There are four steps required to build the cleaned input data for use in the model:
- Partition
- Calibrate
- Share
- Disaggregate
The buildstream process and all of the notation included in the SLiDE documentation here is meant to replicate the WiNDC buildstream. For more information, please reference:
- Thomas F. Rutherford and Andrew Schreiber, "Tools for Open Source, Subnational CGE Modeling with an Illustrative Analysis of Carbon Leakage," J Global Econ Anal 4(2): 1-66.
build
Divide BEA supply/use data into parameters.
partition
calibrate
Divide data into regional component. This will guide how to break the national data into regional components by state or county.
share
disagg