Development of FSUTMS Life Cycle and Seasonal Resident Trip Production Models for Florida Urban Areas
The standard FSUTMS trip production models calibrated in the early 1980's are stratified by dwelling unit type, household size, and auto ownership. These models predict, for example, the same number of work trips for a household resided by two retirees and a household resided by two workers, everything else being the same. When applied to zones dominated by retired residents, these models will grossly overestimate the total work trips. This inability of the existing models in estimating travel demand, coupled with the current trend to relate travel demand decisions to choices of activities and the allocation of roles for performing activities, have brought about an increasing orientation to family life cycles as a major determinant of travel demand.
Life cycle can be defined in terms of numbers of adults in the household, numbers of those adults who work or are in school, number of children present, and ages of children with respect to preschool versus K-12 school, day care, etc. A recent study by Ewing et al. of 29 FSUTMS users found that 87% of the respondents were in favor of refining the current FSUTMS production models with life cycle variables.
In addition to incorporating life cycle variables, another potential refinement to the standard FSUTMS trip production models is to separate the trip rates of seasonal residents from those of permanent residents. The 1992 Lee County Travel Characteristics Study has shown that the two groups have significantly different trip rates. It was found that seasonal residents, who are mostly retirees, make relatively fewer work trips but more shopping trips than permanent residents.
To produce better demand estimates for Florida urban areas, which have a large retired and seasonal population who tends to live in communities that segregate people by life cycle and cultural background, the Florida Statewide Model Task Force (MTF) has recommended the development of FSUTMS trip production models that are sensitive to life cycle and seasonal resident effects for Florida urban areas.
The use life cycle variables to predict travel demand is not new to Florida. In fact, the following three Florida urban regions have implemented life cycle regional trip production models:
- The Tampa Bay Region (Hillsborough, Pinellas, Hernando, Pasco, and sections of Citrus, Polk, and Manatee Counties) has implemented the Tampa Bay Regional Planning Model (TBRPM), which stratifies trips by households without children, households with children, and households with retirees.
- The FDOT District 4 has implemented a life cycle model for Broward and Palm Beach Counties. The model stratifies work trips by number of workers, auto ownership, and presence of children, and non-work trips by auto ownership, presence of children, and household size.
- The Treasure Coast Region (Martin, St. Lucie, and Indian River Counties) has implemented the Treasure Coast Regional Planning Model (TCRPM) with stratification similar to that of the Broward/Palm Beach model.
Many urban areas outside Florida have also implemented such models. A small sample of these models includes:
- San Francisco, California uses a model that stratifies work trips by number of workers and school trips by youthful age cohorts.
- Reading, Pennsylvania uses a model that stratifies trips into three groups: retired, not retired with children, not retired without children.
- Phoenix, Arizona uses the retirement status of an entire traffic zone as a stratifying variable. A zone is classified as a Aretirement zone@ if 50% or more of its population are retired.
- Portland, Oregon uses the age of the head of the household and number of children for work and school trips.
While the calibration of new trip rates and the software implementation of new models are relatively straightforward, the proper design and application of these new models are not. Specifically, the following questions must be considered when developing the models:
- What life cycle variables should be used and how should these variables be segmented? What other variables should be included?
- Can the data for life cycle variables be forecasted with reasonable ease and sufficient degree of confidence?
- What is the spatial transferability of these models? In other words, can existing models be used for other Florida urban areas?
- What is the temporal stability of these models? In other words, will the trip rates classified by life cycle variables remain sufficiently stable over the planing horizon at minimum?
- Should the models be a supplement to, or a replacement for, the present trip production models and what is the implication of the decision to FSUTMS model standardization?
This project will develop trip production models that are sensitive to life cycle and seasonal resident effects for Florida urban areas based on the design considerations listed above. The specific objectives include:
- To evaluate existing life cycle and seasonal resident models considering the sensitivity, spatial transferability, and temporal stability of their trip rates to the model variables, and the forecastability of these variables.
- To develop life cycle and seasonal resident models for other Florida urban areas, including Collier County, Lee County, Polk County Orlando, Volusia County, Gainesville, Tallahassee, West Palm Beach, Lee County, and Jacksonville.
- To develop a detailed application manual to guide the users in applying the new models.
- To develop or modify software programs to implement these new models.