Geography Reference
In-Depth Information
often visit for shopping, entertaining, etc. Most importantly, they were asked to
record each trip and corresponding details such as date, start/end time, activity type,
destination/place name and street address.
To make the data collection process smoothly, participants were encouraged to
charge the battery of GPS device timely. As for the potential problems of GPS signal
blockage and data loss caused by high-rise building and cold/warm start, they were
reminded to pay attention to the indicator light of GPS signal on the device and try
to make sure that the device is at work all the time. Respondents were encouraged
to report any problem with the GPS device during the data collection process.
In most existing studies on GPS applications, tracking data were recoded for
every seconds. For example, Moiseeva et al. (
2010
) collected data with a time
interval of 3 s, while Bohte and Maat (
2008
) logged data every 6 s. With such
detailed data, one is able to reconstruct every single movement or standstill of the
object. However, if a GPS receiver logs individuals' positioning information second
by second continuously for 24 h, the amount of data would be enormous. These
massive data not only cause much trouble in data preparation because of the need to
remove a great volume of redundant data (not to mention the tedious work of data
filtering for removing noise data), they also post high requirement for the memory
capacity and battery life of GPS devices. In fact, if the purpose is to collect activity-
travel behavior data, not the exact and detailed trajectories of movement, in other
words, if the purpose is to distinguish activity and travel, one does not need the
GPS tracking data in such a frequency. For all these reasons, our GPS tracking data
were mostly recorded for every 1-min. To support our argument and for comparison
purposes, some of them were recorded for every 3 min. The 1-min data involves 100
person-days and 549 reported activities, whilst the 3-min ones 68 person-days and
310 reported activities. As evidenced later, the successful rate of detecting trips and
activities from 3-min data is not very much different from that of 1-min data.
13.4.3
Collection of GIS Land Use Data
The following GIS land-use data were collected: Guangzhou administrative bound-
ary, a complete road network, as well as Point of interests (POIs) including
restaurants, banks, post offices, medical and sports facilities, places for entertain-
ment, shopping places, schools, government departments, places of welfare and
religious activities in both point and polygon patterns.
Among these GIS data, Guangzhou administrative boundary and road network
were collected from Statistics Bureau of Guangzhou while the layer of restaurants
came from National Geomatics Center of China (NGCC). These data were in
WGS 84 (short for World Geodetic System 1984) Coordinate System, which is
a commonly used geocentric coordinate system worldwide and is currently the
reference system being applied to the Global Positioning System. All the other
POI layers were collected from Guangzhou Bureau of Urban Planning in the same
Search WWH ::
Custom Search