Joint probability methods for assessing storm surge flood risk involve the use of a collection of hydrodynamic storm simulations to fit a response surface model describing the functional relationship between storm surge and storm parameters like central pressure deficit and the radius of maximum wind speed. However, in areas with a sufficiently low probability of flooding, few storms in the simulated storm suite may produce surge, with most storms leaving the location dry with zero flooding. Analysts could treat these zero-depth, “non-wetting” storms as either truncated or censored data. If non-wetting storms are excluded from the training set used to fit the storm surge response surface, the resulting suite of wetting storms may have too few observations to produce a good fit; in the worst case, the model may no longer be identifiable. If non-wetting storms are censored using a constant value, this could skew the response surface fit. The problem is that non-wetting storms are indistinguishable, but some storms may have been closer to wetting than others for a given location. To address these issues, this thesis proposes the concept of a negative surge, or “pseudo-surge”, value with the intent to describe how close a storm came to causing surge at a location. Optimal pseudo-surge values are determined by their ability to improve the predictive performance of the response surface via minimization of a modified least squares error function. We compare flood depth exceedance estimates generated with and without pseudo-surge to determine the value of perfect information. Though not uniformly reducing flood depth exceedance estimate bias, pseudo-surge values do make improvements for some regions where <40% of simulated storms produced wetting. Furthermore, pseudo-surge values show potential to replace a post-processing heuristic implemented in the state-of-the-art response surface methodology that corrects flood depth exceedance estimates for locations where very few storms cause wetting.

<p><a>Proactive evaluation of road safety is one of the most important objectives of transportation engineers. While current practice typically relies on crash-based analysis after the fact to diagnose safety problems and provide corrective countermeasures on roads, surrogate measures of safety are emerging as a complementary evaluation that can allow engineers to proactively respond to safety issues. These surrogate measures attempt to address the primary limitations of crash data, which include underreporting, lack of reliable insight into the events leading to the crash, and long data collection times. </a></p> <p>Traffic conflicts are one of the most widely adopted surrogate measures of safety because they meet the following two conditions for crash surrogacy: (1) they are non-crash events that can be physically related in a predictable and reliable way to crashes, and (2) there is a potential for bridging crash frequency and severity with traffic conflicts. However, three primary issues were identified in the literature that need to be resolved for the practical application of conflicts: (1) the lack of consistency in the definition of traffic conflict, (2) the predictive validity from such events, and (3) the adequacy of traffic conflict observations.</p> <p>Tarko (2018) developed a theoretical framework in response to the first two issues and defined traffic conflicts using counterfactual theory as events where the lack of timely responses from drivers or road users can produce crashes if there is no evasive action. The author further introduced a failure-based definition to emphasize conflicts as an undesirable condition that needs to be corrected to avoid a crash. In this case, the probability of a crash, given failure, depends on the response delay. The distribution of this delay is adjusted, and the probability is estimated using the fitted distribution. As this formal theory addresses the first two issues, a complete framework for the proper identification of conflicts needs to be investigated in line with the failure mechanism proposed in this theory.</p> <p>The objective of this dissertation, in response to the third issue, is to provide a generalized framework for proper identification of traffic conflicts by considering the failure-based definition of traffic conflicts. The framework introduced in this dissertation is built upon an empirical evaluation of the methods applied to identify traffic conflicts from naturalistic driving studies and video-based tracking systems. This dissertation aimed to prove the practicality of the framework for proactive safety evaluation using emerging technologies from in-vehicle and roadside instrumentation.</p> <p>Two conditions must be met to properly claim observed traffic events as traffic conflicts: (1) analysis of longitudinal and lateral acceleration profiles for identification of response due to failure and (2) estimation of the time-to-collision as the period between the end of the evasion and the hypothetical collision. Extrapolating user behavior in the counterfactual scenario of no evasion is applied for identifying the hypothetical collision point.</p> <p>The results from the SHRP2 study were particularly encouraging, where the appropriate identification of traffic conflicts resulted in the estimation of an expected number of crashes similar to the number reported in the study. The results also met the theoretical postulates including stabilization of the estimated crashes at lower proximity values and Lomax-distributed response delays. In terms of area-wide tracking systems, the framework was successful in identifying and removing failure-free encounters from the In-Depth understanding of accident causation for Vulnerable road users (InDeV) program.</p> <p>This dissertation also extended the application of traffic conflicts technique by considering estimation of the severity of a hypothetical crash given that a conflict occurs. This component is important in order for conflicts to resemble the practical applications of crashes, including the diagnostics of hazardous locations and evaluating the effectiveness of the countermeasures. Countermeasures should not only reduce the number of conflicts but also the risk of crash given the conflict. Severity analysis identifies the environmental, road, driver, and pre-crash conditions that increase the likelihood of severe impacts. Using dynamic characterization of crash events, this dissertation structured a probability model to evaluate crash reporting and its associated severity. Multinomial logistic models were applied in the estimation; and quasi-complete separation in logistic regression was addressed by providing a Bayesian estimation of these models.</p>