Incorrect results in County Scale & Temporal Preagg without passenger cars

[danielbizercox]

If you run at County Scale with temporal pre-aggregation (e.g., Annual Preagg), you may get incorrect results if you do not include passenger cars in the run. This is because MOVES is hardcoded to use passenger car temporal distributions (e.g., hour VMT distributions) as a weighting factor when preforming temporal preaggregation. If these distributions are missing in the input database, the weighting factors will not be calculated correctly. Pollutants affected by temperature adjustments and AC effects are particularly affected by this bug.

Ideally, MOVES would calculate a weighted average of all the source types when performing pre-aggregation. Unfortunately, this is a complicated and time consuming calculation due to the number of ways VMT can be input to model at County Scale, as well as how the relative mileage accumulation rates and age distributions play a role in how the source types should be properly weighted together.

Since pre-aggregation by definition loses modeling precision, we plan to maintain the current algorithm of using a single source type as a weighting factor. However, to fix the issue when passenger cars are not included in the run, we plan to update the pre-aggregation algorithms to choose another source type if necessary.

This does not affect County Scale runs with no pre-aggregation options selected.

This update will have a very small effect on Default Scale runs with temporal pre-aggregation if passenger cars are not included in the run. This is because MOVES always used passenger car weighting factors, regardless if they were included in the run, and now MOVES will choose another source type if passenger cars are not included in the run.

We intend to include this update in the next patch to MOVES. A workaround until we release the next patch is to include passenger cars in runs that use pre-aggregation at County Scale.