I was wondering is it possible to map Traditional Knowledge (TK) of a bigger region (say staus of TK) on the basis of sampling in a smaller region through Arc GIS and modelling? Say we sampled in few villages of a block and we want to show a heat map of TEK for whole block or district. If yes which tool would be the best and what parameters (like distance from road head, availability of resources, population etc.) should always be considered in the same. It would be great if someone can give a step-wise illustration or recommend some good papers on the same.
Thanks in advance.
If you know the characteristics of the region involved in the sample for example socioeconomic variables then one can use TK as the dependent variable and run a GWR model in ArcGIS Geograhically weighted regression.
Naveen Chandra Joshi, what do you mean by traditional knowledge? Is it spatially located knowledge? I do not think how to map (with reasonable resolution) traditional knowledge of the traditional Oceania sailors, for example. Its covered just whole Pacific and Indian Oceans. Or traditional knowledge of nomad tribes? Or traditional spiritual knowledge? Maybe research task should be more limited.
In any way, any GIS could be useful for that purpose. If you are interesting in the unveiling of global superstructure of TK, Digital Earth approach for visualisation is the best (Google Earth, for example). For example, two megalithic complexes with very different history, morphologically identified like twins (Machu Picchu and Gunung Padang), located precisely on the opposite longitudes. Understanding this kind of interconnections between TK in global scale requires appropriate geospatial tools. Classic map-based GIS are not very convenient for that purposes. Of course both approaches (GIS and Digital Earth) should be combined.
О взаимном расположении мегалитических комплексов Гунунг-Пад...
Dear Naveen,
I'm full agreement with Eugene Eremchenko : you should first define the spatial units that adequately discribe the studied phenomenon.
Simple heat maps (i.e. kernel density estimation, KDE) or basic interpolation techniques (e.g. inverse distance weighted, IDW; thin plate spline, TPS) are not able to use information from auxiliary variables, like population or distance from road. Besides geographically weighted regression that Srini Vasan mentioned, you can use regression kriging (RK, a.k.a. kriging with external drift, KED) as well.
HTH,
Ákos
- Badges
- Science topic
- Similar topics
- Geoscience
- Cartography
- Mapping
- Maps