wavelength across a large 3 x 3 mm2 area.

In the second concept, a photothermal mechanism is integrated into the CMAP platform to achieve re-configurable patterning of micro-objects. A layer of light absorbing hydrogenated amorphous silicon (a-Si:H) is incorporated into the bottom of CMAP’s PDMS structure, where the cavities are initially fluid-filled. By focusing a laser light onto the a-Si:H layer to evaporate the fluid within the cavities selectively, we can modify where in the PDMS structure would contain fluid or air. Cavities containing air can also be reverted to contain fluid. Since the acoustic impedance of PDMS matches closely to that of water, acoustic waves could pass through easily. This way, we have the versatility in shaping and re-shaping an array of air cavities, permitting the generation of adjustable potential profile of the acoustic wells which leads to field-programmable functionality for highly complex and non-periodic patterning.

  • 1 supplemental ZIP
Cover page of Cultural and Psychobiological Processes in Pregnant Latina Women

Cultural and Psychobiological Processes in Pregnant Latina Women


Research indicates that women who experience higher levels of anxiety related to a current pregnancy have a higher risk of preterm birth (Blackmore, Gustafsson, Gilchrist, Wyman, & O’Connor, 2016; Dunkel Schetter, 2010). Previous studies demonstrate that pregnancy anxiety is associated with higher levels of placental corticotropin-releasing hormone (pCRH) which in turn, triggers the timing of delivery. Some evidence suggests that Latinas experience heightened pregnancy anxiety (Ramos et al., 2019), but no study has documented their specific experiences of pregnancy anxiety.

This dissertation tested a model that incorporates components of biopsychosocial and cultural processes in pregnancy across two studies. The first study examined ethnic and cultural moderation linking pregnancy anxiety to the length of gestation in a sample of pregnant Latina and non-Latina White women (N = 125). These women completed interviews and provided blood samples on three separate occasions during their pregnancy. Pregnancy anxiety in the first, second, and third trimesters each predicted shorter length of gestation. Latina and non-Latina White women did not differ in pregnancy anxiety and had similar gestational length at birth. Moderation analyses on the full sample revealed that pregnancy anxiety was associated with gestational length for Latinas only. A similar pattern emerged in analyses on the subsample of Latinas for moderation by acculturation such that pregnancy anxiety predicted the timing of delivery only among low acculturated Latinas. Neither levels nor slopes of pCRH mediated the associations between pregnancy anxiety and length of gestation. This work adds to existing evidence that pregnancy anxiety predicts timing of birth and affirms that ethnicity and acculturation are relevant for understanding cultural and biopsychosocial processes leading to birth outcomes.

The second study used a qualitative methodology to investigate pregnancy anxiety and cultural beliefs surrounding pregnancy among Latinas. One focus group and 11 individual interviews were conducted in Spanish with pregnant Latinas. Thematic analysis (Braun & Clarke, 2006) revealed that Latinas felt that anxiety during pregnancy was normal, and that they were concerned about childbirth, losing their baby, their baby being born with a birth defect, and the current issues in the United States. Latinas felt lucky to be pregnant, believed that pregnancy was a blessing, and stressed the importance of maintaining a healthy pregnancy. Themes about family involvement and culturally-driven privileged status also emerged from the data.

Taken together, these two studies further affirm that pregnancy anxiety contributes to risk for preterm birth and adds evidence that associations of pregnancy anxiety and gestational length may be driven in part by ethnicity and by cultural factors among Latinas. Future studies using larger samples should further investigate these issues utilizing prospective designs with a closer examination of cultural processes. Such work can further identify the role of culture in prenatal processes affecting maternal and child outcomes and help to understand the nature and risks of pregnancy anxiety in Latinas.

Decoding the Relationship between Composition and Various Properties of Calcium Silicate and Calcium Aluminosilicate Glasses


This dissertation contributes to decode the relationship between glass composition and various properties that are crucial to novel glass design by performing molecular dynamics simulations. The topics include the prediction of glass transition temperature, the origin of glass-forming ability, and the impact of cooling rate on glass relaxation. In general, the first two properties are analyzed by utilizing topological constraint theory. By combining molecular dynamics simulations and topological constraint theory, a fully analytical model is developed to predict the fictive glass transition temperature of (CaO)x(SiO2)1-x glass system. To be specific, this model takes composition as input and provides the prediction of glass transition temperature as output. On the other hand, glass-forming ability is an important factor that guides the manufacturing process while sometimes imposes limitations to glass engineering. Despite many empirical successes to identify and characterize glass-forming ability in various glass systems, there exists a lack of knowledge of physical details in the glass structure. Here, we conduct molecular dynamics simulations of a series of calcium silicate glasses. We show that the flexible-to-rigid topological transition coincides with the compositional window that has optimal glass-forming ability. By explaining this transition from the aspect of internal flexibility and internal stress within the network, we aim to provide an alternative topological explanation for the nature of glass-forming ability. Chapter 3 is a reprint of a previously published journal article. It demonstrates the impact of cooling rate on glass transition by analyzing glass relaxation and hysteresis of a series of silicate glasses. It proves that by extrapolating simulation data, one can access the results that are close to those generated from experimental cooling rates.

