Calvin Thigpen

Modeling the Impact of Pavement Roughness on Bicycle Ride Quality

Many cities and states rely on aggregate seal coats (chip seals) to maintain roads. Chip seals are economical as a surface for lower-volume roads or for preservation treatments on asphalt roads, and the technology for increasing the life span of chip seals is continually improving. Chip seals often have higher macrotexture than do asphalt or concrete surfaces, which provides high skid resistance for motor vehicles. However, bicycling has increased in many parts of the country, for both recreation and commuting, and the high macrotexture of chip seals has led bicyclists to protest that the use of chip seals has decreased the comfort of recreational bicycle rides. In California, recreational bicyclists have contacted the California Department of Transportation (Caltrans) regarding the texture of chip seals on popular bicycling routes. Caltrans sponsored research to measure pavement surface macrotexture and roughness, to conduct bicycle ride quality surveys in partnership with six California bicycling clubs, and to do preliminary modeling. These data have been used to develop further a multilevel binomial regression model that indicated that surface macrotexture and roughness have a strong impact on perceived bicycle comfort levels. The model could be used to develop design guidelines for urban and rural bicycle routes.

Modeling Bicycling to Elementary and Junior High Schools with Bike Rack Counts

The decline in active travel to school and the concomitant rise in numbers of children being driven to school in the United States over recent decades have affected the health of school-age children and contributed to environmental problems. In response, communities throughout the country are stepping up efforts to increase active travel, including bicycling, but they have few tools available to them to assess the potential effectiveness of proposed strategies. The purpose of this study was to develop a model with aggregated school-level data of the factors associated with bicycling to elementary and junior high schools and to examine the effectiveness of this model in predicting bicycling to school. With the use of repeated observations of bike rack counts at 11 public schools in Davis, California, binomial multilevel regression models that included factors thought to influence bicycling to school were specified. The models indicated that comfortable bicycling routes, the racial and economic makeup of the student population, and various factors that represented the daily context (e.g., day of week, season, weather) all were likely to influence rates of bicycling to school. The results indicated that models based on aggregated school-level data were not sufficient to predict the amount of bicycling to a given school on a given day but were sufficient to predict mean bicycling to a given school over a period of time. Thus this method may be sufficient for policy analysis whose aim is to increase average bicycling to school.