Geoscience Reference
In-Depth Information
conjunction with the expected increase in parking search time (some of which may
occur outside the CBD).
In contrast to the procedure used for reducing parking supply, this second
measure is modeled by changing the attractiveness of private-car trips to the
CBD, compared to alternative transit modes. The main assumption here is that
the demand for vehicle trips is elastic; that is, the total demand for vehicle trips is
not fixed. Each scenario is composed of a different increase in parking fees. The
transportation model is applied to obtain new equilibrium link flows and travel
times for each scenario. The pollutant emissions are calculated for each scenario,
based on the VKT and travel speeds on each link of the transportation network.
21.3.3 Model Specification and Data
The transportation model used in this work was constructed by NTA—a semi-
public consulting group that led the Mass Transit project for the TAMA (NTA
2001
). It is a four-step model with feedback procedures for the mode-choice and
trip-distribution steps; that is, travel time from the traffic-assignment model serves
as input to previous steps of trip distribution and mode choice. The model
converges after a number of iterations between the steps described.
Figure
21.2
presents the division of the TAMA into ten super zones, based on
Israel's Central Bureau of Statistics. The transportation model calculates the VKT,
VHT, and average speed for every link of the transportation network. These results
are then processed by the emission model and aggregated at the super-zone level.
The trip-generation model is based on NTA's data, which were calibrated using
the 1995 National Census of Population and the 1996 National Travel Habit Survey
(NTHS). The Tel-Aviv metropolitan O-D matrix and network were updated to the
year 2003.
The parking-policy coefficients (parking cost and parking search time) were
obtained from NTA's mode-choice model, which employed a nested logit model.
The point elasticity of demand in the logit model is given as follows (Ben-Akiva
and Lerman
1985
):
E
¼
ð
1
P
Þ
β
x
ð
21
1
Þ
:
Where:
E
¼
Direct point elasticity
P
¼
Probability of traveling by private vehicle (60 % for the base case in TAMA)
β
¼
Parking cost coefficient
0.02883 and parking search time coefficient
0.0587
x
Parking cost or parking search time variables (approximately NIS 10
1
parking per hour and 5 min parking search).
¼
1
NIS 4.2
¼
$1.