Prevalence of Preexisting or Development of Temporomandibular Disorders In Obstructive Sleep Apnea Patients Prior to and During Therapy With Intra-oral Mandibular Advancement Appliance


Introduction: Obstructive sleep apnea (OSA) is a common sleep-related breathing disorder that affects more than 12 % of the US population. It is considered to be a serious problem associated with multiple physiological processes that affect normal body function including stroke, cardiovascular disease, hypertension, and type 2 diabetes. Risk factors for OSA include BMI higher than 35 kg/m2, male gender, neck circumference, and high blood pressure. The STOP-BANG questionnaire is a reliable screening tool for OSA; nevertheless, the gold standard for diagnosis of OSA is the overnight polysomnography (PSG) test. If the physician diagnoses a patient with OSA, different treatment modalities can be provided based on individual patient needs and preferences. OSA treatment modalities include positive airway pressure (PAP) therapy, surgery, and mandibular advancement devices (MADs). Even though PAP therapy is the gold standard treatment method for obstructive sleep apnea in adults, a large number of patients cannot tolerate it; therefore, for many of those patients, MADs are a valuable treatment alternative. MADs reduce the upper airway collapse by advancing the mandible and associated soft tissue; as a result, there is generally an increase in the upper airway volume at the oropharynges level and improvement in pharyngeal patency. Occasionally, treatment of sleep apnea with MADs may be associated with the development of symptoms of Temporomandibular Disorder (TMD) and posterior open bite (POB). The definition of TMD includes clusters of signs and symptoms affecting the masticatory musculature, temporomandibular joints (TMJ), and associated structures. Pain is one of the most common symptoms of TMD. TMD disorders have a high prevalence in the general population, which increases the probability that OSA patients who are referred for MAD therapy could already present with TMD, or possibly could develop it during MAD therapy. Another side effect of MAD is POB, defined as no occlusal contact between the maxillary and mandibular posterior teeth when the patients try to hold their jaws in a maximum intercuspation position. This study aims to evaluate the incidence and prevalence of TMD in OSA patients prior to receiving MAD, and to determine the overall prevalence of TMD problems and POB during MAD therapy. This study also aims to evaluate the effectiveness of daily physical medicine exercise instructions on management of TMD and POB development during usage of MAD.

Materials and Methods: This was a retrospective study for which medical records of UCLA patients were reviewed. In order to determine prevalence of TMD in OSA patients prior to receiving MAD, the charts of 195 patients who were referred to the UCLA Orofacial Pain Clinic between June 2016 to June 2019 for MAD therapy were screened as the study group. One hundred and fifty-five patients met the study group criteria. Additionally, we screened the data of 400 new patients in the UCLA School of Dentistry’s general dental clinic from June 2016 to June 2019 as control group. A total of 188 patients were included in the control group.

One hundred and four subjects of the study group met the criteria to investigate the prevalence of TMD problems and POB during MAD therapy. One hundred and one of the 104 patients whose data contained records of physical exercise compliance were included to explore the effectiveness of physical exercise protocols to manage and prevent TMD and POB during MAD therapy. The follow-up PSG data with oral appliance in place were available for 16 patients. These were reviewed to evaluate the efficacy of MAD in reduction of AHI (Apnea–Hypopnea Index). Successful management of OSA with MAD therapy was considered as at least 50% reduction of AHI compared to the respective baseline AHI.

Results: At baseline, TMD prevalence was 94% among the OSA patients and 28% in the control group, which was significantly different (p<0.05). Sixty-six percent of OSA patients who received MAD did not experience painful TMD, 24% had myalgia and only 8% had both myalgia and arthralgia during MAD therapy. The number of patients with painful TMD was significantly reduced from 98 patients at baseline to 34 patients after receiving MAD (p<0.05). Physical exercise protocols effectively and significantly reduced painful TMD during MAD therapy (p < 0.05). Incidence of POB was 11.5%, but daily jaw repositioning physical exercises reduced the prevalence and incidence of POB during MAD therapy. MAD therapy had successful outcome in 75% of the patients.

