Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
controls distinguish âsingleâ from â2+â persons per household. For families with householders over 65 years of age, the distinction by presence of children is ignored. For the forecast year, fewer controls are defined for each transportation analysis zone (TAZ). These capture ARC TAZ- level forecasts of household income and household size. However, ARC also forecasts some ele- ments at a regional level that can be used for regional controls, such as the average number of workers within a household and the size of age cohorts. The PopSyn creates a synthetic population for a base year and for each forecast year. There are two key differ- ences between the base year and the forecast year. The initial distribution for the base year comes from PUMS, whereas for the forecast year it comes from the base- year distribution. The controls for the base year come from census tables, but for the forecast years they come from the land use forecasts. In both cases, the PopSyn pro- duces a synthetic population, and it also produces a val- idation report that compares synthetic population characteristics with known characteristics. To validate the synthesizerâs ability to generate a fore- cast population, ARC uses Year 2000 as the base year and validates a back- cast to 1990. The initial distribution comes from the base- year PopSyn. The controls then emu- late a 1990 forecast data set and synthesize a 1990 popu- lation, which is then compared with 1990 census, testing the ability to generate a synthetic population with limited forecast information. In this process, it is assumed that the forecast input, though limited in amount and detail, is cor- rect. In other words, the procedure validates the synthe- sizer but does not validate the land use model forecasts. The procedure validates by calculating both aggregate characteristics of the synthetic population and the same characteristics directly from the detailed census tables. It then compares them to see how well they match. There are four levels of geographic aggregation: tract, Public Use Microdata Area (PUMA), county, and supercounty. Reports are then repeated for multiple synthetic popula- tions to identify the variability caused by the Monte Carlo draws used in the synthesizer. As for software, it is object- oriented Java, Version 1.5, and consists entirely of subprograms called classes. Each class consists of member objects (that is, the information it holds) and methods (functions it can accomplish). Each class can be individually coded and tested. The PopSyn has four major groups of classes tied together by PopSyn class. DEVELOPMENT STATUS AND VALIDATION OBJECTIVES The initial programming of the ARC PopSyn is complete. Some improvements are known to be needed, including (a) improving the quality of the rounding procedure used after iterative proportional fitting (IPF) before drawing the households from PUMS, (b) enhancing user friendli- ness, and (c) adjusting the PopSyn to accommodate the recently expanded 20-county geographic scope. Enhancements would also be advisable to take advan- tage of enhanced inputs that may become available from the economic and land use models. Through use of the current synthesizer, base- year and back- cast synthesis have been tested, and preliminary validation results have been produced. The ARC PopSyn allows the user to implement a vari- ety of versions without reprogramming. For initial test- ing and validation, three versions were created, the simplest with 52 household demographic categories and the others with 128 and 316 categories, respectively. As more categories are used, more detail can be used from the census tables (base year) or ARC demographic and land development forecasts (forecast year) to control the synthesis procedure so that more household attributes should be synthesized precisely. However, the computa- tion takes longer; an increase in the number of sparsely populated categories causes more rounding error; and the use of regional values and averages for the additional controls might increase the noise and introduce bias. So one of the primary purposes of the validation is to choose the best version of household categories; preliminary conclusions are reported below. The three versions are shown in Table 1, with their number of categories within each of six dimensions. They will be identified subse- quently by their overall number of categories (e.g., Ver- sion 52). Validation allows better understanding of the level of geographic detail at which the aggregate population attributes can be trusted and which household variables are synthesized well enough to be used in the travel fore- casting models. Results of this analysis are reported later. Also reported is the testing of other setup parameters, including the convergence criterion for IPF and the aggregation level used in the seed distribution for the forecast year. Validation can also be used to evaluate the level of variation in results that is caused by the stochastic nature of the simulation procedure used to generate the syn- thetic population. Several base- year runs have been made 55VALIDATION OF ATLANTA, GEORGIA, REGIONAL COMMISSION POPULATION SYNTHESIZER TABLE 1 Three Basic PopSyn Versions Number of Categories Dimension Simple Middle Complex Overall 52 128 316 Household income 4 4 4 Household size 4 5 5 Number of workers in household 4 4 4 Family or nonfamily 1 2 2 Age of householder 1 1 2