Quality Measure of Short-Duration Bicycle Counts

Abstract

The average annual daily bicyclists (AADB) measure is commonly used in research and practice as a metric for cycling studies, such as bike ridership analysis, infrastructure planning, and injury risk. It is estimated in one of two ways: by averaging the daily cyclist totals measured throughout the year with a long-term automated bicycle counter, or by using a long-term bicycle counter to extrapolate data from a short-term counting site. Unfortunately, extrapolation of a short-term bicycle counting site can produce inaccurate AADB estimates as a result of different error sources; the range of possible error is highly correlated to several characteristics of the short-term count, such as the counting period, flow intensity, and time of year. This paper proposes a simple method to estimate the quality of a short-term count through a single metric combining five factors associated with the count variation: duration, average demand, time of year, stability, and correlation with the reference count. The method is validated with the use of a relatively large data set of automated bicycle counts. The quality measure, with a range from 0 to 10, is negatively correlated with the absolute relative error (ARE) of the AADB estimation. Results show distinct ARE distributions for different quality measure classes. The average ARE for the lowest quality class is 13.5% compared with an average ARE of 3.0% for the highest quality class. The maximum ARE (95% confidence) is 35% for the lowest quality class compared with 7.5% for the highest quality class.

Identifying High-Risk Intersections for Walking and Bicycling Using Multiple Data Sources in the City of San Diego

“…Extrapolation has been used in several studies Journal of Advanced Transportation 3 for estimating Annual Average Daily Pedestrian (AADP) and Annual Average Daily Bicyclist (AADB) [4,6,25,44,45]. When collecting short-term counts, several factors such as counting period length, time of day, month, and year could potentially impact the accuracy of yearly count estimation [3,4,6,[45][46][47]. In addition to collecting short-term data, long-term count data at several locations are also required to perform the extrapolation.…”

“…Several extrapolation methods have been used in previous studies such as traditional or standard method, Dayby-month, day-of-year, and weather model [4,6,9,45,47]. The selection of the method depends on the geographic location and weather variation of the study area as well as the availability and duration of short-term and long-term counts.…”

“…Therefore, some methods were proposed to incorporate the influence of weather on AADP and AADP [7,9]. Another consideration in the extrapolation process is to apply a suitable method for matching short-term counters to appropriate long-term counters [6,7,9,30,[47][48][49][50].…”

Minimizing Annual Average Daily Nonmotorized Traffic Estimation Errors: How Many Counters Are Needed per Factor Group?

“…MAPE in AADBT ranges from 7.8 to 19.1% depending on length of sample count (1-14 days) and season of count (April-October versus year-round) Use of correction equations reduces range of error to 7.9 to 15.6% Day-of-year factors produce lower AADBT error than traditional, day-of-month, and weather model methods MAPE in AADBT for one-day counts was 12-13% for day-ofyear factors and nearly 22% for traditional factors. For sevenday counts, MAPE was 10% for day-of-year factors and 13% Beitel et al(22) Day-of-year for four factor groups (adapted Miranda-Moreno et al (14)) K-means clustering to sort shortduration counts into factor groups Quality of short-duration counts can be characterized by duration of count, average bicycle demand, time of year, stability of count, and correlation with the reference count Average relative error ranges from 3% for highest quality counts to 13.5% for lowest quality counts (continued)…”

Automated Validation and Interpolation of Long-Duration Bicycle Counting Data

“…On the other hand, estimating AADB with a short-term count and extrapolation techniques, although relatively inexpensive, will produce results of varying accuracy. The accuracy of an AADB estimation is known to vary with several characteristics of the short-term count, including the time of year the data were collected, the count duration, and the level of cycling demand ( 3 , 5 ), as well as the goodness of match between the short-term count and the reference(s) used in extrapolation ( 6 ).…”