Conclusion: The prevalence of TMD in OSA patients is significantly higher compared to the control group. Therefore, it is crucial for dentists who provide MAD for OSA patients to have appropriate training and background to assess, diagnose, and document preexisting TMD, and to be able to manage signs and symptoms of TMD before MAD therapy. Physical exercise protocols are effective and valuable tools to manage and prevent possible side effects of MAD usage including TMD and POB. Therefore, physical exercise instructions should be given at the MAD-delivery visit and be emphasized during follow-up visits.

Cover page of Latinas for Housing Justice: The Fight for Rent Control in Unincorporated East Los Angeles

Latinas for Housing Justice: The Fight for Rent Control in Unincorporated East Los Angeles


The hyper privatization of housing and gentrification in Los Angeles has dramatically shifted the racial, class, and cultural landscape of aggrieved neighborhoods. The timely need to address housing injustices––eviction, landlord harassment, and drastic rent increases in gentrifying neighborhoods is needed to further address mechanisms of capitalism that dehumanize and commodify housing. In this thesis, I analyze how Latina Mothers transform moments of housing injustices into empowerment and activism Focusing on LA County's rent control campaign led by local Los Angeles tenant coalition, Unincorporated Tenants United (UTU), I highlight the role of the Latina Mother canvassers. It was witnessed and analyzed that the Latina Mother canvassers enacted forms of placemaking, which I describe as homemaking while engaging with vulnerable tenants facing different forms of housing injustices. Informed by previous housing trauma, the Latina Mother canvassers displayed forms of homemaking, that I describe as: testimonio sharing, comadriando (gossiping), and (re)imagination of community through housing justice. Applying a homemaking framework to Woman of Color (WOC) Mother activism in aggrieved communities then allows for a nuanced analysis of social order, mechanisms of dispossession that influence local organizing, and collective community care that sustains a movement.

Diamond Integration for Thermal Management Applications: Transport, Boundaries, and Interfacial Morphology


Heat management problem limits the potential of further development of high-power devices. Diamond, with the highest thermal conductivity among all the materials, can be integrated near the hot spot region as a heat management component. The goal of this dissertation is to study grain morphology and thermal properties in polycrystalline diamond across boundaries using advanced characterization techniques. This work is anticipated to lead to an understanding of how to control and improve thermal transport in diamond. Diamond in device integration is achieved by chemical vapor deposition (CVD) at 750˚C with methane as the carbon source. A hydrogen plasma is used in the CVD process to remove any non-diamond carbon. In this research, transmission electron microscopy (TEM) and electron backscatter diffraction (EBSD) were used for grain morphology analysis. Both of the techniques can be used to separate the individual grains, which leads to an accurate size determination. In-plane thermal conductivity was measured with Raman thermography. Cross-plane thermal properties, including thermal boundary resistance (TBR), were measured with Time-domain thermoreflectance (TDTR). The combination of the advanced techniques has been done in several aspects to improve the diamond integration for better thermal transport.

The challenging issue for polycrystalline CVD diamond is the relatively lower thermal conductivity compared with single crystal diamond due to phonon scattering at grain boundaries, and the thermal conductivity continues to decrease closer to the nucleation region. At the interface, where the grain starts to nucleate and grow, the diamond structure is composed of small size, randomly oriented diamond grains. In this research, a larger diamond seed size is shown to increase the average diamond grain size near the nucleation region, which leads to a higher in-plane thermal transport for 1 μm diamond films. Also, with the combination of EBSD and spatially resolved TDTR, ~ 60% of thermal conductivity reduction is directly observed at the grain boundaries. While in the traditional model, the grain boundaries act as discrete thermal boundary resistances, this research has pointed out the impact is not restricted to the grain boundaries and can be observed in up to ~ 10 μm nearby region. Another challenge exists when integrating diamond on GaN-based high-power devices. The presence of hydrogen plasma at a high temperature can etch into the GaN substrate and cause a rough Diamond/GaN interface, which could increase the TBR value. In this research, an ultra-thin SiN is shown to be an effective protection layer, and a low TBR value of 9.5 m2K/GW is achieved.

Cover page of Housing Analysis and Prediction in Melbourne Australia

Housing Analysis and Prediction in Melbourne Australia


Using Melbourne, Australia dataset from, we analyzed the factors that determine the type of housing properties from Single-family house, Townhouse, and Apartment using multinomial logistic regression. We also predicted the property price using various machine learning algorithms such as Random Forest, Gradient Boosting, etc.