Survey Information
Methodology
Survey Calendar Periods and Dates
Survey flights were normally conducted on Tuesday, Wednesday or Thursday mornings and evenings (one Friday morning was also authorized / Nov. 16th 2007). Morning survey periods were from 6:30 to 9:30 a.m.; evening periods were from 4:00 to 7:00 p.m. Coverage of the primary highway system (inside the NYC boundaries) occurred during the two-week window of November 5th to 16th, with additional flights in January 2008 after the end of the holiday season. Supplementary aerial flights covering the Nassau and Westchester County highways occurred between Nov. 28th and Dec 6th 2007, or later between January and March 2008.
Survey Operations
Highways to be surveyed were designated by Parsons Brinckerhoff. Altogether, approximately 275 centerline miles of highway were chosen for primary coverage inside the NY City boundaries; an additional 100 highway miles in adjacent Nassau and Westchester Counties were also designated. Skycomp divided this network into separate tours (routes) that could each be covered by a fast-moving airplane in about 45 minutes. During morning and evening survey periods, multiple airplanes were launched, each with the mission to follow its assigned route four times around, producing four observations per flight. Flights were repeated on multiple days to produce an average of 12 observations of each highway link during the morning survey period, plus 12 samples during the evening survey period. While in-flight, a camera-equipped observer acquired overlapping still digital photographs of all designated highway segments; resolution of these images in daylight was high enough to clearly show vehicle travel lanes and pavement striping; the field-of-view was also wide enough to show traffic flow conditions on crossroads and parallel streets. (Because of the time-of-year, about 2/3rds of evening coverage occurred during periods of darkness; while photography acquired during periods of darkness was not as detailed as in daylight, in most cases less detail was offset by the clarity of vehicle headlights shown against darker backgrounds.)
Data extraction and analysis
The archive of overlapping digital photographs was used to count the population of vehicles on each highway link, using a computer-assisted manual counting procedure. For uninterrupted-flow highways, raw counts were taken (classified as cars, trucks, tractor trailers or buses when ambient lighting permitted), and then entered into a computer program for the computation of densities. For interrupted-flow (signalized) arterials, a surrogate level-of-service measurement system was used; this is a comprehensive set of guidelines created by Skycomp to classify traffic flow qualitatively based on platoon characteristics and the nature of queues at traffic signals (select the link named "Survey Information" to the left for a detailed description of the performance rating systems). After initial data reduction and key punching into a database, an analyst scrubbed output data tables to identify and confirm zones of congestion, and to identify and exclude data affected by incidents or other anomalies. Underlying photographs were examined to determine the apparent cause of all bottlenecks, and a bottlenecks database table was produced detailing the characteristics of each congested zone. Highlight photographs were also selected that best depict the congested conditions found at each bottleneck.
Creation of deliverables: Skycomp then prepared standard time/space level-of-service display graphics, including associated descriptions of the apparent underlying causes of each congested zone taken from the bottlenecks database table. These graphics form the main body of the primary report, delivered in digital PDF format (select the link named "Additional Resources" to download a copy of this report). Skycomp also prepared this web-based slide show, with separate morning and evening graphics showing the location of all bottlenecks. In this product, red and orange bottleneck arrowheads are linked to pop-up windows displaying bottleneck characteristics and the associated highlight photographs. Separate arrowhead graphics were also created depicting the traffic conditions associated with major incidents which occurred during the survey flights, aligned with the maps showing normal bottleneck arrowheads. Skycomp also prepared a GIS "look-up" table, allowing traditional GIS and Google Earth programmers to display Skycomp's performance ratings, bottleneck characteristics and highlight photo modules in GIS and Google Earth environments. This slide show is also linked to the underlying database of performance ratings, which can be accessed through the data extraction module at the link named "Survey Data". This module allows users to filter data to produce custom comparative bar charts, LOS displays and data tables for personal use.