search

cpmodel / FTT_StandAlone

Future Technology Transformation models
GNU General Public License v3.0
10 stars 1 forks source link

Discontinuity in TEWK data FTT:Tr #105

Open Femkemilene opened 6 months ago

Femkemilene commented 6 months ago

The historical shares data has a discontinuity between 2017 and 2018. This is particularly acute for the US, but is also visible in other countries. I'm computing the stock of cars by age group, and this poses some problems there:

image

pv-camecon commented 6 months ago

Any chance this has to do with the total fleet size estimates, i.e. is there a jump going from history to simulation period?

Femkemilene commented 5 months ago

The FTT:Tr data is now updated to 2022, so it's not a jump from history to simulation. The issue is RFLT, which sees this jump in the input files. I imagine there may have been reclassification of what constitutes a car (given that the US has loads of van-like huge cars, that may have been freight before). Where do we get RFLT?

pv-camecon commented 5 months ago

I believe RFLT is based on a Gompertz equation that takes in GDP per cap, saturation levels of vehicle ownership, and estimated parameters. We plug those projections in as an exogenous input.