Date of Award

Spring 5-6-2022

Level of Access Assigned by Author

Open-Access Thesis

Degree Name

Master of Science (MS)


Civil Engineering


Mohammadali Shirazi

Second Committee Member

Per Garder

Third Committee Member

Shaleen Jain


Lane departure crashes account for approximately 34% of all roadway crashes, and over 70% of all roadway fatalities in Maine. Despite an 18% decrease in average daily traffic volume during the half of the year with colder weather, the months of November to April comprise over 64% of the yearly lane departure crashes. The majority of roadways and roadway crashes in Maine are in rural regions. Moreover, Maine has aging infrastructure, houses the oldest population in the United States, has diverse terrain and land use, and experiences several extreme weather events. The combination of these factors impacts the frequency and severity of crashes in Maine.

This thesis first investigates the impact of various weather factors on frequency of monthly crashes in Maine. A Negative binomial model with panel data was used to analyze monthly crashes on Interstates, minor arterials, major collectors, and minor collectors from 2015 to 2019 for winter (November to April) and non-winter periods (May to October). The data include monthly average daily traffic, geometric characteristics, and monthly weather variables. Second, the impact of roadway, driver and weather factors on the severity of single-vehicle lane departure crashes occurring from 2017 to 2019 on rural roadways in Maine is analyzed using a multinomial logit (MNL) model. Four facility types: Interstates, minor arterials, major collectors, and minor collectors were considered for analysis. Four severity outcomes were considered including, fatal-incapacitating injury crashes (KA), non-incapacitating injury (B), possible injury (C) and property damage only (PDO). The PDO outcome was used as the reference (or base) category in the MNL model.

To better account for weather factors, instead of using police reported weather, daily weather data was obtained from National Oceanic and Atmospheric Administration online resources for a total of 16 weather stations and matched to road segments in the created regions. To analyze the impact of monthly weather factors on frequency of crashes, the weather variables were aggregated to monthly values. To analyze severity of crashes, daily weather variables for the time of crash was considered.

Factors that were found to positively impact frequency of lane departure crashes on all considered facilities include the number of days in a month that experience more than one inch of precipitation or snowfall and posted speed limit. Factors that were found to impact higher crash severity for all facilities considered include not using a seatbelt, a crash resulting in a vehicle rollover, speeding, operating under the influence and higher temperatures. Factors that were found to lower crash severity include days with more than one inch of snowfall, precipitation and winter season crashes.

The analysis provides safety analysts and practitioners in Maine a comprehensive study of factors that influences the frequency and severity of rural lane departure crashes in Maine at different facilities to improve maintenance strategies, enhance safety using proper safety countermeasures, or increase awareness across the state.