I have two shapefiles that represent two slightly different zonings of the same area (municipality zoning and transport zones), about 3000 polygons each. Several smaller municipalities may form a single transport zone, while a larger municipality may be split into several transport zones.
To these shapefiles, two datasets are linked: Traffic flow is available for the municipality zoning, and travel time is available between transport zones. I need to approximate the travel time between municipalities. My idea is to estimate the travel time between two municipalities as the weighted average of the travel times between all related transport zones. The weighting would happen according to the share of area that a transport zone contributes to the corresponding municipality.
I have come up with an R code that seems to do the right thing. However, I've also read about the MAUP (Modifiable Areal Unit Problem), and it seems to apply here. My questions are:
- Is the construction of a spatial mapping between incompatible zonings a common practical problem?
- If yes, how is this problem solved in practice -- in R, QGIS, ArcGIS (How to merge two 'incompatible' polygon layers in ArcGIS?) or using other tools?
- Is there a more sensible weighting, perhaps one that incorporates centrality (i.e., those transport zones that are located more centrally would have more influence on the result)?
References to academic publications are more than welcome.