Editing Land Use Data

Land-use inputs to the model (households, employment estimates, school enrollment, park space, etc) is stored in the Mainzone (MAZ) Attributes table. Editing this table is fairly straightforward; simply replace the existing value with a new value for the MAZs that you want to change.

However, you should be aware of several requirements:

1. Blanks are not allowed by the ABM in numeric fields. Zeros need to be used instead of blanks. This is an important warning because Visum will allow blanks to exist, but the ABM code does not correctly process blanks, so any non-value numeric cell needs to be input as a zero not a blank.
2. If you edit either the grade school or high school district fields (ECH_DIST and HCH_DIST respectively), you must make sure that there is at least some grade school and high school enrollment in each district. If not, students in those districts will not have any available school locations for their school tours, and the ABM will throw an error in school location choice.
3. If you modify the number of households in an MAZ, you will need to run the population synthesizer to create a new or updated synthetic population. The population synthesizer is described Running the Population Synthesizer.
4. Significantly increasing or decreasing employment may also require re-running the population synthesizer, since the total number of workers in the synthetic population should be consistent with the new employment totals. Otherwise the employment change will result in a re-distribution of work travel rather than generate more or less work travel. In other words, it is important when using an ABM to ensure that some reasonable balance between the number of workers and the coded employment exists between scenarios. The ABM does not correct for these imbalances, so the responsibility falls on the analyst operating the model to review and ensure that the balance makes sense.

Additional Reminders for School Enrollment

The ABM only differentiates K-8 versus high school, not K-5 and then 6-8. Keep in mind that the synthetic population isn’t trying to match the proportion of 6th versus 7th graders (populations by grade) either. This is a regional travel model; don’t expect it to match auto trips to any specific elementary school perfectly. It’s more important to think of this as getting the total enrollment relatively consistent with the size of the school; more enrollment means proportionally more trips.

Similarly, one shouldn't worry too much about K-12 schools that are only part-time. The model has home-schooled kids explicitly represented; generally the proportion of kids who are home-schooled would stay constant into the future, which is a good assumption to make in the absence of any data and safe given that we don’t want to change what we don’t know. The model doesn't capture their part-time school travel explicitly but the related employment for these schools would attract some non-school trips, which is good enough.

The ABM allows full-time private schools to be modeled, and if the data is available (location and enrollment by K-8 versus 9-12) and it’s 10% or more of the student population/total enrollment, it is suggested to include them explicitly in the land-use data, and keep their enrollment constant into the future, since the larger private schools attract a fair number of peak period trips.

For future year scenarios, it is important to try to anticipate the location of potential future schools, at least generally, since enrollment fluctuates annually anyway and is far less certain than base-year enrollment. The issue is that future scenarios can include a lot of new development in the urban growth areas, with no commensurate increases in enrollment or new schools to serve that development. This leads to unrealistic future-year increases in school trip lengths simply because there’s no local schools coded for synthetic students to attend.