For many chocolate is not a health food. It has been linked to medical conditions including acne (1) and in most forms it is high in fat and sugar. Recent data, however, has shed light on chocolate's cardio-protective roles. Specifically, cocoa has been found to have flavonoids which are high in antioxidant activity (3). Cocoa's role in platelet activity has also been investigated10 and has been shown to inhibit platelet activity which can further prevent cardiovascular disease. Investigators have also looked at chocolate's fat content and noticed that, although it is high in fat, most of it is in the form of stearic acid which is not well absorbed and excreted as feces (2). Although the data is limited, there is promise that chocolate may provide benefits beyond great taste. Until more research is completed, however, the quantity of chocolate should be limited, but each bite should be relished, not only for its flavor but for its health benefits as well.

Numerous studies have highlighted that water resources and hydrologic extremes are sensitive to climate change. An interesting research question is what the role of climate change is in occurrence of extreme events. More importantly, how climate extremes may change under future climate conditions and emission scenarios. Therefore, there exists a strong need to study water resources and hydrologic cycle under different climate change scenarios at the global scale.

In the past decades, numerous methods and models have been developed for assessing climate change impacts on water resources. However, there are still major research gaps from uncertainties in climate model simulations to limitations in the current large scale water cycle (or global hydrologic) models. Some of the current research gaps include: (I) high uncertainty of climate model simulations; (II) limitations and high uncertainties of the global hydrologic model simulations because of calibration challenges at the global scale; and (III) lack of frameworks for accounting for the local resilience and man-made infrastructure in climate impact assessment studies. The overarching goal of this study is to address the above mentioned research gaps.

In this dissertation, several novel evaluation metrics are introduced that can be used for evaluation of errors and biases in input data which is a key factor in the overall uncertainty of climate change studies. Furthermore, this study leads to a better representation of the hydrologic cycle at the global scale through a comprehensive multi-objective calibration framework for global hydrologic models. Then, a modeling framework is presented for accounting for local resilience in climate change studies. Finally, this study outlines a framework for combining top-down and bottom-up approaches for climate change impact assessment.

Electrocorticography (ECoG) arrays are used in clinical mapping for neurosurgical resection and hold the promise for less damaging brain-machine interfaces. Current clinical ECoG electrodes face physical limits to the number of contact sites, spatial resolution (centimeter scale), and contact diameter (millimeter scale), and thus cannot resolve the dynamically changing neural activity over sub-millimeter scales . In addition to these practical limitations, current clinical electrode arrays are constrained to non-conformal electrode-carriers/substrates and to less-optimal metal electrochemical interfaces. Increasing the flexibility of clinical electrodes may lead to higher signal-to-noise ratios as well as higher spatial specificity and this also requires overcoming substantial physical barriers due to the compromised metal electrochemical interface properties.

The objectives of this thesis, described in seven chapters, are to develop high performance, safe, and durable neural electrode interfaces to yield stable, high signal-to-noise ratio cortical recordings in animal models as well as in humans.

In the second chapter, we demonstrate that sterilization of PEDOT:PSS electrophysiology devices can be performed using an autoclave. We find that autoclaving is a viable sterilization method, leaving morphology unaltered and causing only minor changes in electrical properties. These results pave the way for the widespread utilization of PEDOT:PSS electrophysiology devices in the clinic.

In the third chapter, we translate the use of robust PEDOT:PSS microelectrode arrays for safe intraoperative monitoring of the human brain. PEDOT:PSS micro-electrodes measured significant differential neural modulation under various clinically relevant conditions. We report the first evoked (stimulus-locked) cognitive activity with changes in amplitude across pial surface distances as small as 400 μm, potentially enabling basic neurophysiology studies at the scale of neural micro-circuitry.

In the fourth and fifth chapters, we present the first systematic study of scaling effects on the electrochemical properties of Pt and Au metallic and PEDOT:PSS organic electrodes from neural recording and stimulation perspectives. PEDOT:PSS coating reduced the impedances of metallic electrodes by up to 18X. The overall reduced noise of the PEDOT:PSS microelectrodes enable a lower noise floor for recording action-potentials with high fidelity. We observed a substantial enhancement in charge injection capacity up to 9.5X for PEDOT:PSS microelectrodes compared to metal ones and 88% lower required power for injecting the same charge density. These results permit quantitative optimization of contact material and diameter for different ECoG applications.

In the sixth chapter, We report an effective method of mechanically anchoring the PEDOT within the Au nanorod (Au-nr) structure and demonstrate that it provides enhanced adhesion and overall PEDOT layer stability under various electrochemical (charge injection) and In vivo stability tests.

In the seventh chapter, we report the fabrication of pure Pt nanorods (PtNRs) by utilizing low-temperature selective dealloying to develop scalable and biocompatible 1D platinum nanorod (PtNR) arrays that exhibit superb electrochemical properties at various length scales for high-performance neurotechnologies. PtNR arrays record brain activity with cellular resolution from the cortical surfaces in birds, mice, and non-human primates; demonstrating the PtNR microelectrode system as a robust system for high performance and stable neural electrode interfaces.

Existing face recognition algorithms perform well under controlled conditions; however, in real-world scenarios, where variations in pose, illumination, scale, rotation, and expression are encountered, they are unable of performing accurately and/or efficiently. As an example, in the case of the Boston marathon bombings on April 15th, 2013, although the suspects' face images existed in government databases (e.g., DMV), the face recognition algorithms were unable to identify the suspects before they were identified by other means. Motivated by the demand of reliable face recognition techniques for real-world scenarios, we are tasked with identifying an unknown face, or verifying a match between a pair of unknown faces. The process of doing so requires overcoming many uncertainties in real-world images including pose and scale variation, misalignment, varying illumination conditions, etc. Towards this goal, two new recognition frameworks, a graph-based probabilistic pose estimation classifier, and a hash-based retrieval algorithm were designed to accurately perform face recognition in large-scale real-world applications. In addition, a novel real-time video-based classification system was developed for robust vehicle classification.