Development of Time-space Diagram showing Changes in Visibility Information along the Route in Winter

Abstract

In Hokkaido, Japan, poor visibility caused by snowstorms can lead to a disaster that may affect human lives. The present study proposes a time-space diagram showing changes in visibility information every 5 minutes at 0.1 km intervals by combining the visibility measured by closed-circuit television cameras located along the national highways and by on-board video cameras in the intercity bus. To find out when and where visibility is poor along the route, the regression models were used to estimate the visibility information. This allows the road users and the road administrator to know the visibility information every 5 minutes at 0.1 km intervals. However, the estimation accuracy of the regression models was not high. It is necessary to increase the amount of data for poor visibility conditions. Also, the estimation model should be improved by using a non-linear model.