2

I have taken a step interpolation and heat map of the data point shapefile (building Centroid) with QGIS ver-2.8.1, as shown below.

My question are:

  1. What is the difference of the two methods?
  2. How to explain the results?(Interpolation and heat Map of points) enter image description here
Improve this question
5
  • Just to clarify the question - are the two rasters created simply a count of the number of properties (building centroids?) or is there other data being represented (i.e. building prices)? If its the former, what you're producing is essentially a surface of property density - and that is the key aspect, both approaches are giving you a count of properties at the relative bandwidth of the interpolation / heatmap (the radius) ... with some smoothing of counts. So the bandwidth is the starting point of your explanation ... x properties per km (for example). – Andrew Tice May 28 '15 at 23:10
  • ...also, I'd check the interpolation process you used. There appears to be a concentration of centroids in the lower south west which isn't being picked up. Looks like you may have generated the raster clipped to the administrative boundaries, that might cause this artifact. Try running the interpolation unclipped and then clipping the resulting raster. – Andrew Tice May 28 '15 at 23:15
  • Thank you PolyGeo.Building shp sourced from the open street map. polygon shape file centroid point of making the building was done, from the centroid point is done by heat map and interpolation methods, such as picture attached. I can not interpretation the results of the two methods. – Syafruddin Rauf May 28 '15 at 23:23
  • OK (although I'm not Poly!). You are creating a raster surface which is showing you the concentration of centroids - both the heat map and the interpolation values are an averaged count of the number of centroids falling within the raster cell size you specify (see Underdark's answer here: gis.stackexchange.com/questions/30252/…). What you are showing in the map is the smoothed average number of properties in the bandwidth / radius you set. – Andrew Tice May 29 '15 at 0:22
  • You need to think a little about what you want the maps to show - are you trying to illustrate property density in the Netherlands, what property density are you wanting to demonstrate (i.e. properties per square kilometer)... that will help you think about how you will explain the results – Andrew Tice May 29 '15 at 0:25
0

Interpolation (which one of the methods of interpolation? there are a lot of them) is a grid of VALUES (not number of points) of the points you have. On the other hand Heat Map is just a density grid (only the density of points per area unit that you set).

Interpretation of the Heat Map is easy: red areas have higher density of buildings per area unit. (We do not know the area unit, you set it in QGIS when you used the method). The Interpolation grid on the other hand is not interpretable since the legend doesn't provide us neither the value that was used nor the interpolation method (which is necessary to provide in a scientific output, since there are numerous available). Guessing from the different values in the two maps, the value of the 1st interpolation grid is not the density of the buildings. (a little warning: when using an interpolation method, some users provide as a value a default field set by the method or the software used - in many cases it's the ID or FID field of a shapefile - in this case the resulting grid is useless and doesn't represent something of value)

Improve this answer

Not the answer you're looking for? Browse other questions tagged or ask your own question.