SAFETYLIT WEEKLY UPDATE

We compile citations and summaries of about 400 new articles every week.
RSS Feed

HELP: Tutorials | FAQ
CONTACT US: Contact info

Search Results

Journal Article

Citation

Matsuyama Y, Miyoshi T, Matsuda Y, Matsuzawa M, Kajiya Y. Transp. Res. Circular 2008; (E-C126): 612-624.

Copyright

(Copyright © 2008, U.S. National Academy of Sciences Transportation Research Board)

DOI

unavailable

PMID

unavailable

Abstract

Roads in Hokkaido, the northernmost island of Japan, suffer poor visibility from snowfall, snowstorm and dense fog. Fog often occurs at mountain passes separating the Pacific Coast of Hokkaido from inland areas. Such visibility reduction increases the risk of rear-end collisions, because leading vehicles are recognized later than in clear weather. To support trip planning, it is important to provide drivers with fog forecasts in addition to conventional winter road weather information. The authors provided fog forecasts via Internet on a trial basis for Nissho Pass on National Highway 274, where fog occurs in every season. Fog was forecast using a statistical model based on analyses of weather and fog data for the past three years. They found that fog tends to occur when the humidity is high and weak winds blow from cardinal points ranging from east to south and that the most influential weather factors in fog occurrence are wind speed, wind direction, and humidity. Based on these three factors, they categorized weather conditions into six combinations. They divided the road section that includes Nissho Pass into six subsections. The fog percentage (hours of fog/hours of that weather condition × 100) for each of the six combinations of weather conditions on each of the six road sub-sections was calculated from meteorological records. This percentage is regarded as the probability of fog occurrence, and the probabilities were classified into the four ranges of 0%–10%, 10%–40%, 40%–70%, and 70%–100%. The model output a current fog forecast and a forecast for each of the next 6 h as one of the four fog probability ranges, and these were provided by web page in July 2006. To determine the forecast accuracy, the authors analyzed road image data recorded during the same period. A fog percentage was calculated for each of the six combinations of weather conditions on each of the two road subsections of the mountain pass by visually judging image data for fog. Each observed fog percentage fell within the probability range of the fog forecast. Finally, to evaluate the usefulness of the fog forecast web page, the authors surveyed website visitors by online questionnaire. Eighty percent of the respondents responded that they thought fog forecasts were useful, attesting to their high acceptance of the system.

NEW SEARCH


All SafetyLit records are available for automatic download to Zotero & Mendeley
Print