Eric Yamashita

Influence of Land Use, Population, Employment, and Economic Activity on Accidents

In this study, the relationships between land use, population, employment by sector, economic output, and motor vehicle accidents are explored. Through the use of comprehensive data from the largest county in Hawaii, the relationships are modeled in a uniform 0.1-mi2 (0.259-km2) grid structure and with various linear regression models. This method has an advantage over other approaches that have typically used unevenly sized and shaped traffic analysis zones, census tracts, or block groups. Positive, statistically significant relationships among population, job counts, economic output, and accidents are identified. After some of the general effects are sorted through, a negative binomial (NB) model is used to look at the absolute and relative effects of these factors on the number of pedestrian, bicycle, vehicle-to-vehicle, and total accidents. With a multivariate model, the different effects can be compared and the specific nature of the relationships between zonal characteristics and accidents can be id...

MOTOR VEHICLE CRASHES AND LAND USE: EMPIRICAL ANALYSIS FROM HAWAII

At first glance, the relationship between crashes and land use appears obvious. Various types of land uses tend to generate and attract different types of trips, and trip-making behavior affects the nature and volume of traffic. As the use of land intensifies, it does not seem unreasonable to expect that the potential exposure to crashes would also increase. Yet upon closer inspection, it is evident that crashes are more a function of the characteristics of drivers and travelers than the underlying uses of land. Using comprehensive police crash data linked to a land use database, the relationships between land use and automobile crashes in Hawaii are investigated. Recent developments in geographic information system technology and the availability of spatial databases provide a rich source of information with which to investigate the relationships between crashes and the environments in which they occur. Although the patterns in terms of crashes and the underlying use of land are difficult to describe, the implications of these findings for planners, developers, and those interested in traffic safety are apparent.

Hit-and-Run Crashes: Use of Rough Set Analysis with Logistic Regression to Capture Critical Attributes and Determinants

In this paper, an innovative mathematical tool, rough set analysis (RSA), combined with logistic regression modeling, is used to understand the key factors associated with hit-and-run collisions in Hawaii. After a description of the nature of the problem in Hawaii and some background on the RSA, the methods are applied to a comprehensive database of police-reported accidents over the period 2002 to 2005. RSA is used to extract the key determinants of hit-and-run collisions. With the information from the RSA, a logistic regression model is built to explain the key factors associated with hit-and-run crashes in Hawaii. Factors such as being (a) a male, (b) a tourist, and (c) intoxicated and driving a stolen vehicle are strong predictors of hit-and-run crashes. In addition to the obvious human factors associated with these crashes, there are interesting roadway features, such as horizontal alignment, weather, and lighting, that are also significantly related to hit-and-run crashes. Some suggestions for reducing hit-and-run crashes as well as opportunities for additional research are identified.

TYPOLOGY OF MOTORCYCLE CRASHES: RIDER CHARACTERISTICS, ENVIRONMENTAL FACTORS, AND SPATIAL PATTERNS

Research was conducted on motorcycling and alcohol-impaired motor-cycling in Hawaii to develop a crash typology. The investigation expanded the scope of rider characteristics analysis by examining the combined effects on motorcycle crashes of rider behavior (helmet use, speeding, actions taken before a crash, etc.), environmental factors (urban versus rural locations, roadway alignment, etc.), and spatial patterns. The crash typology was used to derive logistic regression models for explaining alcohol-involved crashes, single-vehicle crashes, and injury outcomes (including fatalities) associated with motorcycle crashes. The logistic models allowed comparisons of the relative importance of various rider characteristics, temporal and environmental correlates associated with motorcycle crashes, and the associated crash types and injury outcomes. Finally, a spatial cluster analysis was also performed with both geographic information system and spatial analytical tools. The analysis suggests that overall behavioral and temporal factors are more significant predictors of alcohol-involved crash patterns than environmental or roadway features. The findings are qualified in terms of the usefulness of the methods to motorcycle safety researchers and relevance to motorcycle safety initiatives in the state of Hawaii.

