Despite these recent findings for the existence of well defined MFDs for urban areas, these MFDs should not be universally expected. In particular, networks with an uneven and inconsistent distribution of congestion may exhibit traffic states that are well below the upper bound of an MFD and much too scattered to line along a well defined diagram. An inconsistent distribution of congestion is typical of freeway networks with multiple recurrent and non-recurrent bottlenecks, and of large urban areas with multiple congested sub-centers. Hence, a proper parameterization and description of heterogeneity that would yield good estimates of macroscopic representation of the network is of critical importance. Several questions related to this issue are part of LUTS research agenda. Some examples are:
- How a heterogeneous urban network can be partitioned in neighborhoods with well-defined MFDs?
- Should this partitioning be time-dependent?
- How are the traffic dynamics of this multi-partition system?
- How can we control inter-transfers between successive neighborhoods to increase cities’ mobility and accessibility measures?
- What is the level of information needed from different types of sensors (fixed, mobile etc) to accurately monitor the system and give answers to the above questions?
- How can we integrate complex phenomena in the macroscopic modeling of urban networks (e.g. the effect of parking, elastic demand etc)
- How can we model multimodal systems with many different modes (buses, cars) competing for the same space?
Geroliminis, N., (2009), Dynamics of the rush hour and the effect of parking for congested cities, Transportation Research Board 2009 Annual Meeting, Washington, DC