Hi Tom, in Atlanta, we have developed an algorithm for TAZ disaggregation, which basically consists in a GIS- based regional control forecast allocation tool, currently operable as an ESRI ArcMap 9.3 extension with the requirement of a Spatial Analyst license, and written in VB.net. The tool uses a tab structure in the form of a GUI interface to load data, to set and calibrate the model parameters, to run the spatial allocation model, and to generate an animation of results.  It furthermore provides an optimal weighting matrix, for a predetermined set of spatial influence layers, that most replicates historical time series spatial growth patterns.  Currently, our TAZ-D extension makes the user adjust the shares (by Superdistrict area) of control totals of employment and households manually, with subsequent machine interpolation of the shares matrix. This interpolation function tool will be improved so that rebalancing of shares between Superdistricts occurs in a more automated fashion.

 

Additional details are available at:

 

http://www.ampo.org/assets/604_arctazdisaggregator.pdf

 

Animation of results & visualization examples are available at:

 

http://www.youtube.com/user/ARCModeling

 

Thanks,

 

Guy

 

Guy Rousseau

Traffic Modeling Manager

Atlanta Regional Commission

grousseau@atlantaregional.com

 

 

 

From: ctpp-news-bounces@chrispy.net [mailto:ctpp-news-bounces@chrispy.net] On Behalf Of Tom Krenzke
Sent: Tuesday, February 23, 2010 6:13 PM
To: ctpp-news@chrispy.net
Subject: [CTPP] Aggregating Census TAZs

 

Has anyone tried aggregating Census TAZs for the nation into something larger (e.g.,  20K in population) that still makes sense for transportation planning (e.g., corridors, MIS, Districts, planning areas etc.)? Or do you have an algorithm that combines TAZs?  If you do, would you have a correspondence file to share with us, and the purpose for which this aggregation was made?

 

Thanks,

 

--Tom Krenzke

Westat