Corresponding Characteristics and Circumstances of Collision-Involved Pedestrians in Hawaii

Correspondence analysis is used to explore the relationships between the characteristics and circumstances of collision-involved pedestrians in Hawaii. The purpose of this study is twofold. First, it is to apply and understand how best to use a powerful exploratory data technique (correspondence analysis) for safety research. The second purpose is to understand better the nature and causes of pedestrian accidents. The circumstances associated with pedestrian fault, serious injury, and the effects of location and temporal factors are examined. Correspondence analysis is a technique that allows one to examine the relationships between various discrete or categorical variables and to consider the effects of gender, age groupings, time, and location on pedestrian fault and injury. The technique is useful not only in clarifying the important relationships but also in designing appropriate strategies and interventions for the reduction of pedestrian accidents. The analysis suggests that greater emphasis be directed toward drivers, who are 13.8 times more likely to be classified as being at fault than the pedestrian they collide with. It is also evident that there are differences between seniors and other pedestrians in terms of injuries sustained. Moreover, specified programs might be targeted to men, children, and young adults.

ALCOHOL-IMPAIRED MOTORCYCLE CRASHES IN HAWAII, 1986 TO 1995: AN ANALYSIS

The purpose of this research is to review and describe data on the nature and extent of alcohol-impaired motorcycle-riding crashes in Hawaii. Using comprehensive police crash data, the differences between impaired and nonimpaired riders involved in crashes, in terms of various demographic characteristics, helmet use, license status, roadway environments, and injury outcomes, are described. In addition to various demographic and behavioral factors associated with impaired rider crashes, clear temporal patterns also have been detected. After summarizing these effects, a logistic regression model is built to explain the likelihood of an impaired motorcycle crash as a function of rider characteristics as well as environmental and temporal factors. In addition to age-specific determinants of impaired crash involvement, time factors such as weekends and nighttime, and licensing (out-of-state or no licenses) are strongly associated with impaired rider crashes. These findings could be used to enhance enforcement efforts as well as public information and safety education programs.

Walking in Waikiki, Hawaii: Measuring Pedestrian Level of Service in an Urban Resort District

This study examines pedestrian conditions in Waikiki, the most urbanized area in Honolulu, Hawaii. Focusing on one side of approximately a four-block stretch of Kalakaua Avenue, a series of 15-min pedestrian counts was conducted to establish overall sidewalk volumes and flows. Detailed maps were produced to characterize the physical environment and to identify potential factors associated with measurement of the pedestrian environment. Next, the pedestrian level of service (LOS) was estimated. A variety of factors affect the pedestrian environment. In addition to the width of the sidewalk, there are also movable and immovable objects, street furniture, plantings, and other activities that affect access and use of sidewalks. The intent of this study was to understand and measure better the impacts of sidewalk activities on LOS. To test and refine the approach, the impacts of street performers who occupied the sidewalks in the area were analyzed in terms of the effects on pedestrian flow and LOS. Street performers create the following kinds of impacts to the pedestrian environment: (a) increased delay, (b) direction changes and longer travel paths for pedestrians, (c) increased traffic conflicts as pedestrians step into moving motorized traffic to avoid congestion, (d) disruption of pedestrian platoons or preferred side of sidewalk for travel (generally, most pedestrians prefer to walk on the right side of the sidewalk), and (e) increased confusion associated with crowding and forced changes in direction. Policy considerations as well as directions for research are discussed in a concluding section.

Evacuation planning for plausible worst case inundation scenarios in Honolulu, Hawaii

Honolulu is susceptible to coastal flooding hazards. Like other coastal cities, Honolulu&s long-term economic viability and sustainability depends on how well it can adapt to changes in the natural and built environment. While there is a disagreement over the magnitude and extent of localized impacts associated with climate change, it is widely accepted that by 2100 there will be at least a meter in sea level rise (SLR) and an increase in extreme weather events. Increased exposure and vulnerabilities associated with urbanization and location of human activities in coastal areas warrants serious consideration by planners and policy makers. This article has three objectives. First, flooding due to the combined effects of SLR and episodic hydro-meteorological and geophysical events in Honolulu are investigated and the risks to the community are quantified. Second, the risks and vulnerabilities of critical infrastructure and the surface transportation system are described. Third, using the travel demand software, travel distances and travel times for evacuation from inundated areas are modeled. Data from three inundation models were used. The first model simulated storm surge from a category 4 hurricane similar to Hurricane Iniki which devastated the island of Kauai in 1992. The second model estimates inundation based on five tsunamis that struck Hawaii. A 1-m increase in sea level was included in both the hurricane storm surge and tsunami flooding models. The third model used in this article generated a 500-year flood event due to riverine flooding. Using a uniform grid cell structure, the three inundation maps were used to assess the worst case flooding scenario. Based on the flood depths, the ruling hazard (hurricane, tsunami, or riverine flooding) for each grid cell was determined. The hazard layer was analyzed with socioeconomic data layers to determine the impact on vulnerable populations, economic activity, and critical infrastructure. The analysis focused both on evacuation needs and the critical elements of the infrastructure system that are needed to ensure effective response and recovery in the advent of flooding. This study shows that the coastal flooding will seriously affect the economy and employment. Extreme flooding events could affect 38 percent of the freeways, 44 percent of the highways, 69 percent of the arterial roads, and 40 percent of the local streets in the area examined. Approximately 80 percent of the economy and 76 percent of the total employment in the urban core of Honolulu is exposed to flooding. Evacuation modeling, shelter accessibility, and travel time to shelter analyses revealed that there is a significant shortage in sheltering options, as well as increases in travel times and distances as inundation depth increases. The findings are useful for evacuation and shelter planning for extreme coastal events, as well as for climate change adaptation planning in Honolulu. Recommendations for emergency responders as well as those interested in the integration of long-term SLR and low probability, high consequence coastal hazards are included. The study shows how to integrate travel demand modeling across multiple hazards and threats related to evacuating, sheltering, and disaster risk reduction.

Use of Safety Viewgrams to Visualize Driver and Pedestrian Interactions

The purpose of this paper is to describe the development of a tool for visualizing, at a glance, the salient features of the interaction between motorists and pedestrians. Simple, yet robust and flexible, tools are needed to describe and summarize real-world roadway conditions for problem identification and to monitor the progress that has been made toward achieving safety goals. In the study described in this paper, the safety viewgram has been developed to help visualize various types of traffic safety problems. Although this particular tool has been developed by using observational data for pedestrians and drivers in Hawaii, it could be used for other types of safety studies in other locales. After the development and use of this tool are described, a number of different illustrative examples of how it might be used are provided. Although this is largely a visualization tool, efforts are under way to devise both more rigorous statistical applications and a user-friendly interface for data entry, data manipulation, and presentation of the traffic safety viewgram.

Asleep at the Wheel: Spatial and Temporal Patterns of Fatigue-Related Crashes in Honolulu

As an island state located in the middle of the Pacific Ocean, where there is limited opportunity for long-distance driving, Hawaii provides an interesting context in which to study fatigue-related crashes. Data from the Hawaii Crash Outcome Data Evaluation System are used to analyze and map fatigue-related collisions. Injury outcomes of fatigue-related crashes are analyzed by using police crash data, emergency medical service records, and insurance claims records. There are distinct temporal and spatial patterns as well as relationships between fatigue-related crashes and driver characteristics. Recommendations for preventing fatigue-related crashes are developed. Roadway segments where fatigue-related crashes occur are identified as possible sites for various engineering treatments. Temporal and demographic information also can be used to design and implement more effective programs and systems for fatigue-related crashes.